General Mathematics VCE Units 3&4 Second Edition

GENERAL
MATHEMATICS
VCE UNITS 3 & 4
CAMBRIDGE SENIOR MATHEMATICS VCE

SECOND EDITION
PETER JONES | KAY LIPSON | MICHAEL EVANS
ROSE HUMBERSTONE | PETER KARAKOUSSIS | KYLE STAGGARD
GENERAL
MATHEMATICS
VCE UNITS 3 & 4
CAMBRIDGE SENIOR MATHEMATICS VCE

SECOND EDITION
PETER JONES | KAY LIPSON | MICHAEL EVANS
ROSE HUMBERSTONE | PETER KARAKOUSSIS | KYLE STAGGARD
ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Photocopying is restricted under law and this material must not be transferred to another party.
Shaftesbury Road, Cambridge CB2 8BS8EA, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
Cambridge University Press is part of Cambridge University Press & Assessment, a department
of the the University of Cambridge.
We share the University’s mission to contribute to society through the pursuit of education,
learning and research at the highest international levels of excellence. It furthers the University’s
mission by disseminating knowledge in the pursuit of education, learning and research at the
highest international levels of excellence.
www.cambridge.org
First Edition c Peter Jones, Michael Evans, Kay Lipson and Kyle Staggard 2016
Second Edition c Peter Jones, Kay Lipson, Michael Evans, Rose Humberstone, Peter
Karakoussis and Kyle Staggard 2023
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2016
Second Edition 2023
20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Cover design by Denise Lane (Sardine)
Text design by Jane Pitkethly
Typeset by Integra Software Services Pvt. Ltd
Printed in China by C & C Offset Printing Co. Ltd.
A catalogue record for this book is available from the National Library of
Australia at www.nla.gov.au
ISBN 978-1-009-11041-9 Paperback
Additional resources for this publication at www.cambridge.edu.au/GO
Reproduction and Communication for educational purposes
The Australian Copyright Act 1968 (the Act) allows a maximum of
one chapter or 10% of the pages of this publication, whichever is the greater,
to be reproduced and/or communicated by any educational institution
for its educational purposes provided that the educational institution
(or the body that administers it) has given a remuneration notice to
Copyright Agency Limited (CAL) under the Act.
For details of the CAL licence for educational institutions contact:
Copyright Agency Limited
Level 12, 66 Goulburn Street
Sydney NSW 2000
Telephone: (02) 9394 7600
Facsimile: (02) 9394 7601
Email: [email protected]
Reproduction and Communication for other purposes
Except as permitted under the Act (for example a fair dealing for the
purposes of study, research, criticism or review) no part of this publication
may be reproduced, stored in a retrieval system, communicated or
transmitted in any form or by any means without prior written permission.
All inquiries should be made to the publisher at the address above.
Cambridge University Press & Assessment has no responsibility for the persistence or
accuracy of URLS for external or third-party internet websites referred to in
this publication and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate. Information regarding prices, travel
timetables and other factual information given in this work is correct at
the time of first printing but Cambridge University Press & Assessment does not guarantee
the accuracy of such information thereafter.
Cambridge University Press & Assessment acknowledges the Australian Aboriginal and Torres
Strait Islander peoples of this nation. We acknowledge the traditional custodians of the lands on
which our company is located and where we conduct our business. We pay our respects to
ancestors and Elders, past and present. Cambridge University Press & Assessment is committed
to honouring Australian Aboriginal and Torres Strait Islander peoples’ unique cultural and
ISBN 978-1-009-11041-9 © seas
spiritual relationships to the land, waters and Peter Jones
and their et al contribution
rich 2023 to society. Cambridge University Press
Contents
Introduction and overview vii
Acknowledgements xii
1 Investigating data distributions

1A . . . . . . . . . . . . . . . . . . . . . . . . .
Types of data
1
2
1B Displaying and describing the distributions of categorical
variables . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1C Displaying and describing numerical data . . . . . . . . . . . 13
1D Dot plots and stem plots . . . . . . . . . . . . . . . . . . . . 28
1E Using a logarithmic (base 10) scale to display data . . . . . . 34
1F Measures of centre and spread . . . . . . . . . . . . . . . . 43
1G The five-number summary and the boxplot . . . . . . . . . . 61
1H The normal distribution and the 68–95–99.7% rule . . . . . . 76
1I Populations and samples . . . . . . . . . . . . . . . . . . . 89
Review of Chapter 1 . . . . . . . . . . . . . . . . . . . . . . 90
2 Investigating associations between two variables. . . . . . . . . . . 103

2A Bivariate data – Classifying the variables 104
2B Investigating associations between categorical variables . . . 108
2C Investigating the association between a numerical and a
categorical variable . . . . . . . . . . . . . . . . . . . . . . 122
2D Investigating associations between two numerical variables . 130
2E How to interpret a scatterplot . . . . . . . . . . . . . . . . . 134
2F Strength of a linear relationship: the correlation coefficient . . 140
2G The coefficient of determination . . . . . . . . . . . . . . . . 146
2H Correlation and causality . . . . . . . . . . . . . . . . . . . 151
2I Which graph? . . . . . . . . . . . . . . . . . . . . . . . . . 155

iv Contents
3 Investigating and modelling linear associations

3A Fitting a least squares regression line to numerical data . . . 169
168
3B Using the least squares regression line to model a relationship

between two numerical variables . . . . . . . . . . . . . . . 179
3C Conducting a regression analysis using data . . . . . . . . . 195
4 Data transformation
4A The squared transformation. . . . . . . . . . . . . . . . .
211
212
4B The log transformation . . . . . . . . . . . . . . . . . . . . 221
4C The reciprocal transformation . . . . . . . . . . . . . . . . 230
4D Choosing and applying the appropriate transformation . . . . 239
5 Investigating and modelling. time

5A
series
. . . . . . . . . . . . . . . . . . . . . .
Time series data
252
253
5B Smoothing a time series using moving means . . . . . . . . . 268
5C Smoothing a time series plot using moving medians . . . . . . 277
5D Seasonal indices . . . . . . . . . . . . . . . . . . . . . . . 283
5E Fitting a trend line and forecasting. . . . . . . . . . . . . . 295
6 Revision: Data analysis

6A Exam 1 style questions: Univariate data . . . . . . . . . . . . 311
310
6B Exam 1 style questions: Associations . . . . . . . . . . . . . 316

6C Exam 1 style questions: Regression and data transformation . 320
6D Exam 1 style questions: Time series . . . . . . . . . . . . . . 324
6E Exam 2 style questions . . . . . . . . . . . . . . . . . . . . 328
7 Modelling growth and decay using recursion . . . . . . . . . . . . . 333

7A Sequences and recurrence relations 334
7B Modelling linear growth and decay . . . . . . . . . . . . . . 340
7C . . . . . . . 350
Using an explicit rule for linear growth or decay
7D Modelling geometric growth and decay . . . . . . . . . . . . 359
7E Using an explicit rule for geometric growth or decay . . . . . 366
7F Interest rates over different time periods and effective inter-
est rates . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
8 Reducing balance loans, annuities and investments

8A Combining linear and geometric growth or decay to model com-
391
pound interest investments with additions to the principal . . 392

Contents v
8B Using recurrence relations to analyse and model reducing

balance loans and annuities . . . . . . . . . . . . . . . . . . 400
8C Amortisation tables . . . . . . . . . . . . . . . . . . . . . . 407
8D Analysing financial situations using amortisation tables . . . . 416
8E Using a finance solver to find the balance and final payment . . 422
8F Using a finance solver to find interest rates, time taken and
regular payments . . . . . . . . . . . . . . . . . . . . . . . 432
8G Solving harder financial problems. . . . . . . . . . . . . . 440
8H Interest-only loans . . . . . . . . . . . . . . . . . . . . . . 446
8I Perpetuities . . . . . . . . . . . . . . . . . . . . . . . . . . 451
9 Revision: Recursion and financial. modelling

9A . . . . . . . . . . . . . . . . . . .
Exam 1 style questions
465
466
9B Exam 2 style questions . . . . . . . . . . . . . . . . . . . . 469
10 Matrices 10A What is a matrix? . . . . . . . . . . . . . . . . . . . . . . . 473

472
10B Using matrices to represent information . . . . . . . . . . . 485

10C Matrix arithmetic: addition, subtraction and scalar
multiplication . . . . . . . . . . . . . . . . . . . . . . . . . 491
10D Matrix arithmetic: the product of two matrices . . . . . . . . 499
10E Matrix inverse, the determinant and matrix equations . . . . . 513
10F Binary, permutation and communication matrices . . . . . . 523
10G Dominance matrices . . . . . . . . . . . . . . . . . . . . . 531
Review of Chapter 10 . . . . . . . . . . . . . . . . . . . . . 538
11 Transition matrices and Leslie matrices

11A Transition matrices - setting up a transition matrix . . . . . . 551
550
11B . . . . . . . . .
Interpreting transition matrices . . . . . . 555
11C Transition matrices – using recursion . . . . . . . . . . . . . 559
11D Transition matrices – using the rule Sn+1 = TSn + B . . . . . . . 570
11E Leslie matrices . . . . . . . . . . . . . . . . . . . . . . . . 574
12 Revision: Matrices
12A . . . . . . . . . . . . . . . . . . . . 596
Exam 1 type questions
595
13 Graphs, networks and trees: travelling

13A
and connecting problems
. . . . . . . . . . . . . . . . . . . . .
Graphs and networks
608
609
13B Adjacency matrices . . . . . . . . . . . . . . . . . . . . . . 619

vi Contents
13C Exploring and travelling . . . . . . . . . . . . . . . . . . . . 624

13D Weighted graphs and networks . . . . . . . . . . . . . . . . 633
13E Dijkstra’s algorithm . . . . . . . . . . . . . . . . . . . . . . 637
13F Trees and minimum connector problems . . . . . . . . . . . 640
14 Flow, matching and scheduling

14A
problems
. . . . . . . . . . . . . . . . . . . . . . . .
Flow problems
659
660
14B . . . .
Matching and allocation problems . . . . . . . . . . 674
14C Precedence tables and activity networks . . . . . . . . . . . 683
14D Scheduling problems . . . . . . . . . . . . . . . . . . . . . 693
14E Crashing . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
15 Revision: Networks and decision. mathematics

15A . . . . . . . . . . . . . . . . . . .
727
728
16 Revision of Chapters 1–15 . . . . . . . . . . . . . . . . . . . . . . . 741

16A Exam 1 questions 742
16B Exam 2 questions . . . . . . . . . . . . . . . . . . . . . . . 753
Glossary 765
Answers 774
Online appendices accessed through the Interactive Textbook or PDF Textbook
Appendix A Guide to the TI-Nspire CAS calculator in VCE mathematics

Appendix B Guide to the Casio ClassPad II CAS calculator in VCE mathematics

Introduction and overview
Cambridge General Mathematics VCE Units 3&4 Second Edition provides a complete
teaching and learning resource for the VCE Study Design to be first implemented in 2023.
It has been written with understanding as its chief aim, and with ample practice offered
through the worked examples and exercises. The work has been trialled in the classroom, and
the approaches offered are based on classroom experience and the responses of teachers to
earlier editions of this book and the requirements of the new Study Design.
The textbook and its accompanying digital resources provide comprehensive coverage of the
assumed knowledge and skills required.
General Mathematics Units 3 and 4 provide the following four Areas of Study:
Data analysis
Recursion and financial modelling
Matrices and their applications
Networks and decision mathematics
Separate revision chapters containing Multiple-choice (Exam 1-style questions) and

Written-response (Exam 2-style questions) are included for each study area. Additional
Multiple-choice and Written-response questions, covering the year’s course, are included in
the final revision chapter.
Other key features of Cambridge General Mathematics VCE Units 3&2 Second Edition
For each topic within a chapter:
The topic starts with a clear outline of its Learning intentions
Exam 1-style questions are included at the end of every exercise.
The Review section at the end of each chapter includes:

Key ideas and a chapter summary
A Skills checklist, which provides a question to check mastery of each skill listed in the
Learning intentions
Multiple-choice and written-response questions review the skills and concepts of the
chapter.
The textbook is supported by an extensive range of online teacher and student resources, as
outlined in the ‘How to use this resource’ section on the following pages.
The TI-Nspire calculator examples and instructions have been completed by Peter Flynn,
and those for the Casio ClassPad by Mark Jelinek, and we thank them for their helpful
contributions.

Overview of the print book
1 Learning Intentions added at the beginning of each section to clearly outline the goals of
the lesson.
2 Graded step-by-step worked examples with precise explanations (and video versions)
encourage independent learning, and are linked to exercise questions.
3 Exam 1-style questions added to every exercise.
4 Additional linked resources in the Interactive Textbook are indicated by icons, such as
skillsheets and video versions of examples.
5 End-of-chapter content includes a Skills Checklist and a summary of the key terminology.
6 Chapter reviews contain multiple-choice and written-response questions.
7 Revision chapters include Exam 1-style and Exam 2-style questions, providing
comprehensive revision and preparation for assessment.
8 The glossary includes page numbers of the main explanation of each term.
Numbers refer to descriptions above.

Introduction and overview ix
Overview of the
downloadable PDF
textbook
9 The convenience of a downloadable PDF
textbook has been retained for times when
users cannot go online.
10 PDF annotation and search features are
enabled.
10 10
Overview of the Interactive Textbook
The Interactive Textbook (ITB) is an online HTML version of the print textbook powered
by the HOTmaths platform, included with the print book or available as a separate purchase.
11 The material is formatted for on screen use with a convenient and easy-to-use navigation
system and links to all resources.
12 Workspaces for all questions, which can be enabled or disabled by the teacher, allow
students to enter working and answers online and to save them. Input is by typing, with the
help of a symbol palette, handwriting and drawing on tablets, or by uploading images of
writing or drawing done on paper.
13 Self-assessment tools enable students to check answers, mark their own work, and rate
their confidence level in their work. This helps develop responsibility for learning and
communicates progress and performance to the teacher. Student accounts can be linked to
the learning management system used by the teacher in the Online Teaching Suite, so that
teachers can review student self-assessment and provide feedback or adjust marks.
14 All worked examples have video versions to encourage independent learning.
15 Worked solutions are included and can be enabled or disabled in the student ITB accounts
by the teacher.
16 An expanded and revised set of Desmos interactives and activities based on embedded
graphics calculator and geometry tool windows demonstrate key concepts and enable
students to visualise the mathematics.
17 The Desmos graphics calculator, scientific calculator, and geometry tool are also
embedded for students to use for their own calculations and exploration.
18 Revision of prior knowledge is provided with links to diagnostic tests and Year 10
HOTmaths lessons.
19 Quick quizzes containing automarked multiple-choice questions have been thoroughly
expanded and revised, enabling students to check their understanding.
20 Definitions pop up for key terms in the text, and are also provided in a dictionary.
21 Messages from the teacher assign tasks and tests.

x Introduction and overview
INTERACTIVE TEXTBOOK POWERED BY THE HOTmaths

PLATFORM
A selection of features is shown. Numbers refer to the descriptions on pages xi–xii.
HOTmaths platform features are updated regularly
WORKSPACES AND SELF-ASSESSMENT

Introduction and overview xi
Overview of the Online Teaching Suite powered by the

HOTmaths platform
The Online Teaching Suite is automatically enabled with a teacher account and is integrated
with the teacher’s copy of the Interactive Textbook. All the teacher resources are in one place
for easy access. The features include:
22 The HOTmaths learning management system with class and student analytics and reports,
and communication tools.
23 Teacher’s view of a student’s working and self-assessment which enables them to modify
the student’s self-assessed marks, and respond where students flag that they had diffculty.
24 A HOTmaths-style test generator.
25 An expanded and revised suite of chapter tests, assignments and sample investigations.
26 Editable curriculum grids and teaching programs.
27 A brand-new Exam Generator, allowing the creation of customised printable and online
trial exams (see below for more).
More about the Exam Generator

The Online Teaching Suite includes a comprehensive bank of VCAA exam questions,
augmented by exam-style questions written by experts, to allow teachers to create custom trial
exams.
Custom exams can model end-of-year exams, or target specific topics or types of questions
that students may be having difficulty with.
Features include:
Filtering by question-type, topic and degree of difficulty

Searchable by key words
Answers provided to teachers
Worked solutions for all questions
VCAA marking scheme
Multiple-choice exams can be auto-marked if completed online, with filterable reports
All custom exams can be printed and completed under exam-like conditions or used as
revision.

Acknowledgements
The author and publisher wish to thank the following sources for permission to reproduce
material:
Cover: c Getty Images / m_pavlov
Images: c Getty Images / OsakaWayne Studios, Chapter 1 Opener / Andriy Onufriyenko,

Chapter 2 Opener / MirageC, Chapter 3 Opener / MR.Cole_Photographer, Chapter 4 Opener /
Chapter 4 Opener, Chapter 5 Opener / oxygen, Chapter 6 Opener / Andriy Onufriyenko,
Chapter 7 Opener / Gremlin, Chapter 8 Opener / DENIS Starostin, Chapter 9 Opener / Ilona
Nagy, Chapter 10 Opener / carloscastilla, Chapter 11 Opener / Mrljanica, Chapter 12 Opener
/ PM Images, Chapter 13 Opener / matdesign24, Chapter 14 Opener / SEGU JEON, Chapter
15 Opener / Photo by Alex Tihonov, Chapter 16 Opener.
Every effort has been made to trace and acknowledge copyright. The publisher apologises for
any accidental infringement and welcomes information that would redress this situation.

Chapter
1
Investigating data
distributions
Chapter questions
I What is categorical data?
I What is numerical data?
I How do we summarise and display categorical data?
I How do we use the distribution of a categorical variable to answer
statistical questions?
I What is a dot plot?
I What is a stem plot?
I What is a histogram?
I What is a boxplot?
I What are summary statistics, and how do we choose which ones to use?
I How do we use the distribution of a numerical variable to answer statistical
questions?
I What is the normal distribution?
I What is the 68-95-99.7% rule and why is it useful?
I What are standardised values and why are they useful?
We can think of data as factual information about a person, object or situation which has
been collected and recorded. In General Mathematics Units 1&2 we learned a range of
statistical procedures to help us analyse such sets of data. In this chapter, we will review
and extend our knowledge of those procedures for data which has been collected from a
single variable, which is called univariate data.

2 Chapter 1 Investigating data distributions
1A Types of data
Learning intentions
I To be able to classify data as categorical or numerical.
I To be able to further classify categorical data as nominal or ordinal.
I To be able to further classify numerical data as discrete or continuous.
A group university of students were asked to complete a survey, and the information
collected from eight of these students is shown in the following table:
Study mode Fitness level

(C on-campus, (1 high, Pulse rate
Height (cm) Weight (kg) Age (years) O online ) 2 medium, 3 low) (beats/min)
173 57 18 C 2 86
179 58 19 C 2 82
167 62 18 C 1 96
195 84 18 O 1 71
173 64 18 C 3 90
184 74 22 O 3 78
175 60 19 O 3 88
140 50 34 C 3 70
Since the answers to each of the questions in the survey will vary from student to student,
each question defines a different variable namely:
height (in centimetres) study mode (C = on-campus, O = online)

weight (in kilograms) fitness level (1 = high, 2 = medium, 3 = low)
age (in years) pulse rate (beats/minute).
The values we collect about each of these variables are called data.
Types of variables
Variables come in two general types, categorical and numerical:
Categorical variables
Categorical variables classify or name a quality or attribute– for example, a person’s

eye colour, study mode, or fitness level.
Data generated by a categorical variable can be used to organise individuals into one of
several groups or categories that characterise this quality or attribute.

1A Types of data 3
For example, a ‘C’ in the Study mode column indicates that the student studies on-campus,
while a ‘3’ in the Fitness level column indicates that their fitness level is low.
Categorical variables can be further classified as one of two types: nominal and ordinal.
Nominal variables
Nominal variables have data values that are simply names.
The variable Study mode is an example of a nominal variable because the values of the
variable, on-campus or online, simply name the group to which the students belong.
Ordinal variables
Ordinal variables have data values that can be used to both name and order.
The variable Fitness level is an example of an ordinal variable. The data generated by
this variable contains two pieces of information. First, each data value can be used to
group the students by fitness level. Second, it allows us to logically order these groups
according to their fitness level – in this case, as ‘low’, ‘medium’ or ‘high’.
Numerical variables
Numerical variables have data values which are quantities, generally arising from
counting or measuring.
For example, a ‘179’ in the Height column indicates that the person is 179 cm tall, while an
‘82’ in the Pulse rate column indicates that they have a pulse rate of 82 beats/minute.
Numerical variables can be further classified as one of two types: discrete and continuous.
Discrete variables
Discrete variables are those which may take on only a countable number of distinct
values such as 0, 1, 2, 3, 4, . . .
Discrete random variables often arise when the situation involves counting. The number
of mobile phones in a house is an example of a discrete variable.
As a guide, discrete variables arise when we ask the question ‘How many?’
Continuous variables
Continuous variables are ones which take an infinite number of possible values, and
are often associated with measuring rather than counting.
Thus, even though we might record a person’s height as 179 cm, their height could be any
value between 178.5 and 179.4 cm. We have just rounded to 179 cm for convenience, or
to match the accuracy of the measuring device.

As a guide, continuous variables arise when we ask the question ‘How much?’
Comparing numerical and categorical variables

The interrelationship between categorical (nominal and ordinal) and numerical variables
(discrete and continuous) is displayed in the diagram below.
Nominal variable
(e.g. eye colour)
Categorical variable
Ordinal variable
(e.g. house number)
Variable
Discrete data
(e.g. number of cars in a car park)
Numerical variable
Continuous variable
(e.g. weight)
Numerical or categorical?
Deciding whether data are numerical or categorical is not an entirely trivial exercise. Two
things that can help your decision-making are:
1 Numerical data can always be used to perform arithmetic computations. This is not the
case with categorical data. For example, it makes sense to calculate the average weight of
a group of individuals, but not the average house number in a street. This is a good test to
apply when in doubt.
2 It is not the variable name alone that determines whether data are numerical or
categorical; it is also the way the data are recorded. For example, if the data for variable
weight are recorded in kilograms, they are numerical. However, if the data are recorded as
‘underweight’, ‘normal weight’, ‘overweight’, they are categorical.
Example 1 Types of data
Classify the following data as nominal, ordinal, discrete or continuous.

a The number of chocolate chips in each of 10 cookies is counted.
b The time taken for 20 students to complete a puzzle is recorded in seconds.
c Member of a football club were asked to rate how they felt about the current coach, 1=
Very satisfied, 2 = Satisfied, 3 = Indifferent, 4 = Dissatisfied, 5 = Very dissatisfied.
d Students are asked to each choose their preferred colour from the list 1= Blue, 2 =
Green, 3 = Red, 4 = Yellow.
e Students weights were classified as ‘less than 60kg’, ‘60kg - 80kg’ or ‘more than
80kg’.
Solution
a Discrete, as the number of chocolate chips will only take whole number values.
b Continuous, as the data can take any value, limited only by the accuracy to which the
time can be measured.
1A 1A Types of data 5
c Ordinal, as the numbers in this data do not represent quantities, they represent each
person’s level of approval of the coach.
d Nominal, as the numbers in this data again do not represent quantities, they represent
colours.
e Ordinal, as each student’s weight has been recorded into three categories which can
be ordered.
Skill-
sheet Exercise 1A
Types of variables: categorical or numerical

Example 1 1 Classify each of the following variables (in italics) as categorical or numerical:
a time (in minutes) spent exercising each e time spent playing computer games
day (hours)
b number of frogs in a pond f number of people in a bus
c bank account numbers g eye colour (brown, blue, green )
d height (short, average, tall) h post code
Categorical variables: nominal or ordinal

2 Classify the categorical variables identified below (in italics) as nominal or ordinal.
a The colour of a pencil
b The different types of animals in a zoo
c The floor levels in a building (0, 1, 2, 3 . . . )
d The speed of a car (less than 50 km/hr, 50 to 80 km/hr, above 80 km/hr)
e Shoe size (6, 8, 10, . . . )
f Family names
Numerical variables: discrete or continuous

3 Classify the numerical variables identified below (in italics) as discrete or continuous.
a The number of pages in a book
b The cost (to the nearest dollar) to fill the tank of a car with petrol
c The volume of petrol (in litres) used to fill the tank of a car
d The speed of a car in km/h
e The number of people at a football match
f The air temperature in degrees Celsius

6 Chapter 1 Investigating data distributions 1A

4 Respondents to a survey question “Are you concerned about climate change?” were
asked to select from the following responses
1 = not at all, 2 = a little, 3 = somewhat, 4 = extremely
The data which was collected in response to this question is:
A nominal B ordinal C discrete
D continuous E numerical
5 The variables weight (light, medium, heavy) and age (under 25 years, 25-40 years, over
40 years) are:
A a nominal and an ordinal variable respectively
B an ordinal and a nominal variable respectively
C both nominal variables
D both ordinal variables
E both continuous variables
6 Data relating to the following five variables were collected from a group of university
students:
course
study mode (1= on campus, 2 = online)
study load (1= full-time, 2-part-time)
number of contact hours per week
number of subject needed to complete degree
The number of these variables that are discrete numerical variables is:
A 1 B 2 C 3 D 4 E 5
1B Displaying and describing the distributions of

categorical variables
Learning intentions
I To be able to construct frequency and percentage frequency tables for categorical data.
I To be able to construct bar charts and percentage bar charts from frequency tables.
I To be able to interpret and describe frequency tables and bar charts.
I To be able to answer statistical questions about a categorical variable.

1B Displaying and describing the distributions of categorical variables 7
The frequency table

With a large number of data values, it is difficult to identify any patterns or trends in the raw
data.
For example, the following set of categorical data, listing the study mode (C = on-campus,
O = online) for 60 individuals, is hard to make sense of.
O O C O O O C O C C C O C O O O C C C O
C O C O C C C O C O C O C O O O C O C O
C O O O O O O C C O C O O O C O C C C O
To help make sense of the data, we first need to organise them into a frequency table.
The frequency table

A frequency table is a listing of the values a variable takes in a data set, along with how
often (frequently) each value occurs.
Frequency can be recorded as a:
number: the number of times a value occurs, or
percentage: the percentage of times a value occurs (percentage frequency):
number of times a value occurs
percentage frequency = × 100%
total number of values
Example 2 Frequency table for a categorical variable
A group of 11 preschool children were asked to choose between chocolate and vanilla
ice-cream (C = chocolate, V = vanilla):
C V V C C V C C C V V
Construct a frequency table (including percentage frequencies) to display the data.
Explanation Solution
1 Set up a table as shown.
Frequency
2 Count up the number of chocolate (6) and vanilla
flavour Number Percentage
(5), and record in the Number column.
chocolate 6 54.5
3 Add to find the total number, 11 (6 + 5).
vanilla 5 45.5
4 Convert the frequencies into percentage
frequencies. e.g. percentage chocolate= Total 11 100.0
6
× 100% = 54.5%
11
5 Finally, total the percentages and record.
Note that the total should always equal the total number of observations – in this case, 11,
and that the percentages should add to 100%. However, if percentages are rounded to one
decimal place a total of 99.9 or 100.1 is sometimes obtained. This is due to rounding error.
Totalling the count and percentages helps check on your tallying and percentaging.
The bar chart

Once categorical data have been organised into a frequency table, it is common practice to
display the information graphically to help identify any features that stand out in the data.
The statistical graph we use for this purpose is the bar chart.
The bar chart represents the key information in a frequency table as a picture. The bar chart
is specifically designed to display categorical data.
In a bar chart:
frequency (or percentage frequency) is shown on the vertical axis
the variable being displayed is plotted on the horizontal axis
the height of the bar (column) gives the frequency (number or percentage)
the bars are drawn with gaps to show that each value is a separate category
there is one bar for each category.
Example 3 Constructing a bar chart from a frequency table
The climate type of 23 countries is classified

Frequency
as ‘cold’, ‘mild’ or ‘hot’. The results are
Climate type Number Percentage
summarised in the table opposite.
Cold 3 13.0
a What is the level of measurement of the
variable climate type? Mild 14 60.9
b Construct a frequency bar chart to display Hot 6 26.1
this information. Total 23 100.0
a The data enable us to both group the countries a Ordinal
by climate type and put these groups in some sort
of natural order according to the ‘warmth’ of the
different climate types. The variable is ordinal.
b 1 Label the horizontal axis with the variable name, b 15
Climate type. Mark the scale off into three equal
intervals and label them ‘Cold’, ‘Mild’ and ‘Hot’.
10
Frequency
2 Label the vertical axis ‘Frequency’. Scale allowing

for the maximum frequency, 14.
3 For each climate type, draw a bar. There are gaps 5
between the bars to show that the categories are

separate. The height of the bar is made equal to 0
Cold Mild Hot
the frequency (given in the ‘Number’ column).
Climate type

1B Displaying and describing the distributions of categorical variables 9
The segmented bar chart

A variation on the standard bar chart is the segmented bar chart. It is a compact display that
is particularly useful when comparing two or more categorical variables.
In a segmented bar chart, the bars are stacked one on top of 25
another to give a single bar with several parts or segments. Climate
The lengths of the segments are determined by the frequencies. 20 Hot
Mild
The height of the bar gives the total frequency. A legend is
Frequency
15 Cold
required to identify which segment represents which category (see
opposite). The segmented bar chart opposite was formed from the 10
climate data used in Example 3. In a percentage segmented bar
chart, the lengths of each segment in the bar are determined by 5
the percentages. When this is done, the height of the bar is 100%.
0
Example 4 Constructing a percentage segmented bar chart from a frequency

table
The climate type of 23 countries is classified as Frequency

‘cold’, ‘mild’ or ‘hot’.
Climate type Number Percentage
Construct a percentage frequency segmented Cold 3 13.0
bar chart to display this information.
Mild 14 60.9
Hot 6 26.1
Total 23 100.0
1 In a segmented bar chart, the horizontal axis has no 100
Climate
label. 90 Hot
Mild
80
2 Label the vertical axis ‘Percentage’. Scale allowing Cold
70
for the maximum of 100 (%), Mark the scale in tens.
Percentage
60
3 Draw a single bar of height 100. Divide the bar into 50
three by inserting dividing lines at 13% and 73.9% 40
30
(13 + 60.9%).
20
4 Shade (or colour) the segments differently. 10
5 Insert a legend to identify each shaded segment by 0
climate type.
The mode
One of the features of a data set that is quickly revealed with a frequency table or a bar
chart is the mode or modal category.
The mode is the most frequently occurring value or category.

In a bar chart, the mode is given by the category with the tallest bar or longest segment.
For the previous bar charts, the modal category is clearly ‘mild’. That is, for the countries
considered, the most frequently occurring climate type is ‘mild’.
Modes are particularly important in ‘popularity’ polls. For example, in answering questions
such as ‘Which is the most watched TV station between 6:00 p.m and 8:00 p.m.?’ or ‘When
is the time a supermarket is in peak demand: morning, afternoon or night?’
Note, however, that the mode is only of real interest when a single category stands out
from the others.
Answering statistical questions for categorical variables

A statistical question is a question that depends on data for its answer.
Statistical questions that are of most interest when working with a single categorical variable
are of these forms:
Is there a dominant category into which a large percentage of individuals fall or
are the individuals relatively evenly spread across all of the categories? For example, are
the shoppers in a department store predominantly male or female, or are there roughly
equal numbers of males and females?
How many and/or what percentage of individuals fall into each category? For example,
what percentage of visitors to a national park are ‘day-trippers’ and what percentage
of visitors are staying overnight?
A short written report is the standard way to answer these questions.
The following guidelines are designed to help you to produce such a report.
Guidelines for writing a report describing a categorical variable

Briefly summarise the context in which the data were collected including the number
of individuals involved in the study.
If there is a clear modal category, ensure that it is mentioned.
Include frequencies or percentages in the report. Percentages are preferred.
If there are a lot of categories, it is not necessary to mention every category, but the
modal category should always be mentioned.
Example 5 Describing the distribution of a categorical variable
In an investigation of the variation of climate type Frequency

across countries, the climate types of 23 countries
Climate type Number %
were classified as ‘cold’, ‘mild’ or ‘hot’. The data
are displayed in a frequency table to show the Cold 3 13.0
percentages. Mild 14 60.9
Hot 6 26.1
Use the information in the frequency table to write
a concise report on the distribution of climate types Total 23 100.0
across these 23 countries.
1B 1B Displaying and describing the distributions of categorical variables 11
Solution
Report
The climate types of 23 countries were classified as being, ‘cold’, ‘mild’ or ‘hot’. The
majority of the countries, 60.9%, were found to have a mild climate. Of the remaining
countries, 26.1% were found to have a hot climate, while 13.0% were found to have a
cold climate.
Exercise 1B
Constructing a frequency table for categorical data

Example 2 1 Construct appropriately labelled frequency tables showing both frequencies and
percentage frequencies for each of the following data sets:
a Grades: A A C B A B B B B C C
b Shoe size: 8 9 9 10 8 8 8 9 8 10 12 8
Example 3 2 The following data identify the state of residence of a group of people where
1 = Victoria, 2 = South Australia and 3 = Western Australia.
2 1 1 1 3 1 3 1 1 3 3
a Is the variable state of residence, categorical or numerical?
b Construct a frequency table (with both numbers and percentages) to show the
distribution of state of residence for this group of people.
c Construct a bar chart of the percentaged frequency table.
3 The size (S = small, M = medium, L = large) of 20 cars was recorded as follows.

S S L M M M L S S M
M S L S M M M S S M
a Is the variable size in this context numerical or categorical?
b Construct a frequency table (with both numbers and percentages) to show the
distribution of size for these cars.
c Construct a percentage bar chart.
Constructing a percentage segmented bar chart from a frequency table

Example 4 4 The table shows the frequency distribution of the Place of birth Percentage
place of birth for 500 Australians.
Australia 78.3
a Is place of birth an ordinal or a nominal
Overseas 21.8
variable?
Total 100.1
b Display the data in the form of a percentage
segmented bar chart.

12 Chapter 1 Investigating data distributions 1B
5 The table records the number of new

Frequency
cars sold in Australia during the first
Type of vehicle Number Percentage
quarter of one year, categorised by type
of vehicle (private, commercial). Private 132 736
a Is type of vehicle an ordinal or a Commercial 49 109

nominal variable? Total
b Copy and complete the table giving the percentages correct to the nearest whole
number.
c Display the data in the form of a percentage segmented bar chart.
Using the distribution of a categorical variable to answer statistical questions

Example 5 6 The table shows the frequency
Frequency
distribution of school type for a number
School type Number Percentage
of schools. The table is incomplete.
Catholic 4 20
a Write down the information missing
from the table. Government 11
b How many schools are categorised as Independent 5 25
‘independent’? Total 100
c How many schools are there in total?
d What percentage of schools are categorised as ‘government’?
e Use the information in the frequency table to complete the following report
describing the distribution of school type for these schools.
Report
schools were classified according to school type. The majority of these

schools, %, were found to be . Of the remaining schools,
were while 20% were .
7 Twenty-two students were asked the question,

Frequency
‘How often do you play sport?’, with the
Plays sport Number Percentage
possible responses: ‘regularly’, ‘sometimes’
or ‘rarely’. The distribution of responses is Regularly 5 22.7
summarised in the frequency table. Sometimes 10
a Write down the information missing from Rarely 31.8
the table. Total 22
b Use the information in the frequency table to complete the following report
describing the distribution of student responses to the question, ‘How often do you
play sport?’

1B 1C Displaying and describing numerical data 13
Report
When students were asked the question, ‘How often do you play sport’,
the dominant response was ‘Sometimes’, given by % of the students. Of
the remaining students, % of the students responded that they played sport
while % said that they played sport .
8 The table shows the frequency distribution of

Frequency
the eye colour of 11 preschool children.
Eye colour Number Percentage
Use the information in the table to write a brief Brown 6 54.5
report describing the frequency distribution of Hazel 2 18.2
eye colour.
Blue 3 27.3
Total 11 100.0

9 In a survey people were asked to select how much of the time they felt happy from
the alternatives ‘almost none of the time’, ‘some of the time’, ‘most of the time’, or
‘almost all of the time’. Their responses are summarised in the following barchart of
the variable happiness.
30
25
Frequency
20
15
10
5
0 almost none some of most of almost all
of the time the time the time of the time
happiness
The percentage of people who chose the modal response to this question is closest to:
A 30% B 43% C 50% D 57% E 70%
1C Displaying and describing numerical data

Learning intentions
I To be able to construct frequency tables for discrete numerical data.
I To be able to construct frequency tables for grouped numerical data (discrete and
continuous).
I To be able to construct a histogram from a frequency table for numerical data.
I To be able to construct a histogram from numerical data using a CAS calculator.
I To be able to describe the distribution of numerical data according to its key
characteristics of shape, centre, spread and outliers.
Frequency tables can also be used to organise numerical data. For a discrete variable which
only takes a small number of values the process is the same as that for categorical data, as
shown in the following example.
Example 6 Constructing a frequency table for discrete numerical data taking a

small number of values
The number of bedrooms in each of the 24 properties sold in a certain area over a one
month period are as follows:
2 3 4 3 3 4 3 4 4 1 3 2 1 2 2 2 4 5 3 4 4 5 3 4
Construct a table for these data showing both frequency and percentage frequency, giving
the values of the percentage frequency rounded to one decimal place.
1 Find the maximum and the minimum
Number of Frequency
values in the data set. Here the
bedrooms Number %
minimum is 1 and the maximum is 5.
1 2 8.3
2 Construct a table as shown, including
2 5 20.8
all the values between the minimum
3 7 29.2
and the maximum.
4 8 33.3
3 Count the number of 1s, 2s, etc. in the
5 2 8.3
data set. Record these values in the
number column and add the frequencies Total 24 99.9
to find the total.

4 Convert the frequencies to percentages,
and record in the per cent (%) column.
For example, percentage of 1s equals
2
× 100 = 8.3%.
24
5 Total the percentages and record.
The grouped frequency distribution

When the variable can take on a large range of values (e.g., age from 0 to 100 years) or when
the variable is continuous (e.g. response times measured in seconds to 2 decimal places), we
group the data into a small number of convenient intervals.
These grouping intervals should be chosen according to the following principles:
Every data value should be in an interval.
The intervals should not overlap.
There should be no gaps between the intervals.
The choice of intervals can vary but there are some guidelines.
A division which results in about 5 to 15 groups, is preferred.
1C Displaying and describing numerical data 15
Choose an interval width that is easy for the reader to interpret such as 10 units, 100 units
or 1000 units (depending on the data).
By convention, the beginning of the interval is given the appropriate exact value, rather
than the end. As a result, intervals of 0–49, 50–99, 100–149 would be preferred over the
intervals 1–50, 51–100, 101–150 etc.
When we then organise the data into a frequency table using these data intervals we call this
table a grouped frequency table.
Example 7 Constructing a grouped frequency table
The data below give the average hours worked per week in 23 countries.
35.0 48.0 45.0 43.0 38.2 50.0 39.8 40.7 40.0 50.0 35.4 38.8
40.2 45.0 45.0 40.0 43.0 48.8 43.3 53.1 35.6 44.1 34.8
Construct a grouped frequency table with five intervals.
1 Set up a table as shown. Use five
Average Frequency
intervals: 30.0–34.9, 35.0–39.9, . . . ,
hours worked Number Percentage
50.0–54.9.
30.0−34.9 1 4.3
2 List these intervals, in ascending order,
under Average hours worked. 35.0−39.9 6 26.1
3 Count the number of countries whose 40.0−44.9 8 34.8
average working hours fall into each of 45.0−49.9 5 21.7
the intervals. Record these values in the 50.0−54.9 3 13.0
‘Number’ column. Total 23 99.9
4 Convert the counts into percentages
and record in the ‘Percentage’ column.
5 Total the number and percentage columns.
The histogram and its construction

As with categorical data, we would like to construct a visual display of a frequency table for
numerical data. The graphical display of a frequency table for a numerical variable is called
a histogram. A histogram looks similar to a bar chart but, because the data is numerical,
there is a natural order to the plot and the bar widths depend on the data values.
Constructing a histogram from a frequency table

In a frequency histogram:
frequency (count or per cent) is shown on the vertical axis
the values of the variable being displayed are plotted on the horizontal axis

each bar in a histogram corresponds to a data interval

the height of the bar gives the frequency (or the percentage frequency).
Example 8 Constructing a histogram from a frequency table
Construct a histogram for the frequency

Average hours worked Frequency
table opposite.
30.0–34.9 1
35.0–39.9 6
40.0–44.9 8
45.0–49.9 5
50.0–54.9 3
Total 23
1 Label the horizontal axis with the variable 9
name, Average hours worked. Mark the 8
scale using the start of each interval: 30, 7
35, . . . 6
Frequency
2 Label the vertical axis ‘Frequency’. Scale

5
allowing for the maximum frequency, 8.
4
3 Finally, for each interval draw a bar,
3
making the height equal to the frequency.
2
1
0
25 30 35 40 45 50 55 60
Average hours worked
Constructing a histogram from raw data

It is relatively quick to construct a histogram from a frequency table. However, if you have
only raw data (as you mostly do), it is a very slow process because you have to construct the
frequency table first. Fortunately, a CAS calculator will do this for you.
CAS 1: How to construct a histogram using the TI-Nspire CAS

Display the following set of 27 marks in the form of a histogram.
16 11 4 25 15 7 14 13 14 12 15 13 16 14
15 12 18 22 17 18 23 15 13 17 18 22 23

Steps
1 Start a new document by pressing / + N
(or c>New. If prompted to save an existing
document, move the cursor to No and press ·.
2 Select Add Lists & Spreadsheet.
Enter the data into a list named marks.
a Move the cursor to the name cell of column
A and type in marks as the list variable.
Press ·.
b Move the cursor down to row 1, type in the first data value and press ·.
Continue until all the data have been entered. Press · after each entry.
3 Statistical graphing is done through the Data &
Statistics application. Press / + I (or
alternatively press / ) and select Add Data &
Statistics.
a Press e · (or click on the Click to add
variable box on the x-axis) to show the list of
variables. Select marks.
Press · to paste marks to that axis.
b A dot plot is displayed as the default. To change
the plot to a histogram, press b>Plot Type>
Histogram. The histogram shown opposite has
a column (or bin) width of 2, and a starting
point (alignment) of 3. See Step 5 below for
instructions on how to change the appearance
of a histogram.
4 Data analysis
a Move the cursor over any column; a { will
appear and the column data will be displayed
as shown opposite.
b To view other column data values, move the
cursor to another column.
Note: If you click on a column, it will be selected.
Hint: If you accidentally move a column or data point, / + d · will undo the move.
5 Change the histogram column (bin) width to 4 and the starting point to 2.
a Press / + b to access the context menu as shown (below left).
Hint: Pressing / + b · with the cursor on the histogram gives you a context menu that
relates only to histograms. You can access the commands through b>Plot Properties.
b Select Bin Settings>Equal Bin Width.

c In the settings menu (below right) change the Width to 4 and the Starting Point
(Alignment) to 2 as shown. Press ·.
d A new histogram is displayed with column width of 4 and a starting point of

2 but it no longer fits the window (below left). To solve this problem, press
/ + b>Zoom>Zoom-Data and · to obtain the histogram as shown
below right.
6 To change the frequency axis to a percentage axis, press / + b»Scale>Percent

and then press ·.
CAS 1: How to construct a histogram using the ClassPad

Display the following set of 27 marks in the form of a histogram.
16 11 4 25 15 7 14 13 14 12 15 13 16 14
15 12 18 22 17 18 23 15 13 17 18 22 23

Steps
1 From the application menu
screen, locate the built-in
Statistics application.
Tap to open.
Tapping from the icon panel
(just below the touch screen) will
display the application menu if it
is not already visible.
2 Enter the data into a list named
marks.
To name the list:

a Highlight the heading of the first list by tapping it.
b Press on the front of the calculator and tap the tab.
c To enter the data, type the
word marks and press .
Tap and to
return to the list screen.
d Type in each data value and
press or (which
is found on the cursor
button on the front of the
calculator) to move down to
the next cell.
The screen should look like the one shown above right.
3 Set up the calculator to plot a
statistical graph.
a Tap from the toolbar. This
opens the Set StatGraphs dialog
box.
b Complete the dialog box as
given below.
Draw: select On.
Type: select Histogram ( ).
XList: select main\marks ( ).
Freq: leave as 1.
c Tap Set to confirm your selections.

Note: To make sure only this graph is drawn, select SetGraph from the menu bar at the top and
confirm that there is a tick only beside StatGraph1 and no others.
4 To plot the graph:
a Tap in the toolbar.
b Complete the Set Interval
dialog box as follows.
HStart: type 2 (i.e. the
starting point of the
first interval)
HStep: type 4 (i.e. the
interval width).
Tap OK to display
histogram.
Note: The screen is split into two halves, with the graph displayed in the bottom half, as shown
above. Tapping from the icon panel allows the graph to fill the entire screen. Tap again to
return to half-screen size.
5 Tapping from the toolbar
places a marker (+) at the
top of the first column of the
histogram (see opposite) and
tells us that:
a the first interval begins at
2 (x c = 2)
b for this interval, the
frequency is 1 (F c = 1).
To find the frequencies
and starting points of the
other intervals, use the cursor key arrow ( ) to move from interval to interval.
What to look for in a histogram

A histogram provides a graphical display of a 8
Frequency
data distribution. For example, the histogram 6

4
opposite displays the distribution of test marks
2
for a group of 32 students. 0
10 20 30 40 50 60 70 80 90 100
Marks
The purpose of constructing a histogram is to help understand the key features of the data
distribution. These features are:
shape centre spread outliers

Shape
How are the data distributed? Is the histogram peaked? That is, do some data values tend to
occur much more frequently than others, or is the histogram relatively flat, showing that all
values in the distribution occur with approximately the same frequency?
Symmetric distributions
If a histogram is single-peaked, does the histogram region tail off evenly on either side of the
peak? If so, the distribution is said to be symmetric (see Histogram 1).
lower tail peak upper tail peak peak
10 10
8 8
Frequency
Frequency
6 6
4 4
2 2
0 0
Histogram 1 Histogram 2
A single-peaked symmetric distribution is characteristic of the data that derive from
measuring variables such as intelligence test scores, weights of oranges, or any other data for
which the values vary evenly around some central value.
The double-peaked distribution (histogram 2) is symmetric about the dip between the two
peaks. A histogram that has two distinct peaks indicates a bimodal (two modes) distribution.
A bimodal distribution often indicates that the data have come from two different
populations. For example, if we were studying the distance the discus is thrown by Olympic-
level discus throwers, we would expect a bimodal distribution if both male and female
throwers were included in the study.
Skewed distributions
Sometimes a histogram tails off primarily in one direction. If a histogram tails off to the
right, we say that it is positively skewed (Histogram 3). The distribution of salaries of
workers in a large organisation tends to be positively skewed. Most workers earn a similar
salary with some variation above or below this amount, but a few earn more and even fewer,
such as the senior manager, earn even more. The distribution of house prices also tends to be
positively skewed.
peak long upper tail long lower tail peak
10 10
8 +ve skew 8 −ve skew
Frequency
Frequency
6 6
4 4
2 2
0 0
If a histogram tails off to the left, we say that it is negatively skewed (Histogram 4). The
distribution of age at death tends to be negatively skewed. Most people die in old age, a few
in middle age and fewer still in childhood.

Centre
Histograms 6 to 8 display the 8
distribution of test scores for three 7
different classes taking the same 6
Frequency
subject. They are identical in shape, 5
4
but differ in where they are located
3
along the axis. In statistical terms we
2
say that the distributions are ‘centred’
1
at different points along the axis. But 0
what do we mean by the centre of a 50 60 70 80 90 100 110 120 130 140 150
distribution? Histograms 5 to 7
This is an issue we will return to in more detail later in the chapter. For the present we will
take the centre to be the middle of the distribution.
The middle of a symmetric distribution is reasonably easy to locate by eye. Looking at
histograms 5 to 7, it would be reasonable to say that the centre or middle of each distribution
lies roughly halfway between the extremes; half the observations would lie above this point
and half below. Thus we might estimate that histogram 5 (yellow) is centred at about 60,
histogram 6 (light blue) at about 100, and histogram 7 (dark blue) at about 140.
For skewed distributions, it is more difficult to estimate the middle of a distribution by eye.
The middle is not halfway between the extremes because, in a skewed distribution, the
scores tend to bunch up at one end.
However, if we imagine a cardboard 5
cut-out of the histogram, then the middle line that divides
4 the area of the
lies on the line that divides the histogram histogram in half
Frequency
into two equal areas (Histogram 8). 3
Using this method, we would estimate 2

the centre of the distribution to lie
somewhere between 35 and 40, but 1
closer to 35, so we might opt for 37. 0

However, remember that this is only an 15 20 25 30 35 40 45 50
Histogram 8
estimate.
Spread
If the histogram is single-peaked, is it narrow? This would indicate that most of the data
values in the distribution are tightly clustered in a small region. Or is the peak broad? This
would indicate that the data values are more widely spread out. Histograms 9 and 10 are
both single-peaked. Histogram 9 has a broad peak, indicating that the data values are not
very tightly clustered about the centre of the distribution. In contrast, Histogram 10 has a
narrow peak, indicating that the data values are tightly clustered around the centre of the
distribution.

wide central region narrow central region

10 20
Frequency8 16
Frequency
6 12
4 8
2 4
0 0
2 4 6 8 10 12 14 16 18 20 22 2 4 6 8 10 12 14 16 18 20 22
Outliers
Outliers are any data values that stand out from the main body of data. These are data values
that are atypically high or low. See, for example, Histogram 11, which shows an outlier. In
this case it is a data value that is atypically low compared to the rest of the data values.
In the histogram shown there appears to be 10 main body of data
an outlier, a data value which is lower than 8 outlier
Frequency
than rest of the data values. Such values 6
should be checked, they may indicate an 4
unusually low value, but they may also 2
indicate an error in the data. 0
Histogram 11
Example 9 Describing the features of a distribution from a histogram
The histogram shows the 300
gestation period (completed

Frequency
weeks) for a sample for 1000 200

babies born in Australia one
year. Describe this histogram
100
in terms of shape, centre,
spread and outliers.
0
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
Gestation period (weeks)
1 Determine the shape of the distribution. 1 The distribution is clearly negatively
skewed, with a long lower tail.
2 Locate the (approximate) centre of the 2 The centre of the distribution is around
distribution, the value seems to divide 38-39 weeks.
the area of the histogram in half.

24 Chapter 1 Investigating data distributions 1C
3 Consider the spread of the distribution, 3 The data values range from 25-42
are the majority of the values close to weeks, but most of the data values are
the centre, or quite spread out? close to the centre, in the range of 36-41
weeks.
4 Can we identify any outliers? 4 Those values which are less than 34
weeks seem to be small in comparison
to the rest of the data, but which values
are outliers cannot be determined from
the histogram.
We can see from this example that it is very difficult to give an exact values for centre and
spread, and to clearly identify outliers, from the histogram. We will return to this example
later in the chapter once we have discussed which measures of centre and spread are
appropriate for this distribution, and when we have an exact definition on an outlier.
Exercise 1C
Constructing a frequency table for discrete numerical data taking a small number of
values
Example 6 1 The number times a sample of 20 people bought take-away food over a one week
period is as follows:
0 5 3 0 1 0 2 4 3 1 0 2 1 2 1 5 3 0 0 4
a Construct a frequency table for the data, including both the frequency and
percentage frequency.
b What percentage of people bought take away food more than 3 times?
c What is the mode of this distribution?
2 The number of chocolate chips per biscuit in a sample 40 biscuits was found to be as as
follows:
2 5 4 4 5 4 6 4 4 4 5 4 4 5 6 6 5 5 4 6
4 5 5 4 5 4 6 4 6 4 5 4 5 4 6 5 5 6 4 6
a Construct a frequency table for the data, including both the frequency and
percentage frequency.
b What percentage of biscuits contained three or less chocolate chips?
c What is the mode of this distribution?
Constructing a grouped frequency table

Example 7 3 The following are the heights of the 25 players in a women’s football team, in
centimetres.

1C 1C Displaying and describing numerical data 25
188 175 176 161 183 Height (cm) Frequency

169 171 177 165 166
160−164
162 170 174 168 178
165−169
169 181 173 164 179
170−174
163 170 164 175 182
175−179
180−184
185−189
Total 25
a Use the data to complete the grouped frequency table.
b What is the model height for this group of players?
c What percentage of the players are 180 cm or more in height?
Constructing a histogram from a frequency table

Example 8 4 Construct a histogram to display the information
Population density Frequency
in the frequency table opposite. Label axes and
0–199 11
mark scales.
200–399 4
400–599 4
600–799 2
800–999 1
Total 22
Reading information from a histogram
5 The histogram opposite displays the distribution 35
of the number of words in 30 randomly selected 30
sentences. 25
Percentage
a What percentage of these sentences contained: 20

i 5–9 words? 15
ii 25–29 words? 10
iii 10–19 words? 5
iv fewer than 15 words? 0

5 10 15 20 25 30
Number of words
in sentence
Write answers correct to the nearest per cent.
b How many of these sentences contained:
i 20–24 words? ii more than 25 words?
c What is the modal interval?

6 The histogram opposite displays the 4

distribution of the average batting
3
Frequency
averages of cricketers playing for a
district team. 2
a How many players have their 1
averages recorded in this histogram?
0
b How many of these cricketers had a 0 5 10 15 20 25 30 35 40 45 50 55
batting average: Batting average
i 20 or more?
ii less than 15?
iii at least 20 but less than 30?
iv of 45?
c What percentage of these cricketers had a batting average:
i 50 or more? ii at least 20 but less than 40?
Constructing a histogram from raw data using a CAS calculator

7 The pulse rates of 23 students are given below.
86 82 96 71 90 78 68 71 68 88 76 74
70 78 69 77 64 80 83 78 88 70 86
a Use a CAS calculator to construct a histogram so that the first column starts at 63
and the column width is two.
b i What is the starting point of the third column?
ii What is the ‘count’ for the third column? What are the actual data values?
c Redraw the histogram so that the column width is five and the first column starts
at 60.
d For this histogram, what is the count in the interval ‘65 to <70’?
8 The numbers of children in the families of 25 VCE students are listed below.
1 6 2 5 5 3 4 1 2 7 3 4 5
3 1 3 2 1 4 4 3 9 4 3 3
a Use a CAS calculator to construct a histogram so that the column width is one and
the first column starts at 0.5.
b What is the starting point for the fourth column and what is the count?
c Redraw the histogram so that the column width is two and the first column starts
at 0.
d i What is the count in the interval from 6 to less than 8?
ii What actual data value(s) does this interval include?

1C 1C Displaying and describing numerical data 27
Determining shape, centre and spread from a histogram

Example 9 9 Identify each of the following histograms as approximately symmetric, positively
skewed or negatively skewed, and mark the following.
i The mode (if there is a clear mode)
ii Any potential outliers
iii The approximate location of the centre
a 20 b 80
Frequency
Frequency
15 65
10 40
5 20
0 0
Histogram A Histogram B
c 20 d 20
Frequency
Frequency
15 15
10 10
5 5
0 0
Histogram C Histogram D
e 20 f 20
Frequency
Frequency
15 15
10 10
5 5
0 0
Histogram E Histogram F
10 These three histograms show the 10

marks obtained by a group of 9
students in three subjects. 8
7
Frequency
a Are each of the distributions 6

approximately symmetric or 5
skewed? 4
b Are there any clear outliers?
3
2
c Determine the interval 1
containing the central mark for 0
2 6 10 14 18 22 26 30 34 38 42 46
each of the three subjects.
Subject A Subject B Subject C
d In which subject was the Marks
spread of marks the least?


Use the following information to answer questions 11 and 12
The annual salaries for all the sales staff in a large company are summarised in the following
histogram.
40
30
Frequency
20
10
0
40000 50000 60000 70000 80000 90000 100000 110000 120000 130000 140000 150000
salary($)
11 The number of people in the company who earn from $65,000 to less than $70,000 per
year is equal to:
A 20 B 30 C 50 D 32 E 62
12 The shape of this histogram is best described as:

A positively skewed with possible outlliers
B positively skewed with no outliers
C approximately symmetric
D negatively skewed with no outliers
E negatively skewed with outliers
1D Dot plots and stem plots

Learning intentions
I To be able to construct a dot plot for numerical data.
I To be able to construct a stem plot for numerical data.
Dot plots and stem plots are two simple plots used to display numerical data. They are
generally constructed by hand (that is, without using a calculator), from a data set that is
reasonably small.
The dot plot

The simplest way to display numerical data is to construct a dot plot. A dot plot is
particularly suitable for displaying discrete numerical data and provides a very quick way to
order and display a small data set.
1D Dot plots and stem plots 29
A dot plot consists of a number line with each data point marked by a dot. When several
data points have the same value, the points are stacked on top of each other.
Example 10 Constructing a dot plot
The ages (in years) of the 13 members of a cricket team are:

22 19 18 19 23 25 22 29 18 22 23 24 22
Construct a dot plot.
1 Draw a number line, scaled to
17 18 19 20 21 22 23 24 25 26 27 28 29 30
include all data values. Label Age (years)
the line with the variable
being displayed.
2 Plot each data value by
marking a dot above the
corresponding value on the
17 18 19 20 21 22 23 24 25 26 27 28 29 30
number line.
Age (years)
While some CAS calculators will construct a stem plot, they were designed to be a quick
and easy way of ordering and displaying a small data set by hand.
The stem plot

The stem-and-leaf plot, or stem plot for short, is another quick and easy way to display
numerical data. Stem plots work well for both discrete and continuous data. They are
particularly useful for displaying small- to medium-sized sets of data (up to about 50 data
values). Like the dot plot, they are designed to be a pen and paper technique.
In a stem plot, each data value is separated into two parts: its leading digits, which make up
the ‘stem’ of the number, and its last digit, which is called the ‘leaf’.
For example, in the stem-and-leaf plot opposite, the Key: 1|2 = 12
data values 21 and 34 are displayed as follows:
0 8
Stem Leaf
1 2499
21 is displayed as 2 1
2 111122236699
34 is displayed as 3 4
3 24
A key should always be included to show how the
4 4
numbers in the plot should be interpreted.
5 9

Example 11 Constructing a stem plot
University participation rates (%) in 23 countries are given below.

26 3 12 20 36 1 25 26 13 9 26 27
15 21 7 8 22 3 37 17 55 30 1
Display the data in the form of a stem plot.
1 The data set has values in the units, tens,
0
twenties, thirties, forties and fifties. Write the
1
stems 0, 1, 2, 3, 4, and 5 in ascending order,
2 6
followed by a vertical line. Now attach the
leaves. The first data value is ‘26’. The stem 3
is ‘2’ and the leaf is ‘6’. Opposite the 2 in the 4
stem, write down the number 6, as shown. 5
2 Continue systematically working through the
0 3 1 9 7 8 3 1
data, following the same procedure until all
1 2 3 5 7
points have been plotted. You will then have
2 6 0 5 6 6 7 1 2
the stem plot, as shown.
3 6 7 0
4
5 5
3 To complete the task, write the leaves on each

rate (%) Key: 1|2 = 12
stem in ascending order, then add the variable
0 1 1 3 3 7 8 9
name and a key.
1 2 3 5 7
2 0 1 2 5 6 6 6 7
3 0 6 7
4
5 5
Stem plots with split stems

In some instances, using the simple process outlined above produces a stem plot that is too
cramped to give an good overall picture of the variation in the data. This happens when the
data values all have the same one or two first digits.

1D 1D Dot plots and stem plots 31
For example, consider the marks obtained by 17 VCE students on a statistics test.
2 12 13 9 18 17 7 16 12 10 16 14 11 15 16 15 17
If we use the process described in Example 11 to form a stem plot, we end up with a
‘bunched-up’ plot like the below.
0 2 7 9
1 0 1 2 2 3 4 5 5 6 6 6 7 7 8
We can solve this problem by ‘splitting’ the stems.
Generally the stem is split into halves or fifths as shown below.
Key: 1|6 = 16 Key: 1|6 = 16 Key: 1|6 = 16
0 2 7 9 0 2 0
1 0 1 2 2 3 4 5 5 6 6 6 7 7 8 0 7 9 0 2
Single stem 1 0 1 2 2 3 4 0
1 5 5 6 6 6 7 7 8 0 7
Stem split into halves 0 9
1 0 1
1 2 2 3
1 4 5 5
1 6 6 6 7 7
1 8
Stem split into fifths
Splitting the stems is useful when there are only a few different values for the stem.
Exercise 1D
Constructing a dot plot
Example 10 1 The following data gives the number of rooms in 11 houses.
4 6 7 7 8 4 4 8 8 7 8
a Is the variable number of rooms discrete or continuous?
b Construct a dot plot.
2 The following data give the number of children in the families of 14 VCE students:
1 6 2 5 5 3 4 4 2 7 3 4 3 4
a Is the variable number of children discrete or continuous?
c Write down the value of the mode. What does the mode represent in the context of
the data?
3 The following data give the average life expectancies in years of 13 countries.
76 75 74 74 73 73 75 71 72 75 75 78 72
32 Chapter 1 Investigating data distributions 1D
a Is the variable life expectancy discrete or continuous?

c Write down the value of the mode. What does the mode represent in the context of
the data?
4 Describe the shape of each of the following distributions (negatively skewed, positively
skewed, or approximately symmetric).
a b
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
5 The ages of each member of a running club are as follows:

22 20 20 23 21 22 21 25 21 24 18
20 19 22 23 25 19 21 20 21 21 22
a Construct a dot plot of the ages of the players.

b What is the mode of this distribution?
c What is the shape of the distribution of runners ages?
d What percentage of runners are younger than 20? Give your answer to the nearest
whole percentage.
Constructing a stem plot

Example 11 6 The data below give the urbanisation rates (%) in 23 countries.
54 99 22 20 31 3 22 9 25 3 56 12
16 9 29 6 28 100 17 99 35 27 12
a Is the variable urbanisation rate discrete or continuous?
b Construct a stem plot with an appropriate key.
Constructing a stem plot with split stems

7 The data below give the wrist circumference (in cm) of 15 men.
16.9 17.3 19.3 18.5 18.2 18.4 19.9 16.7
17.7 16.5 17.0 17.2 17.6 17.1 17.6
a Is the variable wrist circumference discrete or continuous?
b Construct a stem plot for wrist circumference using:
i stems: 16, 17, 18, 19
ii these stems split into halves, that is: 16, 16, 17, 17, . . .
Interpreting a stem plot

8 Describe the shape of each of following distributions (negatively skewed, positively
skewed, or approximately symmetric).

1D 1D Dot plots and stem plots 33
a key: 4|1 represents 4.1 b key: 3|1 represents 31

0 2 3 3 2
0 5 6 6 6 7 7 8 8 9 9 4 2 7
1 0 0 1 2 2 2 3 3 3 3 3 5 0 1 5 9
1 5 5 6 7 8 6 1 3 3 5 7 7 7 9
2 0 2 3 7 0 2 3 3 4 4 6 7 8 9 9
2 2 5 8 2 5
3 0 0
3 7
9 The stem plot on the right shows the ages, Age (years) key: 2|0 represents 20
in years, of all the people attending a
2 2 3
meeting. 2 5 6 6 7 7 8 9
a How many people attended the meeting? 3 0 1 3 3 4 4 4 4
b What is the shape of the distribution of 3 5 5 5 6 7 7 7 8 8 8 9 9
ages? 4 0 2 3 3 4 4
4 5 5 6 8
c How many of these people were less
5 0
than 33 years old?

The following stem plot shows the distribution of the time it took (in minutes) for each of a
group of 25 people to solve a complex task.
Time (minutes) key: 4|0 represents 4.0

4 2 6
5 1 3 6 8
6 0 1 5 6 7
7 1 3 4 5 7 8 9
8 0 2 5 9
9 5 5
10 6
10 The shape of this distribution is best described as:

A positively skewed with a possible outlier
B positively skewed with no outliers
C approximately symmetric
D negatively skewed with no outliers
E negatively skewed with outliers

34 Chapter 1 Investigating data distributions 1D
11 The time taken by the slowest 20% of people was:

A more than 8.5 minutes B 8.5 minutes or more C less than 5.6 minutes
D 5.6 minutes or less E more than 5.6 minutes
1E Using a logarithmic (base 10) scale to display

data
Learning intentions
I To be able to revise the concept of log10 x.
I To be able to investigate the effect of the logarithmic scale on the features of a
distribution.
Many numerical variables that we deal with in statistics have values that range over
several orders of magnitude. For example, the populations of countries range from a few
thousand to hundreds of thousands, to millions, to hundreds of millions to just over 1 billion.
Constructing a histogram that effectively locates every country on the plot is impossible.
One way to solve this problem is to use a scale that spreads out the countries with small
populations and ‘pulls in’ the countries with huge populations.
A scale that will do this is called a logarithmic scale (or, more commonly, a log scale).
Consider the numbers:
0.01, 0.1, 1, 10, 100, 1000, 10 000, 100 000, 1 000 000
Such numbers can be written more compactly as:
10−2 , 10−1 , 100 , 101 , 102 , 103 , 104 , 105 , 106
In fact, if we make it clear we are only talking about powers of 10, we can merely
write down the powers:
−2, −1, 0, 1, 2, 3, 4, 5, 6
These powers are called the logarithms of the numbers or ‘logs’ for short.
When we use logarithms to write numbers as powers of 10, we say we are working with
logarithms to the base 10. We can indicate this by writing log10 .
log10 x
If log10 x = b, then 10b = x
Thus we can say for example that:

log10 (100) = 2, since 102 = 100
log10 (1000) = 3, since 103 = 1000
log10 (1000000) = 6, since 106 = 1000000

1E Using a logarithmic (base 10) scale to display data 35
Properties of logarithms to the base 10

1 If a number is greater than one, its log to the base 10 is greater than zero.
2 If a number is greater than zero but less than one, its log to the base 10 is negative.
3 If the number is zero, then its log is undefined.
The effect of the the logarithmic scale

The set of numbers
0.01, 0.1, 1, 10, 100, 1000, 10 000, 100 000, 1 000 000
ranges from 0.01 to 1 million.
Thus, if we wanted to plot these numbers on a scale,
the first seven numbers would cluster together at one
end of the scale, while the eighth (1 million) would
be located at the far end of the scale.
0 300 000 600 000 900 000
Number
By contrast, if we plot the logs of these numbers,

they are evenly spread along the scale. We use this –2 –1 0 1 2 3 4 5 6
idea to display a set of data whose values range over log(number)
several orders of magnitude. Rather than plot the data

values themselves, we plot the logarithms of their
data values.
Logarithmic transformation
A logarithmic transformation involves changing the scale on the horizontal axis from x
to log10 (x), and replacing each of the data values with its logarithm.
For example, the histogram below displays the body weights (in kg) of a number of animal
species. Because the animals represented in this data set have weights ranging from around
1 kg to 90 tonnes (a dinosaur), most of the data are bunched up at one end of the scale and
much detail is missing. The distribution of weights is highly positively skewed, with an
outlier.

80%
60%
Percentage 40%
20%
0%
0 10 000 20 000 30 000 40 000 50 000 60 000 70 000 80 000 90 000
Body weight
However, when a logarithmic transfor- 28%

mation is used, their weights are much 24%
more evenly spread along the scale. The 20%
Percentage
distribution is now approximately symmet- 16%
ric, with no outliers, and the histogram is 12%
considerably more informative. 8%
4%
We can now see that the percentage of
0%
animals with weights between 10 and –2 –1 0 1 2 3 4 5 6
100 kg is similar to the percentage of log(body weight)
animals with weights between 100 and 1000 kg.
Working with logarithms

To construct and interpret a log data plot, like the one above, you need to be able to:
1 Work out the log for any number. So far we have only done this for numbers such as 10,
100, 1000 that are exact powers of 10; for example, 100 = 102 , so log 100 = 2.
2 Work backwards from a log to the number it represents. This is easy to do in your head
for logs that are exact powers of 10 – for example, if the log of a number is 3 then the
number is 103 = 1000. But it is not a sensible approach for numbers that are not exact
powers of 10.
Your CAS calculator is the key to completing both of these tasks in practice.
Example 12 Using a CAS calculator to find logs
a Find the log of 45, correct to two significant figures.

b Find the number with log equal to 2.7125, correct to the nearest whole number.

a Open a calculator screen, type log (45) and press
·. Write down the answer correct to two
significant figures.
b If the log of a number is 2.7125, then the number
is 102.7125 .
Enter the expression 102.7125 and press ·.
Write down the answer correct to the nearest
whole number. a log 45 = 1.65 . . .
= 1.7 (to 2 sig. figs)
b 10 2.7125
= 515.82 . . .
= 516 (to the nearest
whole number)
Analysing data displays with a logarithmic scale

Now that you know how to work out the log of any number and convert logs back to
numbers, you can analyse a data plot using a log scale.
Example 13 Interpreting a histogram with a log scale
The histogram shows the 28%

distribution of the weights of 27 24%
animal species plotted on a log
20%
Percentage
scale.
16%
a What body weight (in kg) is
12%
represented by the number 4 on
8%
the log scale?
4%
b How many of these animals have
0%
body weights more than 10 000 –2 –1 0 1 2 3 4 5 6
kg? log(body weight)
c The weight of a cat is 3.3 kg. Use your calculator to determine the log of its weight
correct to two significant figures.
d Determine the weight (in kg) of the animal with a log(body weight) of 3.4 (the
elephant). Write your answer correct to the nearest whole number.
a If the log of a number is 4 then the a 104 = 10 000 kg
number is 10 = 10 000.
4

b On the log scale, 10 000 is shown as 4. b Two animals
Thus, the number of animals with

a weight greater than 10 000 kg
corresponds to the number of animals
with a log weight of greater than 4.
This can be determined from the
histogram which shows there are two
animals with log weights greater than 4.
c The weight of a cat is 3.3 kg. Use c Cat: log 3.3 = 0.518...
your calculator to find log 3.3. Write = 0.52 kg (to 2 sig. figs)
the answer correct to two significant
figures.
d The log weight of an elephant is 3.4. d Elephant: 103.4 = 2511.88...
Determine its weight in kg by using = 2512 kg
your calculator to evaluate 103.4 . Write
the answer correct to the nearest whole
number.
Constructing a histogram with a log scale

The task of constructing a histogram is also a CAS calculator task.
CAS 2: Using a TI-Nspire CAS to construct a histogram with a log scale

The weights of 27 animal species (in kg) are recorded below.
1.4 470 36 28 1.0 12 000 2600 190 520
10 3.3 530 210 62 6700 9400 6.8 35
0.12 0.023 2.5 56 100 52 87 000 0.12 190
Construct a histogram to display the distribution:
a of the body weights of these 27 animals and describe its shape
b of the log of the body weights of these animals and describe its shape.

Steps
1 a Start a new document by pressing / + N.
b Select Add Lists & Spreadsheet.
Enter the data into a column named weight.
2 a Press / + I and select Add Data &

Statistics.
Click on the Click to add variable on the
x-axis and select the variable weight. A dot
plot is displayed.
b Plot a histogram using b>Plot
Type>Histogram.
c Describe the shape of the distribution. Shape: positively skewed with
outliers
3 a Return to the Lists & Spreadsheet screen.
b Name another list logweight.
c Move the cursor to the formula cell below the
logweight heading. Type in = log(weight).
Press · to calculate the values of logweight.
4 a Plot a histogram using a log scale. That is,

plot the variable logweight.
Note: Use b>Plot Properties>Histogram
Properties>Bin Settings>Equal Bin
Width and set the column width (bin) to 1
and alignment (start point) to −2 and use
b>Window/Zoom>Zoom-Data to rescale.
b Describe the shape of the distribution. Shape: approximately symmetric

CAS 2: Using a ClassPad to construct a histogram with a log scale

The weights of 27 animal species (in kg) are recorded below.
1.4 470 36 28 1.0 12 000 2600 190 520
10 3.3 530 210 62 6700 9400 6.8 35
0.12 0.023 2.5 56 100 52 87 000 0.12 190
Construct a histogram to display the distribution:
a of the body weights of these 27 animals and describe its shape
b of the log of the body weights of these animals and describe its shape.
Steps
1 In the statistics application
enter the data into
a column named weight as
shown.
2 Plot a histogram of the data.
a Tap from the

toolbar.
b Complete the dialog
box.
Draw: select On.
Type: select Histogram ( )
XList: select main\weight( ).
Freq: leave as 1.
Tap Set to confirm your selections.
c Tap in the toolbar.
d Complete the Set Interval dialog box as follows:
HStart: 0
HStep: 5000
Describe the shape of the distribution. Shape: positively skewed with outliers

1E 1E Using a logarithmic (base 10) scale to display data 41
3 a Return to the data entry screen.

b Name another column ‘lwt’, short for log(weight).
c Tap in the calculation cell at the bottom of this
column.
Type log(weight) and tap .
4 Plot a histogram to display the distribution of weights

on a log scale. That is, plot the variable lwt.
a Tap from the toolbar.
b Complete the dialog box.
Draw: select On.
Type: select Histogram ( ).
XList: select main\lwt ( ).
Freq: leave as 1.
c Tap in the toolbar.
d Complete the Set Interval dialog box as follows:
HStart: type -2
HStep: type 1
Tap OK to display histogram.
Describe the shape of the distribution. Shape: approximately symmetric
Skill-
sheet Exercise 1E
Determining the log of a number

Example 12 1 Using a CAS calculator, find the logs of the following numbers correct to one decimal
place.
a 2.5 b 25 c 250 d 2500
e 0.5 f 0.05 g 0.005 h 0.0005

42 Chapter 1 Investigating data distributions 1E
Determining a number from its log

2 Find the numbers with the following logs:
a −2.5 b −1.5 c −0.5 d 0
Write your decimal answers correct to two significant figures.
Constructing a histogram with a log scale

3 The brain weights of the same 27 animal species (in g) are recorded below.
465 423 120 115 5.5 50 4600 419 655
115 26 680 406 1320 5712 70 179 56
1.0 0.4 12 175 157 440 155 3.0 180
a Construct a histogram to display the distribution of brain weights and comment on
its shape.
b Construct a histogram to display the log of the brain weights and note the shape of
the distribution.
Interpreting a histogram with a log scale

Example 13 4 The histogram opposite shows the 9
distribution of brain weights (in g)
of 27 animal species plotted on a log
6
Frequency
scale.
a The brain weight (in g) of a mouse
is 0.4 g. What value would be 3
plotted on the log scale?
b The brain weight (in g) of an
African elephant is 5712 g. What 0
−2 −1 0 1 2 3 4 5 6
is the log of this brain weight (to
log(weight)
two significant figures)?
c What brain weight (in g) is represented by the number 2 on the log scale?
d What brain weight (in g) is represented by the number –1 on the log scale?
e Use the histogram to determine the number of these animals with brain weights:
i 1000 g or more ii from 1 to less than 100 g iii 1 g or more

1E 1F Measures of centre and spread 43

5 The histogram shows the carbon dioxide 70
60
emissions (in thousands of metric tons) for
50
Frequency
239 different countries, plotted on a log10 40
scale. 30
20
10
0
1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
log(CO2)
Based on this histogram, the percentage of countries with carbon dioxide emissions (in
thousands of metric tons) from 10 000 to less than 100 000 is equal to:
A 21 B 25 C 26 D 50 E 60
42
6 The following histogram 40
38
shows the amount spent 36
34
32
by tourists from several 30
28
countries in one year 26
Frequency
24
22
(spending), plotted on a log10 20
18
16
scale. 14
12
10
8
6
4
2
0
6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00
log(spending)
The number of countries where tourists spent from $100 000 000 to less than
$1 000 000 000 per year is equal to:
A 12 B 25 C 33 D 37 E 51
1F Measures of centre and spread

Learning intentions
I To be able to understand the mean and the median as measures of centre.
I To be able to understand the range, interquartile range and standard deviation as
measures of spread.
I To be able to know whether to use the median and interquartile range, or the mean and
standard deviation, for a particular distribution.
I To be able to use a CAS calculator to calculate these summary statistics.

The median, range and interquartile range

The most versatile statistical tools for numerically describing the centre and spread of a
distribution are:
the median (the middle value) as a measure of centre;
the range (the maximum spread of the data values), and the interquartile range (the
spread of the middle half of data values) as measures of spread.
While these statistical values could be estimated only approximately from a histogram, they
can be determined exactly when we use either a dot or stem plot.
Determining the median

We begin by revisiting the rule for locating the median of a data set.
The median
The median is the middle value in an ordered data set.
n + 1
For n data values the median is located at the th position.
2
When:
n is odd, the median will be the middle data value
n is even, the median will be the average of the two middle data values.
Example 14 Finding the median value in a data set
Order each of the following data sets, locate the median, and then write down its value.
a 2 9 1 8 3 5 3 8 1 b 10 1 3 4 8 6 10 1 2 9
a For an odd number of data values, the
median will be the middle data value.
1 Write down the data set in order. 1 1 2 3 3 5 8 8 9
1 1 2 3 3 5 8 8 9
9 + 1
2 Locate the middle data value by eye Median is the th or fifth value.
2
or use the rule.
3 Write down the median. Median = 3
b For an even number of data values, the
median will be the average of the two
middle data values.

1F Measures of centre and spread 45
1 Write down the data set in order. 1 1 2 3 4 6 8 9 10 10

1 1 2 3 4 6 8 9 10 10
2 Locate the two middle data values Median is the average of the 5th and 6th
and find their average or use the values.
rule.
4 + 6
Write down the median. Median = =5
2
Note: You can check that you are correct by counting the number of data values each side of the median.
They should be equal.
Using a dot plot to help locate medians

The process of calculating a median, as outlined above, is very simple in theory but can be
time-consuming in practice. This is because you have to spend most of your time ordering
the data set. For a very large data set this is a calculator task.
However, even for a reasonably large data set, locating a median in a dot or stem plot
requires no more than counting because the data are already ordered for you.
Example 15 Finding the median value from a dot plot

The dot plot opposite displays
the age distribution (in years) of
the 13 members of a local cricket
17 18 19 20 21 22 23 24 25 26 27 28 29 30
team. Age (years)
Determine the median age of these cricketers and mark its location on the dot plot.
The median value is the middle
data value in the dot plot.
1 Locate the middle data value
(or use the rule) and identify it
on the dot plot.
17 18 19 20 21 22 23 24 25 26 27 28 29 30
Age (years)
2 Write down its value. Median = 22 years
Example 16 Finding the median value from a stem plot
The stem plot opposite displays the maximum temperature (in ◦ C) Key: 0|8 = 8◦ C
for 12 days in January. 1 899
Determine the median maximum temperature for these 12 days. 2 0257899
3 13

1 For an even number of data values, as in this Key: 0|8 = 8◦ C
example, the median will be the average of the 1 899
two middle data values. 2 0257899
2 Locate the two middle data values in the dot plot
3 13
by eye (or use the rule) and identify them on the
plot.
25 + 27
3 Determine the median by finding the average of M= = 26◦ C
2
these two data values.
Having found the median value in a dot plot or stem plot, we now look at ways of doing the
same with the first measure of spread, the range.
The range
The range
The range, R, is the simplest measure of spread of a distribution. It is the difference
between the largest and smallest values in the data set.
R = largest data value − smallest data value
Example 17 Finding the range from a stem plot
The stem plot opposite displays the maximum Key: 0|8 = 8◦ C

temperature (in ◦ C) for 12 days in January. 1 899
Determine the temperature range over these 12 days. 2 0257899
3 13
1 Identify the lowest and highest values in the stem Key: 0|8 = 8◦ C
plot and write them down. 1 899
2 0257899
3 13
2 Substitute into the rule for the range and Lowest = 18, highest = 33, range
evaluate. = 33 − 18 = 15◦ C
Because the range depends only on the two extreme values in the data, it is not always an
informative measure of spread. For example, one or other of these two values might be an
outlier. Furthermore, any data with the same highest and lowest values will have the same
range, irrespective of the way in which the data are spread out in between.

A more refined measure of spread that overcomes these limitations of the range is the
interquartile range (IQR).
The interquartile range (IQR)

Quartiles
To determine the value of the IQR, we first need to determine the quartiles.
Just as the median is the point that divides a distribution in half, quartiles are the points
that divide a distribution into quarters. We use the symbols Q1 , Q2 and Q3 to represent the
quartiles. Note that the second quartile, Q2 , is the median.
Determining the interquartile range

To find the interquartile range of a distribution:
arrange all observations in order according to size
divide the observations into two equal-sized groups, and if n is odd, omit the median
from both groups
locate Q1 , the first quartile, which is the median of the lower half of the observations
locate Q3 , the third quartile, which is the median of the upper half of the observa-
tions.
The interquartile range IQR is then: IQR = Q3 − Q1
We can interpret the interquartile range as follows:

Since Q1 , the first quartile, is the median of the lower half of the observations, then it
follows that 25% of the data values are less than Q1 , and 75% are greater than Q1 .
Since Q3 , the third quartile, is the median of the upper half of the observations, then it
follows that 75% of the data values are less than Q3 , and 25% are greater than Q3 .
Thus, the interquartile range (IQR) gives the spread of the middle 50% of data values.
Example 18 Finding the IQR from an ordered stem plot when n is even
Weight (kg) 1 | 2 represents 1.2 kg
Find the interquartile range of the weights of
1 9
the 18 cats whose weights are displayed in
2 1 3 5 8
the ordered stem plot below.
3 0 0 4 9 9
4 0 4 5 8
5 0 3
6 3 4
1 There are 18 values in total. This means that
there are nine values in the lower ‘half’, and nine
in the upper ‘half’.

2 The median of the lower half (Q1 ) is the middle Lower half:
of lower nine values, which is the 5th value from 1.9 2.1 2.3 2.5 2.8 3.0 3.0 3.4 3.9
the bottom. Q1 = 2.8
3 The median of the upper half (Q3 ) is the middle Upper half:
of the upper nine values, which is the 5th value 3.9 4.0 4.4 4.5 4.8 5.0 5.3 6.3 6.4
from the top. Q3 = 4.8
4 Determine the IQR using IQR = Q3 − Q1 . IQR = Q3 − Q1 = 4.8 − 2.8 = 2.0
Example 19 Finding the IQR from an ordered stem plot when n is odd
The stem plot shows the life expectancy (in Stem: 5|2 = 52 years
years) for 23 countries. Find the IQR for life 5 2
expectancies. 5 56
6 4
6 6679
7 122334444
7 556677
1 Since there are 23 values, the median is the 12th Stem: 5|2 = 52 years
value from either end which is 73. Mark the 5 2
value 73 on the stem plot.
5 5 6
2 Since n is odd, to find the quartiles the median
6 4 Q1
value is excluded. This leaves 11 values below
the median and 11 values above the median. 6 6 6 7 9 median
Then:
Q1 = midpoint of the bottom 11 data values 7 1 2 2 3 3 4 4 4 4
Q3
Q3 = midpoint of the top 11 data values.
7 5 5 6 6 7 7
Write these values down.
Q1 = 66, Q3 = 75
3 Determine the IQR using IQR = Q3 − Q1 . IQR = Q3 − Q1 = 75 − 66 = 9
To check that these quartiles are correct, write the data values in order, and mark the median
and the quartiles. If correct, the median divides the data set up into four equal groups.

Q1 Q2 (= M) Q3
52
| 55 {z
56 64 66} 66 67
| 69 {z } 73 73
71 72 72 | 74 {z } 75 75
74 74 74 | 76 {z 76 77 77
}
5 values 5 values 5 values 5 values
Example 20 Finding the median and quartiles from a histogram

The histogram shows the average number of hours per week a group of 23 people spent
on the internet. Find possible values for the median and quartiles of this distribution.
6
5
4
Frequency
0
10 20 30 40 50 60 70 80 90 100
internet hours per week
1 Since there are 23 values, we can locate There is 1 value in the interval 10-20 (total
which interval contains the median by 1), 2 values in the interval 40-50 (total 3),
adding the number of values in each 4 values in the interval 50-60 (total 7), 5
interval moving from left to right. values in the interval 60-70 (total 12). Thus
M = the 12th value from the bottom the median is in the interval 60-70.
(adding from left to right)
2 Similarly
Q1 = the 6th value from the bottom Q1 is in the interval 50-60
(adding from left to right)
Q3 = the 6th value from the top Q3 is in the interval 80-90
(adding from right to left)
Why is the IQR a more useful measure of spread than the range?
The IQR is a measure of spread of a distribution that includes the middle 50% of
observations. Since the upper 25% and lower 25% of observations are discarded, the
interquartile range is generally not affected by the presence of outliers.

The mean and standard deviation

So far, we have looked at methods for describing the centre and spread for distributions of
any shape. We used the median, IQR and range for this purpose. In this section, we will
look at alternative measures of centre (the mean) and spread (the standard deviation) that
are only useful when working with symmetric distributions without outliers. While this may
seem unnecessarily restrictive, these two measures have the advantage of being able to fully
describe the centre and spread of a symmetric distribution with only two numbers.
The mean
The mean of a set of data is what most people call the ‘average’. The mean of a set of data
is given by:
M
sum of data values
mean =
total number of data values 3
For example, consider the set of data:
2 3 4
2 3 3 4
0 1 2 3 4 5 6
The mean of this set of data is given by:
2 + 3 + 3 + 4 12 mean
mean = = =3
4 4
From a pictorial point of view, the mean is the balance point of a distribution (see above).
Note that in this case, the mean and the median coincide; the balance point of the
distribution is also the point that splits the distribution in half. That is, there are two
data points to the left of the mean and two to the right. This is a general characteristic of
symmetric distributions.
However, consider the data set M
2 3 3 8 3
The median remains at M = 3, but:
2 3 8
2 + 3 + 3 + 8 16
mean = = =4
4 4 0 1 2 3 4 5 6 7 8
mean
Note that the mean is affected by changing the largest data value but that the median is not.
Some notation
Because the rule for the mean is relatively simple, it is easy to write in words. However, later
you will meet other rules for calculating statistical quantities that are extremely complicated
and hard to write out in words.
To overcome this problem, we introduce a shorthand notation that enables complex
statistical formulas to be written out in a compact form. In this notation, we use:

P
the Greek capital letter sigma, , as a shorthand way of writing ‘sum of’
a lower case x to represent a data value
a lower case x with a bar, x̄ (pronounced ‘x bar’), to represent the mean of the data values
an n to represent the total number of data values.
P
x
The rule for calculating the mean then becomes: x̄ =
n
Example 21 Calculating the mean from the formula
The following is a set of reaction times (in milliseconds): 38 36 35 43 46 64 48 25

Write down the values of the following, correct to one decimal place.
P
a n b x c x̄
a n is the number of data values. n=8
x = 38 + 36 + 35 + 43 + 46 + 64 + 48 + 25
P
b x is the sum of the data
P
values. = 335
c x̄ is the
P mean. It is defined by
P
x 335
x x̄ = = = 41.9
x̄ = . n 8
n
The relationship between the mean and the median

Whereas the median lies at the midpoint of a distribution, the mean is the balance point of
the distribution. For approximately symmetric distributions, both the median and mean will
be approximately equal in value.
An example of a symmetric distribution is the 15
distribution of mortality rates for 60 US cities
shown opposite. Calculations reveal that the mean 10
mortality rate for the cities is 940 per 100 000
5
while the median mortality rate is 944 per 100 000
people. As expected, the mean and median are
approximately equal in value. 775 950 1125
Mortality
An example of a highly skewed distribution is the 20

population distribution of different cities, shown
opposite. This distribution is clearly positively 10
skewed with two outliers. The mean population
5
is 1.4 million, while the median population is
0.9 million. They are quite different in value. The
0 4500000 9000 000
mean has been affected by the extreme values in the Population
tail and no longer represents the typical city.

When to use the median rather than the mean

Because the value of the median is relatively unaffected by the presence of extreme values
in a distribution, it is said to be a resistant statistic. For this reason, the median is frequently
used as a measure of centre when the distribution is known to be clearly skewed and/or
likely to contain outliers.
For example, median house prices are used to compare housing prices between capital cities
in Australia because the distribution of house prices tends to be positively skewed. There are
always a small number of very expensive houses sold for much higher prices than the rest of
houses sold.
However, if a distribution is symmetric, there will be little difference in the value of the
mean and median and we can use either. In such circumstances, the mean is often preferred
because:
it is more familiar to most people
more can be done with it theoretically, particularly in the area of statistical inference
(which you will learn about if you are doing Mathematics Methods).
Choosing between the mean and the median

The mean and the median are both measures of the centre of a distribution. If the
distribution is:
symmetric and there are no outliers, either the mean or the median can be used to
indicate the centre of the distribution
clearly skewed and/or there are outliers, it is more appropriate to use the median to
indicate the centre of the distribution.
The standard deviation

To measure the spread of a data distribution around the median (M) we use the interquartile
range (IQR). To measure the spread of a data distribution about the mean ( x̄) we use the
standard deviation (s).
The standard deviation rP

(x − x̄)2
The formula for the standard deviation, s, is: s =
n−1
Although not easy to see from the formula, the standard deviation is an average of the
squared deviations of each data value from the mean. We work with the squared deviations
because the sum of the deviations around the mean (the balance point) will always be zero.
Calculating the standard deviation

Normally, you will use your calculator to determine the value of a standard deviation.
Instructions for the TI-Nspire or ClassPad follow.

CAS 3: How to calculate the mean and standard deviation using the
TI-Nspire CAS
The following are the heights (in cm) of a group of women.
176 160 163 157 168 172 173 169
Determine the mean and standard deviation of the women’s heights. Give your answers
correct to two decimal places.
Steps
1 Start a new document by pressing / + N.
Enter the data into a list named height, as shown.
3 Statistical calculations can be done in either
the Lists & Spreadsheet application or the
Calculator application (used here).
Press / + I and select Add Calculator.
a Press b>Statistics>Stat Calculations>One-
Variable Statistics. Press · to accept the
Num of Lists as 1.
b i To complete this screen, use the ¢ arrow
and · to paste in the list name height.
ii Pressing · exits this screen and

generates the results screen shown
opposite.
4 Write down the answers to the required degree The mean height of the women
of accuracy (i.e. two decimal places). is x̄ = 167.25 cm and the stan-
dard deviation is s = 6.67 cm.
Notes: a The sample standard deviation is sx.
b Use the £ ¤ arrows to scroll through the results screen to obtain values for additional
statistical values.

54 Chapter 1 Investigating data distributions 1F
CAS 3: How to calculate the mean and standard deviation using the
ClassPad
The following are all heights (in cm) of a group of women.
176 160 163 157 168 172 173 169
Determine the mean and standard deviation of the women’s heights correct to two
decimal places.
Steps
1 Open the Statistics application
and enter the data into the
column labelled height.
2 To calculate the mean and standard
deviation, select Calc from the
menu One-Variable from the
drop-down menu to open the Set
Calculation dialog box shown
below.
3 Complete the dialog box as shown.
XList: select main\height ( ).
Freq: leave as 1.
4 Tap OK to confirm your selections
and calculate the required statistics,
as shown.
5 Write down the answers to two decimal The mean height of the women is
places. x̄ = 167.25 cm.
The standard deviation is
s x = 6.67 cm.
Notes: a The value of the standard deviation is given by s x .
b Use the side-bar arrows to scroll through the results screen to obtain values for additional
statistical values (i.e. median, Q3 and the maximum value) if required.
Exercise 1F
Determining the median from data

Example 14 1 Locate the medians of the following data sets. For each set of data, check that the
median divides the ordered data set into two equal groups.
a 4 9 3 1 8 6 b 10 9 12 20 14

1F 1F Measures of centre and spread 55
2 The prices of nine second-hand mountain bikes advertised for sale were as follows.
$650 $3500 $750 $500 $1790 $1200 $2950 $430 $850
What is the median price of these bikes?
Determining the median from a dot plot

Example 15 3 The dot plot opposite displays the number of
times 20 shoppers visited their supermarket in a
week. Find the median number of visits.
0 1 2 3 4 5 6
Number of times visited
Determining the median and range from a stem plot

4 The following stem plot shows the distribution of the time it took (in minutes) for each
of a group of 25 people to solve a complex task.
Time (minutes) key: 4|0 represents 4.0

4 2 6
5 1 3 6 8
6 0 1 5 6 7
7 1 3 4 5 7 8 9
8 0 2 5 9
9 5 5
10 6
Example 16 a Find the median time taken.

Example 17 b Find the range of the time taken.
Determining the median and quartiles from a dot plot

Example 18 5 The dot plot shows the distribution of the number of
children in each of 14 families.
a Determine the median, M.
b Determine the quartiles Q1 and Q3 . 0 1 2 3 4 5 6 7
c Calculate the IQR. Children in family
d Calculate the range, R.
e By writing the data values in a line, check that the quartiles and the median have
divided the data set up into four equal groups.

Determining the median and quartiles from a stem plot

6 The stem plot displays the infant mortality rates Key: 0|7 = 7
(deaths per 1000 live births) in 14 countries. 0 7 7 9
a Determine the median, M. 1 0 0 0 0 2 2 4
b Determine the quartiles Q1 and Q3 . 1 5
c Calculate the IQR and the range, R. 2 0 1
2 5
7 The stem plot displays the test scores for 20 students. Key: 1|0 = 10
a Describe the shape of the distribution. 1 0 2
b Determine the median, M. 1 5 5 6 9
c Determine the quartiles, Q1 and Q3 . 2 3 3 4
d Calculate the IQR and the range, R. 2 5 7 9 9 9
3 0 1 2 4
3 5 9
Example 19 8 The dot plot displays the number of goals scored in 23

games.
a Describe the shape of the distribution and note
outliers (if any).
b Without using your calculator determine:
i the median, M.
ii the IQR.
0 1 2 3 4 5 6 7
Goals scored
9 The stem plot displays the university participation rates Key: 0|1 = 1
(%) in 17 countries. 0 1 3 8 9
a Determine the median, M. 1 2 3 7
b Determine the quartiles Q1 and Q3 . 2 0 1 2 5 6 6
c Calculate the IQR and the range, R. 3 0 6 7
4
5 5

Example 20 10 The histogram shows the time taken to complete a complex task by a group of students.
Find possible values for the median and quartiles of this distribution.
12
11
10
9
8
Frequency
7
6
5
4
3
2
1
0
55 60 65 70 75 80 85 90 95 100 105 110 115
time taken to complete task (minutes)
11 A group of 195 people were asked to record (to one decimal place) the average number
of hours they spent on email each week over a 10 week period. The data are shown in
the following histogram:
60
55
50
45
40
Frequency
35
30
25
20
15
10
5
0
0 5 10 15 20 25 30 35
hours spent per week on email
a Find possible values for the median.

b Find the maximum value for the IQR.

Determining the mean, median and mode from data

Example 21 12 For each of the following data sets, write down the value of n, the value of Σx and
hence evaluate x̄.
a 2 5 2 3 b 12 15 20 32 25 c 2 1 3 2 5 3 5
13 Calculate the mean and locate the median and modal value(s) of the following scores.
a 1 3 2 1 2 6 4 5 4 3 2
b 3 12 5 4 3 2 6 5 4 5 5 6
14 The temperature of a hospital patient (in degrees Celsius) taken at 6-hourly intervals
over 2 days was as follows.
35.6 36.5 37.2 35.5 36.0 36.5 35.5 36.0
a Calculate the patient’s mean and median temperature over the 2-day period.
b What do these values tell you about the distribution of the patient’s temperature?
15 The amounts (in dollars) spent by seven customers at a corner store were:
0.90 0.80 2.15 16.55 1.70 0.80 2.65
a Calculate the mean and median amount spent by the customers.
b Does the mean or the median give the best indication of the typical amount spent by
customers? Explain your answer.
16 For which of the following distributions might you question using the mean as a
measure of the centre of the distribution? Justify your selection.
a b c
Age distribution in a country Urban car accident rates Blood cholesterol levels
17 The stem plot shows the distribution of weights (in Weight (kg)
kg) of 22 footballers. 6 9
a Name the shape of the distribution. Which 7 0 2
measure of centre, the mean or the median, do 7 6 6 7 8
you think would best indicate the typical weight
8 0 0 1 2 3 3 4
of these footballers?
8 5 5 5 6
b Determine both the mean and median to check
9 1 2
your prediction.
9 8
10 3

The concept of standard deviation

18 Which measure of spread:
a always incorporates 50% of the scores?
b uses only the smallest and largest scores in the distribution?
c gives the average variation around the mean?
19 Without using the statistical capabilities of your calculator, write down the mean and
standard deviation of the following six data values: 7.1 7.1 7.1 7.1 7.1 7.1
20 For which of the following variables does it not make sense to calculate a mean or
standard deviation?
a Speed (in km/h) b Sex c Age (in years)
d Post code e Neck circumference (in cm)
f Weight (underweight, normal, overweight)
Calculating the mean and standard deviation using a CAS calculator

21 A sample of 10 students were given a general knowledge test with the following
results.
20 20 19 21 21 18 20 22 23 17
a Calculate the mean and standard deviation of the test scores, correct to one decimal
place.
b The median test score is 20, which is similar in value to the mean. What does this
tell you about the distribution of test scores?
22 Calculate the mean and standard deviation for the variables in the table.
Give answers to the nearest whole number for cars and TVs, and one decimal place for
alcohol consumption.
Number of TVs/ Number of cars/ Alcohol consumption

1000 1000 (litres)
378 417 17.6
404 286 12.5
471 435 16.0
354 370 24.1
539 217 9.9
381 357 9.5
624 550 14.6


The stem plot displays the number of times each person in a sample of 16 people bought a
take-away coffee in the last month.
Key: 0|1 = 1
0 0 0 8 9
1 2 3 7
2 0 1 2 5 6 6
3 0 6 7
23 The median, M, of the number of take away coffees bought is equal to:
A 10.5 B 18.5 C 20.5 D 21 E 26
24 The interquartile range, IQR, of the number of take away coffees bought is equal to:
A 10.5 B 15.0 C 15.5 D 18.5 E 20.5

The following histogram shows the length (in cm) for each of 159 fish.
40
36
32
28
Frequency
24
20
16
12
8
4
0
0 5 10 15 20 25 30 35 40 45 50 55 60
length(cm)
25 The median fish length (in cm) could be

A 15.8 B 24.3 C 25.2 D 31.4 E 80.0
26 The first quartile (Q1 ) for this distribution could be

A 12.2 B 16.7 C 29.0 D 20.2 E 25.0

1G The five-number summary and the boxplot 61
1G The five-number summary and the boxplot

Learning intentions
I To be able to construct the boxplot for displaying the distribution of a numerical data.
I To be able to define and identify outliers.
I To be able to construct both simple boxplots and boxplots with outliers.
I To be able to determine the features of a distribution from a boxplot.
I To be able to use a CAS calculator to construct a boxplot.
The five-number summary

Knowing the median and quartiles tells us quite a lot about the centre and spread of the
distribution. If we also knew something about the tails (ends) we would have a good picture
of the whole distribution. This can be achieved by recording the smallest and largest values
of the data set. Putting all this information together gives the five-number summary.
Five-number summary
A listing of the median, M, the quartiles Q1 and Q3 , and the smallest and largest data
values of a distribution, written in the order
minimum, Q1 , M, Q3 , maximum
is known as a five-number summary.
The five-number summary is the starting point for constructing one of the most useful
graphical tools in data analysis, the boxplot.
The boxplot
The boxplot (or box-and-whisker plot) is a graphical display of a five-number summary. The
essential features of a boxplot are summarised below.
The boxplot
A boxplot is a graphical display of a five-number summary.
box
whisker whisker
minimum Q1 M Q3 maximum
median
25% 25% 25% 25%

In a boxplot:
a box extends from Q1 to Q3 , locating the middle 50% of the data values
the median is shown by a vertical line drawn within the box
lines (called whiskers) are extended out from the lower and upper ends of the box to
the smallest and largest data values of the data set respectively
25% of the data values are from the minimum to Q1
25% of the data values are from Q1 to the median M
25% of the data values are from the median M to Q3
25% of the data values are from Q3 to the maximum
Example 22 Constructing a boxplot from a five-number summary
The stem plot shows the distribution of life Key: 5|2 = 52 years
expectancies (in years) in 23 countries. 5 2 minimum
The five-number summary for these data is: 5 5 6
6 4 Q1
minimum 52 6 6 6 7 9 median
first quartile (Q1 ) 66 7 1 2 2 3 3 4 4 4 4
Q3
median (M) 73 7 5 5 6 6 7 7 maximum
third quartile (Q3 ) 75
maximum 77
Use the five-number summary to construct a boxplot.

1 Draw a labelled and scaled number line
50 55 60 65 70 75 80
that covers the full range of values. Life expectancy (years)
2 Draw a box starting at Q1 = 66 and ending

at Q3 = 75.
50 55 60 65 70 75 80
Life expectancy (years)
3 Mark the median value with a vertical line

segment at M = 73.
50 55 60 65 70 75 80

4 Draw the whiskers: lines joining the

midpoint of the ends of the box to the
minimum and maximum values, 52 and 77. 50 55 60 65 70 75 80
Boxplots with outliers

An extension of the boxplot can also be used to identify possible outliers in a data set.
Sometimes it is difficult to decide whether or not an observation is an outlier. For example, a
boxplot might have one extremely long whisker. How might we explain this?
One explanation is that the data distribution is extremely skewed with lots of data values
in its tail.
Another explanation is that the long whisker hides one or more outliers.
By modifying the boxplots, we can decide which explanation is most likely, but firstly we
need a more exact definition of an outlier.
Defining outliers
Outlier
An outlier in a distribution is any data point that lies more than 1.5 interquartile ranges
below the first quartile or more than 1.5 interquartile ranges above the third quartile.
To be more informative the boxplot can be modified so that the outliers are plotted
individually in the boxplot with a dot or cross, and the whisker now ends only to the largest
or smallest data value that is not outside these limits.
An example of a
boxplot with outliers
is shown opposite.
0 10 20 30 40 50 60
Three of the data values 30, 40, and 60 are possible outliers.
To display outliers on a boxplot, we must first determine the location of what we call the
upper and lower fences. These are imaginary lines drawn one and a half interquartile ranges
(or box widths) above and below the box ends, as shown in the diagram following. Data
values outside these fences are then classified as possible outliers and plotted separately.

Using a boxplot to display outliers

In a boxplot, possible outliers are defined as being those values that are:
greater than Q3 + 1.5 × IQR (upper fence)
less than Q1 − 1.5 × IQR (lower fence).
outliers outlier
1.5 IQR 1.5 IQR
lower fence upper fence
When drawing a boxplot, any observation identified as an outlier is shown by a dot. The
whiskers end at the smallest and largest values that are not classified as outliers.
While we have used a five-number summary as the starting point for our introduction to
boxplots, in practice the starting point for constructing a boxplot is raw data. Constructing a
boxplot from raw data is a task for your CAS calculator.
CAS 4: How to construct a boxplot with outliers using the TI-Nspire

CAS
Display the following set of 19 marks in the form of a boxplot with outliers.
28 21 21 3 22 31 35 26 27 33
43 31 30 34 48 36 35 23 24
Steps
1 Start a new document by pressing /+N.
2 Select Add Lists & Spreadsheet. Enter the data
into a list called marks as shown.
3 Statistical graphing is done through the
Data & Statistics application.
Press / + I and select Add Data & Statistics.

Note: A random display of dots will appear – this

indicates that list data are available for plotting. Such
a dot is not a statistical plot.
a Click on the Click to add variable on the
x-axis and select the variable marks. A
dot plot is displayed by default as shown
opposite.
b To change the plot to a boxplot press
b>Plot Type>boxplot. Your screen
should now look like that shown opposite.
4 Data analysis
Key values can be read from the boxplot
by moving the cursor over the plot or using
b>Analyze>Graph Trace.
Starting at the far left of the plot, we see that the:
minimum value is 3 (an outlier)
first quartile is 23 (Q1 = 23)
median is 30 (Median = 30)
third quartile is 35 (Q3 = 35)
maximum value is 48.
CAS 4: How to construct a boxplot with outliers using the ClassPad

Display the following set of 19 marks in the form of a boxplot with outliers.
28 21 21 3 22 31 35 26 27 33
43 31 30 34 48 36 35 23 24
Steps
1 Open the Statistics
application and
enter the data into the column
labelled marks.
2 Open the Set StatGraphs
dialog box by tapping
in the toolbar. Complete the
dialog box as shown below.
Draw: select On.
Type: select MedBox ( ).
XList: select main\marks ( ).
Freq: leave as 1.
Tap the Show Outliers box to tick ( ).

3 Tap Set to confirm your

selections and plot the boxplot
b tapping . The graph is
drawn in an automatically
scaled window, as shown.
4 Tap the icon at the bottom
of the screen for a full-screen
graph.
Note: If you have more than one
graph on your screen, tap the data
screen, select StatGraph and turn
off any unwanted graphs.
5 Tap to read key values. This places a marker on the

boxplot (+), as shown. Use the horizontal cursor arrows
( ) and ( ) to move from point to point on the boxplot.
We see that the:
minimum value is 3 (minX = 3; an outlier)
first quartile is 23 (Q1 = 23)
median is 30 (Med = 30)
third quartile is 35 (Q3 = 35)
maximum value is 48 (maxX = 48).
Interpreting boxplots
Constructing a boxplot is not an end in itself. The prime reason to construct boxplots is to
help us answer statistical questions. To do this, you need to know how to read values from a
boxplot and use them to determine statistics such as the median, the interquartile range and
the range. We also use boxplots to identify possible outliers.
Example 23 Reading values from a boxplot
For the boxplot shown, write down the

values of:
a the median
b the quartiles Q1 and Q3 0 10 20 30 40 50 60 70 80 90 100
c the interquartile range (IQR)

d the minimum and maximum values
e the values of any possible outliers
f the smallest value in the upper end of the data set that will be classified as an outlier
g the largest value in the lower end of the data set that will be classified as an outlier.

a The median (the vertical line in the box) M = 36
b Quartiles Q1 and Q3 (end points of box) Q1 = 30, Q3 = 44
c Interquartile range (IQR = Q3 − Q1 ) IQR = Q3 − Q1 = 44 − 30 = 14
d Minimum and maximum values (extremes) Min = 4, Max = 92
e The values of the possible outliers (dots) 4, 70, 84 and 92 are possible
outliers
f Upper fence (given by Q3 + 1.5 × IQR) Upper fence = Q3 + 1.5 × IQR
= 44 + 1.5 × 14 = 65
Any value above 65 is an outlier.
g Lower fence (given by Q1 − 1.5 × IQR) Lower fence = Q1 − 1.5 × IQR
= 30 − 1.5 × 14 = 9
Any value below 9 is an outlier.
Once we know the location of the quartiles, we can use the boxplot to estimate percentages.
Example 24 Estimating percentages from a boxplot
For the boxplot shown, estimate the percentage of

values:
a less than 54 b less than 55
50 55 60 65 70 75 80
c less than 59 d greater than 59
e between 54 and 59 f between 54 and 86.
a 54 is the first quartile (Q1 ); 25% of values are less than Q1 . a 25%
b 55 is the median or second quartile (Q2 ); 50% of values are less b 50%
than Q2 .
c 59 is the third quartile (Q3 ); 75% of values are less than Q3 . c 75%
d 75% of values are less than 59 and 25% are greater than 59. d 25%
e As 75% of values are less than 59 and 25% are less than 54, 50% of e 50%
values are between 54 and 59.
f As 100% of values are less than 86 and 25% of values are less than f 75%
54, 75% of values are between 54 and 86.

Relating a boxplot to shape

When there are a reasonable number of data values, the shape of a distribution can be
identified from a boxplot.
A symmetric distribution
A symmetric distribution tends to be centred on its
median and have values evenly spread around the
median. As a result, its boxplot will also be symmetric,
its median is close to the middle of the box and its
whiskers are approximately equal in length. Q1 M Q3
Positively skewed distributions

Positively skewed distributions are characterised by a positive skew
cluster of data values around the median at the left-hand
end of the distribution with a gradual tailing off to the
right.
As a result, the boxplot of a positively skewed distribution
Q1 M Q3
will have its median off-centre and to the left-hand side
of its box. The left-hand whisker will be short, while the
right-hand whisker will be long, reflecting the gradual
tailing off of data values to the right.
Negatively skewed distributions

Negatively skewed distributions are characterised by negative skew
a clustering of data values around the median at the
right-hand end of the distribution, with a gradual tailing
off of data values to the left.
As a result, the boxplot of a negatively skewed Q1 M Q3
distribution has the median off-centre and in the right-
hand side of its box. The right-hand whisker will be
short, while the left-hand whisker will be long, reflecting
the gradual tailing off of data values to the left.

Distributions with outliers

Distributions with outliers are characterised by large
gaps between the main body and data values in the
tails. The histogram opposite displays a distribution
with an outlier. In the corresponding boxplot, the box
and whiskers represent the main body of data and the
dot, separated by a gap from the box and whiskers, an Q1 M Q3
outlier.
Using boxplots to describe a distribution

Because of the wealth of information contained in a boxplot, it is an extremely powerful tool
for describing the features of distribution in terms of shape, centre, spread and outliers.
Example 25 Using a boxplot to describe the features of distribution without

outliers
Describe the distribution represented by the boxplot in terms of shape, centre and spread.
Give appropriate values.
0 5 10 15 20 25 30 35 40 45 50
Solution
The distribution is positively skewed with no outliers. The distribution is centred at 10,
the median value. The spread of the distribution, as measured by the IQR, is 16 and, as
measured by the range, 45.
Example 26 Using a boxplot to describe the features of a distribution with

outliers
Describe the distributions represented by the boxplot in terms of shape and outliers,
centre and spread. Give appropriate values.
0 5 10 15 20 25 30 35 40 45 50
Solution
The distribution is symmetric but with outliers. The distribution is centred at 41, the
median value. The spread of the distribution, as measured by the IQR, is 5.5 and, as
measured by the range, 37. There are four outliers: 10, 15, 20 and 25.

70 Chapter 1 Investigating data distributions 1G
Earlier in this chapter we attempted to describe the feature of a distribution from a

histogram. We found that from the histogram it was difficult to give exact values for centre
and spread, and to clearly identify outliers. This is much easier to do from a boxplot.
Example 27 Using a boxplot to answer statistical questions
The boxplot shows the gestation period (completed weeks) for a sample for 1000 babies
born in Australia one year. Describe the distribution of gestation period in terms of shape,
centre, spread and outliers.
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
Gestation period (weeks)
Solution
The distribution of gestational period is negatively skewed with several outliers. The
distribution is centred at 39 weeks, the median value. The range of the distribution is 17
weeks, but the interquartile range is only 2 weeks. Any gestational period less than 35
weeks or less is considered unusual, with outliers at 25, 26, 27, 28, 29, 30, 31, 32, 33 and
34 weeks.
Exercise 1G
Constructing a five-number summary from a dot or stem plot

1 Construct a five-number summary for the dot
plot opposite.
4 5 6 7 8 9
Hours worked
2 Construct a five-number summary for the stem plot Key: 13|6 = 136
opposite. 13 6 7
14 3 6 8 8 9
15 2 5 8 8 8
16 4 5 5 6 7 9
17 8 8 9
18 2 9

1G 1G The five-number summary and the boxplot 71
Constructing a boxplot from a five-number summary

Example 22 3 Use the following five-number summaries to construct boxplots.
a Min = 1, Q1 = 4, M = 8, Q3 = 13.5, Max = 24
b Min = 136, Q1 = 148, M = 158, Q3 = 169, Max = 189
4 a Construct a boxplot from the following five-number summary:

Min = 10, Q1 = 22, M = 40, Q3 = 70, Max = 70
b Explain why the box has no upper whisker.
5 University participation rates (%) in 21 countries are listed below.
3 3 7 8 9 12 13 15 17 20 21
22 25 26 26 26 27 30 36 37 55
a Show that the five number summary for this data is:
Min =3, Q1 = 10.5, M = 21, Q3 = 26.5, Max = 55
b Show that the upper fence is equal to 50.5.
c Explain why this boxplot will show at least one outlier.
d Construct a boxplot showing the outlier.
6 The five-number summary for a data set is:

Min = 14 Q1 = 40 M = 55 Q3 = 62 Max = 99
a Determine the values of the upper and lower fences.

b The smallest three values in the data set are 6, 18, 34 and the largest are 90, 94, 99.
Which of these are outliers?
Constructing a boxplot using a CAS calculator

7 The reaction times (in milliseconds) of 18 people are listed below.
38 36 35 35 43 46 42 64 40 48 35 34 40 44 30 25 39 31
a Use a CAS calculator to construct a boxplot with outliers for the data. Name the
variable rtime.
b Use the boxplot to construct a five-number summary. Identify the outlier.
Reading values from a boxplot

Example 23 8 For each of the boxplots below, estimate the values of:
i the median, M ii the quartiles Q1 and Q3
iii the interquartile range, IQR iv the minimum and maximum values
v the values of possible outliers.

0 5 10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50
30 32 34 36 38 40 42 44 46 48 50 52 54 56
9 For the boxplots below, determine the location of:

i the upper fence ii the lower fence.
0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
10 a Determine the lower

fence for the boxplot
opposite.
0 5 10 15 20 25 30 35 40 45 50

b When the data were originally entered, a value of 31 was incorrectly entered as 35.
Would the 31 be shown as an outlier when the error is corrected? Explain your
answer.
Reading percentages from a boxplot

Example 24 11 Use the boxplot
opposite to estimate the
percentage of values that
are: 0 5 10 15 20 25 30 35 40 45 50
a less than 39 b less than 45 c greater than 45

d between 39 and 45 e between 5 and 45.
12 The boxplot displays the monthly rainfall (in
mm) for 12 months.
Use the boxplot to estimate the percentage of
months in which the monthly rainfall was: 45 50 55 60 65 70
Rainfall (mm)
a greater than 59 mm b less than 49.5 mm c between 49.5 and 59 mm

d between 57 and 59 mm e less than 59 mm f between 57 and 70 mm.
13 Match these boxplots with their histograms.
Boxplot 1 Boxplot 2
Boxplot 3 Boxplot 4
Histogram A Histogram B
Histogram C Histogram D

Describing the features of a distribution from a boxplot

Example 25 14 Describe the distributions represented by the following boxplots in terms of shape,
centre, spread. Give appropriate values.
a
0 5 10 15 20 25 30 35 40 45 50
b
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400
Example 26 15 Describe the distributions represented by the following boxplots in terms of shape,
centre, spread and outliers (if any). Give appropriate values.
a
0 5 10 15 20 25 30 35 40 45 50
b
0 5 10 15 20 25 30 35 40 45
c
0 5 10 15 20 25 30 35 40 45 50
Using a boxplot to answer statistical questions

Example 27 16 Taj recorded his travel time to university (in minutes) each day for 60 days, and
summarised the data in the following boxplot. Write a brief report describing the
distribution of his travel time.
60 65 70 75 80 85 90 95 100 105 110 115 120


The boxplot below shows the distribution of marks scored by 200 students in a test.
0 5 10 15 20 25 30 35 40 45 50
17 The percentage of students who scored more than 24 marks is closest to:
A 15% B 25% C 50% D 75% E 100%
18 The five-number summary for the test scores is closest to:

A 1, 15, 18, 24, 50
B 12, 15, 18, 24, 30
C 1, 12, 15, 18, 24
D 12, 15, 18, 24, 50
E 12, 15, 24, 30, 50

The weights (in gm) of 159 fish were measured, and the table gives the mean and the
five-number summary for this data.
mean 398
minimum 20
first quartile (Q1 ) 120
median (M) 273
third quartile (Q3 ) 650
maximum 1650
19 The shape of the distribution is best described as

A approximately symmetric B positively skewed
C symmetrically skewed D uniform
E negatively skewed
20 The largest five values in the data set are 1100gm, 1250gm, 1550gm, 1600gm and
1650gm. Which of these are outliers?

A {1100, 1250, 1550, 1600,1650} B {1650}

C {1250, 1550, 1600,1650} D {1550, 1600,1650}
E {1600,1650}
1H The normal distribution and the 68–95–99.7%

rule
Learning intentions
I To be able to introduce the normal model for bell-shaped distributions.
I To be able to use the 68 -95 - 99.7% rule to estimate percentages and give meaning to
the standard deviation.
I To be able to calculate standardised scores and use them to compare values across
distributions.
We know that the interquartile range is the spread of the middle 50% of the data set. Can we
find some similar way in which to interpret the standard deviation?
It turns out we can, but we need to restrict ourselves to symmetric distributions that have
an approximate bell shape. Again, while this may sound very restrictive, many of the data
distributions we work with in statistics (but not all) can be well approximated by this type of
distribution. In fact, it is so common that it is called the normal distribution.
The normal distribution

Many data sets that arise in practice are roughly symmetrical and have approximate bell
shapes, as shown in the four examples below.
170 175 180 185 55 60 65 70 75 80 85 90 95

Blood pressure Weight (kilograms)
2.0 2.5 3.0 3.5 4.0 4.5 5 6 7 8 9 10 11 12 13 14

Sepal width of flowers (centimetres) Hand span measurements (centimetres)
Data distributions that are bell-shaped can be modelled by a normal distribution.
The 68–95–99.7% rule

In normal distributions, the percentage of observations that lie within a certain number of
standard deviations of the mean can always be determined. In particular, we are interested
in the percentage of observations that lie within one, two or three standard deviations of the
mean. This gives rise to what is known as the 68–95–99.7% rule.
1H The normal distribution and the 68–95–99.7% rule 77
The 68–95–99.7% rule

For a normal distribution, approximately:
68% of the observations lie within one standard deviation of the mean, that is in the
interval ( x̄ − s, x̄ + s).
95% of the observations lie within two standard deviations of the mean, that is in the
interval ( x̄ − 2s, x̄ + 2s).
99.7% of the observations lie within three standard deviations of the mean, that is in
the interval ( x̄ − 3s, x̄ + 3s).
To give you an understanding of what this rule means in practice, it is helpful to view this
rule graphically.
68-95-99.7% Rule
68%
95%
99.7%
x–3s x–2s x–s x x+s x+2s x+3s
Since the normal distribution is symmetrical, and 100% of the observations are within the
normal curve, we can use the 68–95–99.7% rule to allocate percentages to the tails of the
distribution in each instance.
Since around 68% of the data values will lie within
one standard deviation (SD) of the mean, then we
can also say that around 16% of values lie in each
16% 68% 16%
of the tails.
Since around 95% of the data values will lie within
two standard deviations of the mean, then we can 95%
also say that around 2.5% of values lie in each of 2.5% 2.5%
the tails.

Since around 99.7% of the data values will

99.7%
lie within three standard deviations of the
0.15% 0.15%
mean we can also say that around 0.15%
of values lie in each of the tails.
Putting together all of this information, we can then allocate percentages of the data which
fall into each section of the normal curve, as shown in the following diagram:
0.15% 2.35% 13.5% 34% 34% 13.5% 2.35% 0.15%
x–3s x–2s x–s x x+s x+2s x+3s
Example 28 Applying the 68–95–99.7% rule
The distribution of delivery times for pizzas made by House of Pizza is approximately
normal, with a mean of 25 minutes and a standard deviation of 5 minutes.
a What percentage of pizzas have delivery times of between 15 and 35 minutes?
b What percentage of pizzas have delivery times of greater than 30 minutes?
c In 1 month, House of Pizza delivers 2000 pizzas. Approximately many of these pizzas
are delivered in less than 10 minutes?
a 1 Sketch, scale and label a normal
mean = 25
distribution curve with a mean of SD = 5
25 and a standard deviation of 5.
10 15 20 25 30 35 40
Delivery time

2 Shade the region under the normal

mean = 25
curve representing delivery times of
SD = 5
between 15 and 35 minutes.
95%
3 Note that delivery times of between
15 and 35 minutes lie within two
10 15 20 25 30 35 40
standard deviations of the mean. Delivery time
(15 = 25 − 2 × 5 and 35 = 25 + 2 × 5)
4 95% of values are within two 95% of pizzas will have delivery times of
standard deviations of the mean. between 15 and 35 minutes.
Use this information to write your
answer.
b 1 As before, draw, scale and label a
mean = 25
normal distribution curve with a
SD = 5
mean of 25 and a standard deviation
of 5. Shade the region under the
normal curve representing delivery 16%
times of greater than 30 minutes. 10 15 20 25 30 35 40

Delivery time
2 Delivery times of greater than
30 minutes are more than one
standard deviation above the mean.
(30 = 25 + 1 × 5)
3 16% of values are more than one 16% of pizzas will have delivery times of
standard deviation above the mean. greater than 30 minutes.
Write your answer.
c 1 Write down the number of pizzas Number = 2000
delivered.
2 Delivery times of less than Percentage delivered in less than 10
10 minutes are more than three minutes = 0.15%
standard deviations below the mean.
(10 = 25 − 3 × 5).
3 0.15% of values are more than three
standard deviations below the mean.
Record this.
4 Therefore, the number of pizzas Number of pizzas delivered in less than
delivered in less than 10 minutes is 10 minutes ≈ 0.15% of 2000
0.15% of 2000. 0.15
= × 2000 = 3
100

Standard scores
The 68–95–99.7% rule makes the standard deviation a natural measuring stick for normally
distributed data.
For example, a person who obtained a score of 112 on an IQ test with a mean of 100
and a standard deviation of 15 has an IQ score less than one standard deviation from the
mean. Her score is typical of the group as a whole, as it lies well within the middle 68% of
scores. In contrast, a person who scores 133 stands out; her score is more than two standard
deviations from the mean and this puts her in the top 2.5%.
Because of the additional insight provided by relating the standard deviations to percentages,
it is common to transform data into a new set of units that show the number of standard
deviations a data value lies from the mean of the distribution. This is called standardising
and these transformed data values are called standardised or z-scores.
Calculating standardised (z) scores

To obtain a standard score for an actual score, subtract the mean from the score and then
divide the result by the standard deviation. That is:
actual score − mean x − x̄
standard score = or z =
standard deviation s
Let us check to see that the formula works.

We already know that an IQ score of 115 is one standard deviation above the mean, so it
should have a standard or z-score of 1. Substituting into the formula above we find, as we
had predicted, that:
115 − 100 15
z= = =1
15 15
Standard scores can be both positive and negative:
a positive z-score indicates that the actual score it represents lies above the mean
a z-score of zero indicates that the actual score is equal to the mean
a negative z-score indicates that the actual score lies below the mean.
Example 29 Calculating standard scores
The heights of a group of young women have a mean of x̄ = 160 cm and a standard
deviation of s = 8 cm. Determine the standard or z-scores of a woman who is:
a 172 cm tall b 150 cm tall c 160 cm tall.

1 Write down the data value (x), the a x = 172, x̄ = 160, s = 8
mean ( x̄) and the standard deviation (s). x − x̄ 172 − 160 12
z= = = = 1.5
2 Substitute the values into the formula s 8 8
x − x̄ b x = 150, x̄ = 160, s = 8
z= and evaluate. x − x̄ 150 − 160 10
s z= = =− = −1.25
s 8 8
c x = 160, x̄ = 160, s = 8
x − x̄ 160 − 160 0
z= = = =0
s 8 8
Using standard scores to compare performance

Standard scores are also useful for comparing groups that have different means and/or
standard deviations. For example, consider a student who obtained a mark of 75 in
Psychology and a mark of 70 in Statistics. In which subject did she do better?
Calculating standard scores

We could take the marks at face value and say that Standard
she did better in Psychology because she got a Subject Mark Mean Deviation
higher mark in that subject. The assumption that
Psychology 75 65 10
underlies such a comparison is that the marks for
Statistics 70 60 5
both subjects have the same distribution with
the same mean and standard deviation. However, in this case the two subjects have very
different means and standard deviations, as shown in the table above.
If we assume that the marks are normally distributed, then standardisation and the
68–95–99.7% rule give us a way of resolving this issue.
Let us standardise the marks.
75 − 65
Psychology: standardised mark z= =1
10
70 − 60
Statistics: standardised mark z= =2
5
What do we see? The student obtained a higher score for Psychology than for Statistics.
However, relative to her classmates she did better in Statistics.
Her mark of 70 in Statistics is equivalent to a z-score of 2. This means that her mark was
two standard deviations above the mean, placing her in the top 2.5% of students.
Her mark of 75 for Psychology is equivalent to a z-score of 1. This means that her mark
was only one standard deviation above the mean, placing her in the top 16% of students.
This is a good performance, but not as good as for statistics.

Example 30 Using standardised scores to make comparisons
Another student studying the same two subjects obtained a mark of 55 for both
Psychology and Statistics. Does this mean that she performed equally well in both
subjects? Use standardised marks to help you arrive at your conclusion.
1 Write down her mark (x), the mean ( x̄) Pyschology: x = 55, x̄ = 65, s = 10
and the standard deviation (s) for each x − x̄ 55 − 65 −10
z= = = = −1
subject and compute a standardised s 10 10
score for both subjects. Statistics: x = 55, x̄ = 60, s = 5
x − x̄ 55 − 60 −5
z= = = = −1
s 5 5
2 Write down your conclusion. Yes, her standardised score, z = −1, was
the same for both subjects. In both subjects
she finished in the bottom 16%.
Using standard scores to determine percentages

Now we know how to determine standard scores we can revisit the diagram showing the
percentages associated with each section of the normal curve, replacing the horizontal scale
with the values of Z (the standard normal scores).
0.15% 2.35% 13.5% 34% 34% 13.5% 2.35% 0.15%
z=–3 z=–2 z=–1 z=0 z=1 z=2 z=3
Example 31 Using standard scores to determine percentages
Suppose the weight of a certain species of bird is normally distributed with a mean of 42
grams with a standard deviation of 3 grams.
a If a bird selected at random from this population has a standardised weight of z = −1,
what percentage of birds in this population weigh more than this bird?
b Approximately what percentage of birds would weigh between 39 and 48 grams?

a Locate z = −1 on the graph above. We can see that the percentage of the
distribution below z = −1 is 16%, so the
percentage above z = −1 is 84%.
b 1 Substitute the values of x into the x̄ = 42, s = 3
x − x̄ x − x̄ 39 − 42
formula z = and evaluate. x = 39 z= = = −1
s s 3
2 Locate z = −1 and z = 2 on the x − x̄ 48 − 42
x = 48 z = = =2
graph above, and determine the s 3
The percentage of the distribution
percentage of the distribution
between z = −1 and z = 2 is 81.5%.
between these values.
Converting standard scores into actual scores

Having learnt how to calculate standard scores, you also need to be able to convert a
standardised score back into an actual score. The rule for converting a standardised score
into an actual score is given below.
Converting standardised scores into actual scores

By making the actual score the subject of the rule for calculating standard scores, we
arrive at:
actual score = mean + standard score × standard deviation or x = x̄ + z × s
Example 32 Converting standard scores into actual scores
A class test (out of 50) has a mean mark of x̄ = 34 and a standard deviation of s = 4.
Joe’s standardised test mark was z = −1.5. What was Joe’s actual mark?
1 Write down mean ( x̄), the standard deviation (s) and x̄ = 34, s = 4, z = −1.5
Joe’s standardised score (z).
2 Write down the rule for calculating the actual score x = x̄ + z × s
and substitute these values into the rule. = 34 + (−1.5) × 4 = 28
Joe’s actual mark was 28.
If we know something about the percentages associated with a normal distribution, we can
use this information to find the values of the mean, or standard deviation, or both.
In the following example we are given the value of the mean, and one percentage associated
with the distribution. From this we can determine the value of the standard deviation.

Example 33 Finding the value of the standard deviation given the mean and one
percentage
Suppose the heights of red flowering gum trees have a mean of 10.2 metres, and 2.5% of
these trees grow to more than 11.4 metres tall. Assuming that the heights of these trees
are approximately normally distributed, what is the standard deviation of the height of the
red flowering gum trees?
1 Since 2.5% of the trees are taller than x̄ = 10.2, z = 2
11.4 metres, this height corresponds to
a z-score of 2.
2 Write down the rule for calculating the x = x̄ + z × s
actual score and substitute these values 11.4 = 10.2 + 2 × s
into the rule.
3 Solve this equation for s. 2 × s = 1.2

s = 0.6 metres
Example 34 Finding the value of the standard deviation given the mean and two
percentages
The marks scored in an examination are known to be approximately normally distributed.

If 16% of students score more than 80 marks, and 2.5% of students score less than 20
marks, estimate the mean and standard deviation of this distribution.
1 Since 2.5% of the students score less x̄ − 2 × s = 20 (1)
than 20, this value corresponds to a
z-score of -2.
2 Since 16% of the students score more x̄ + 1 × s = 80 (2)
than 20, this value corresponds to a
z-score of 1.
3 To solve these equations for x̄ and s, (1)-(2) 3 × s = 60

subtract equation (2) from equation (1). Hence s = 20
4 To find x̄ substitute the value of s in x̄ − 2 × 20 = 20

equation 1. Hence x̄ = 60

1H 1H The normal distribution and the 68–95–99.7% rule 85
Skill-
sheet Exercise 1H
The 68–95–99.7% rule

Example 28 1 The blood pressure readings for executives are approximately normally distributed with
a mean systolic blood pressure of 134 and a standard deviation of 20.
Given this information it can be concluded that:
a about 68% of the executives have blood pressures between and
b about 95% of the executives have blood pressures between and
c about 99.7% of the executives have blood pressures between and
d about 16% of the executives have blood pressures above
e about 2.5% of the executives have blood pressures below
f about 0.15% of the executives have blood pressures below
g about 50% of the executives have blood pressures above .
2 The weight of a bag of 10 blood plums picked at U-Pick Orchard is normally

distributed with a mean of 1.88 kg and a standard deviation of 0.2 kg.
Given this information the percentage of the bags of 10 plums that weigh:
a between 1.68 and 2.08 kg is approximately %
b between 1.28 and 2.48 kg is approximately %
c more than 2.08 kg is approximately %
d more than 2.28 kg is approximately %
e less than 1.28 kg is approximately %
f more than 1.88 kg is approximately %.
3 The distribution of times taken for walkers to complete a circuit in a park is normal,
with a mean time of 14 minutes and a standard deviation of 3 minutes.
a What percentage of walkers complete the circuit in:
i more than 11 minutes? ii less than 14 minutes?
iii between 14 and 20 minutes?
b In a week, 1000 walkers complete the circuit. Approximately how many will take
less than 8 minutes?
4 The distribution of heights of 19-year-old women is approximately normal, with a

mean of 170 cm and a standard deviation of 5 cm.
a What percentage of these women have heights:
i between 155 and 185 cm? ii greater than 180 cm?
iii between 160 and 175 cm?
b In a sample of 5000 of these women, approximately how many have heights greater
than 175 cm?

86 Chapter 1 Investigating data distributions 1H
5 The distribution of resting pulse rates of a sample of 2000 20-year-old men is

approximately normal, with a mean of 66 beats/minute and a standard deviation of 4
beats/minute.
a What percentage of these men have pulse rates of:
i higher than 66? ii between 66 and 70? iii between 62 and 74?
b Approximately how many of this sample of 2000 men have pulse rates between 54
and 78 beats/minute?
Calculating standard scores

Example 29 6 A set of scores has a mean of 100 and a standard deviation of 20. Calculate
standardised scores for each of the following test scores:
a 120 b 140 c 80 d 100 e 40 f 110
7 A set of scores has a mean of 30 and a standard deviation of 7. Calculate standardised

scores for each of the following test scores. Give your answers to one decimal place.
a 37 b 23 c 40 d 20
Applying standardised scores

Example 30 8 The table below contains the scores a student obtained in a practice test for each of
their VCE subjects. Also shown are the mean and standard deviation for each subject.
Subject Mark Mean Standard deviation

English 69 60 4
Biology 75 60 5
Chemistry 55 55 6
Further Maths 55 44 10
Psychology 73 82 4
a Calculate the standardised score for each subject.

b Use the standardised score to rate the student’s performance in each subject,
assuming a normal distribution of marks and using the 68–95–99.7% rule.
Using standardised scores to determine percentages

Example 31 9 The volume of soft drink in a small can is normally distributed with a mean of 300 mL
and a standard deviation of 2 mL.
a If a can selected at random from this population has a standardised volume of z = 2,
what percentage of cans in this population contain more soft drink than this can?
b Approximately what percentage of cans contain between between 302 mL and
306 mL?

1H 1H The normal distribution and the 68–95–99.7% rule 87
10 To be considered for a special training program applicants are required to sit for an
aptitude test. Suppose that 2000 people sit for the test, and their scores on the aptitude
test are approximately normally distributed with a mean of 45 and a standard deviation
of 2. People who score more than 49 are selected for the special training program.
People who are not chosen for the training program, but score more than 47, are invited
to resit the aptitude test at a later date.
a What percentage of people who sat for the test are eligible for the training program?
b Approximately how many people would be invited to resit the aptitude test at a later
date?
Calculating actual scores from standardised scores

Example 32 11 A set of scores has a mean of 100 and a standard deviation of 20.
Calculate the actual score if the standardised score was:
a 1 b 0.8 c 2.1 d 0 e −1.4 f −2.5
Find the values of the mean and standard deviation

Example 33 12 The mean salary for retail assistants is $27/hr. If 2.5% of retail assistants earn more
than $30/hr, what is the standard deviation of the salary for retail assistants? Assume
that the salaries are approximately normally distributed.
13 The weights of bananas from a certain grower are approximately normally distributed.
If the standard deviation of the weight of these bananas is 5 g, and 16% of the bananas
weigh less than 96 g, what is the mean weight of the bananas?
Example 34 14 The birth weights of babies are known to be approximately normally distributed. If
16% of babies weigh more than 4.0 kg, and 0.15% of babies weigh more than 5.0 kg,
estimate the mean and standard deviation of this distribution. Give your answers to one
decimal place.
15 The marks scored in an examination are known to be approximately normally

distributed. If 99.7% of students score between 43 and 89 marks, estimate the mean
and standard deviation of this distribution. Give your answers to one decimal place.
Applications
16 The body weights of a large group of 14-year-old girls have a mean of 54 kg and a
standard deviation of 10.0 kg.
a Kate weighs 56 kg. Determine her standardised weight.
b Lani has a standardised weight of –0.75. Determine her actual weight.
c Find:
i percentage of these girls who weigh more than 74 kg
ii percentage of these girls who weigh between 54 and 64 kg

88 Chapter 1 Investigating data distributions 1H
iii percentage of these girls who have standardised weights less than –1
iv percentage of these girls who have standardised weights greater than –2.
17 Suppose that IQ scores are normally distributed with mean of 100 and standard
deviation of 15.
a What percentage of people have an an IQ:
i greater than 115?
ii less than 70?
b To be allowed to join an elite club, a potential member must have an IQ in the
top 2.5% of the population. What IQ score would be necessary to join this club?
c One student has a standardised score of 2.2. What is their actual score?
18 The heights of women are normally distributed with a mean of 160 and a standard
deviation of 8.
a What percentage of women would be:
i taller than 152 cm?
ii shorter than 176 cm?
b What height would put a woman among the tallest 0.15% of the population?
c What height would put a woman among the shortest 2.5% of the population?
d One woman has a standardised height of -1.2. What is her actual height? Give your
answer to one decimal place.

Use the following information to answer questions 19 - 22
A total of 16,000 students sat for a statewide mathematics exam. Their results are are
normally distributed with mean 50 and standard deviation 7.
19 The percentage of students in the state who scored more than 71 marks is closest to:
A 0.15% B 2.5% C 5% D 15% E 0.3%
20 The top 2.5% of the state are to be awarded a distinction. What would be the lowest
mark required to gain a distinction in this exam?
A 36 B 43 C 57 D 64 E 71
21 Approximately how many students gained a mark of 57 or more?

A 400 B 800 C 2560 D 5120 E 15200
22 Approximately how many students gained a mark between 43 and 64?

A 10888 B 11120 C 13040 D 13440 E 15200

1H 1I Populations and samples 89
23 The table below shows Miller’s swimming times (in seconds) for 50 metres in each of
butterfly, backstroke, breaststroke and freestyle. Also shown are the mean and standard
deviation of the times recorded for all of the swimmers in his swimming club. In how
many of these swimming styles is he in the fastest 2.5% of swimmers at his swimming
club?
Style Miller’s time Mean Standard deviation
Butterfly 38.8 46.2 3.2
Breaststroke 51.4 55.1 4.1
Backstroke 53.5 48.3 2.5
Freestyle 33.3 38.2 2.3
A 0 B 1 C 2 D 3 E 4
1I Populations and samples

This material is available in the Interactive Textbook.

Review 90 Chapter 1 Investigating data distributions
Key ideas and chapter summary
Univariate data Univariate data are generated when each observation involves
Assign- recording information about a single variable.
ment
Types of data Data can be classified as numerical or categorical.
Categorical Categorical variables are used to represent characteristics of

variables individuals. Categorical variables come in two types: nominal and
ordinal. Nominal variables generate data values that can only be used
by name, e.g. eye colour. Ordinal variables generate data values that
can be used to both name and order, e.g. house number.
Numerical Numerical variables have data values which are quantities. Numerical
variables variables come in two types: discrete and continuous. Discrete
variables are those which may take on only a countable number of
distinct values such as 0 1 2 3 4 . . . and are often associated with
counting. Continuous variables are ones which take an infinite number
of possible values, and are often associated with measuring rather than
counting.
Frequency table A frequency table lists the values a variable takes, along with how
often (frequently) each value occurs. Frequency can be recorded as:
the number of times a value occurs – e.g. the number of Year 12
students in the data set is 32.
the percentage of times a value occurs – e.g. the percentage of Year
12 in the data set is 45.5%.
Bar chart Bar charts are used to display frequency distribution of categorical
data. Each value of the variable is represented by a bar showing the
frequency, or the percentage frequency.
Segmented bar A segmented bar chart is like a bar chart, but the bars are stacked one
chart on top of another to give a single bar with several segments.
Mode, modal The mode (or modal interval) is the value of a variable (or the interval)
category that occurs most frequently.
Histogram A histogram uses columns to display the frequency distribution of a

numerical variable: suitable for medium to large-sized data sets.

Chapter 1 Review 91
Review
Describing the The distribution of a numerical variable can be described in terms of:
distribution of
shape: symmetric or skewed (positive or negative)
a numerical
variable outliers: values that seem unusually small or large.
centre: the median or mean.
spread: the IQR , range or the standard deviation.
Dot plot A dot plot consists of a number line with each data point marked by a
dot. A dot plot is particularly suitable for displaying a small data set of
discrete numerical data.
Stem plot The stem plot is particularly suitable for displaying a small to medium
sized data sets of numerical data.It shows each data value separated into
two parts: the leading digits, which make up the stem of the number,
and its last digit, which is called the leaf.
Log scales Log scales can be used to transform a skewed histogram to symmetry.
Summary Summary statistics are numerical values for special features of a data
statistics distribution such as centre and spread.
Mean The mean ( x̄) is a summary statistic that can be used to locate the centre
of a symmetric distribution.
P The value of the mean is determined from
x
the formula: x̄ =
n
Range The range (R) is the difference between the smallest and the largest data
values. It is the simplest measure of spread.
Standard The standard deviation (s) is a summary statistic that measures the
deviation spread of the data values around the mean. The value of the standard
deviation is determined from the formula:
s
(x − x̄)2
P
s=
n−1
Median The median (M) is a summary statistic that can be used to locate the
centre of a distribution. It is the midpoint of a distribution, so that
50% of the data values are less than this value and 50% are more. It is
sometimes denoted as Q2
Quartiles Quartiles are summary statistics that divide an ordered data set into
four equal groups.

Interquartile The interquartile range (IQR) gives the spread of the middle
range 50% of data values in an ordered data set. It is found by evaluating
IQR = Q3 − Q1
Five-number The median, the first quartile, the third quartile, along with the
summary minimum and the maximum values in a data set, are known as a
five-number summary.
Outliers Outliers are data values that appear to stand out from the rest of the
data set. They are values that are less than the lower fence or more than
the upper fence.
Lower and upper The lower fence is equal to Q1 − 1.5 × IQR.

fences The upper fence is equal to Q3 + 1.5 × IQR.
Boxplot A boxplot is a visual display of a five-number summary with

adjustments made to display outliers separately when they are present.
The normal Data distributions that have a bell shape can be modelled by a normal
distribution distribution.
The 68-95- For a data distribution which is approximately normally distributed

99.7%rule approximately:
68% of the data will lie within one standard deviation of the mean.
95% of the data will lie within two standard deviations of the mean.
99.7% of the data will lie within three standard deviations of the
mean.
Standardised The value of the standard score gives the distance and direction of a data
scores value from the mean in terms of standard deviations.
The rule for calculating a standardised score is:
actual score − mean
standardised score =
standard deviation

Chapter 1 Review 93
Review
Skills checklist
Download this checklist from the Interactive Textbook, then print it and fill it out to check
Check- your skills. X
list
1A 1 I can identify types of data.
See Example 1, and Exercise 1A Question 1
1B 2 I can construct a frequency table for categorical data.
See Example 2, and Exercise 1B Question 1
1B 3 I can construct a bar chart from a frequency table.
1B 4 I can construct a percentage segmented bar chart from a frequency table.
1B 5 I can describe the distribution of a categorical variable.
1C 6 I can construct a frequency table for discrete numerical data.
See Example 6, and Exercise 1C Question 1
1C 7 I can construct a grouped frequency table.
1C 8 I can construct a histogram from a grouped frequency table.
1C 9 I can describe the features of a distribution from a histogram.
1D 10 I can construct a dot plot.
See Example 10, and Exercise 1D Question 1
1D 11 I can construct a stem plot.
1E 12 I can use a CAS calculator to find logs.
See Example 12, and Exercise 1E Question 1

1E 13 I can interpret a histogram with a log scale.
1F 14 I can find the median value in a data set.
See Example 14, and Exercise 1F Question 1
1F 15 I can find the median value from a dot plot.
1F 16 I can find the median value from a stem plot.
1F 17 I can find the range from a stem plot.
1F 18 I can find the IQR from a stem plot when n is even.
1F 19 I can find the IQR from a stem plot when n is odd.
1F 20 I can calculate the mean using the formula.
1F 21 I can calculate the mean and standard deviation using the CAS calculator.
See CAS 3, and Exercise 1F Question 12
1G 22 I can construct a boxplot from a five number summary.
See Example 22, and Exercise 1G Question 3
1G 23 I can construct a boxplot with outliers from data using the CAS calculator.
See CAS 4, and Exercise 1G Question 7
1G 24 I can read values from a boxplot.
1G 25 I estimate percentages from a boxplot.
1G 26 I can use boxplots for describe a distribution without outliers.

Chapter 1 Review 95
Review
1G 27 I can use boxplots to describe a distribution with outliers.
1G 28 I can use boxplots to answer statistical questions.
1H 29 I can apply the 68-95-99.7% rule.
See Example 28, and Exercise 1H Question 1
1H 30 I can calculate standardised scores.
1H 31 I can use standardised scores to make comparisons.
1H 32 I can use standard scores to determine percentages.
1H 33 I can convert standardised scores into actual scores.
1H 34 I can solve for the values of the mean and standard deviation.
See Example 33 and 34, Exercise 1H Question 12 and Question 14.
Multiple-choice questions
The following information relates to Questions 1 and 2.
Data pertaining to the following five variables was collected about secondhand cars:
colour
model
number of seats
age (1 = less than 2 years, 2 = from 30,000-60,000km, 3 = more than 80,000km)
mileage (in kilometres)
1 The number of these variables that are discrete numerical is:

A 1 B 2 C 3 D 4 E 5
2 The number of ordinal variables is:

A 0 B 1 C 2 D 3 E 4

The percentage segmented bar chart shows the 100 Hair color
distribution of hair colour for 200 students. 90 Other
80 Red
3 The number of students with brown hair is Black
70 Brown
closest to:
Blonde
Percentage
60
A 4 B 34 C 57
50
D 72 E 114
40
4 The most common hair colour is: 30
A black B blonde 20
C brown D red E other 10
0
Questions 5 to 7 relate to the histogram shown below.
The histogram below displays the test scores of a class of students.

6
5
Frequency
4
3
2
1
0
6 8 10 12 14 16 18 20 22 24 26 28
Test score
5 The number of students in the class who obtained a test score less than 14 is:
A 4 B 10 C 14 D 16 E 28
6 The shape of the histogram is best described as:

A negatively skewed B positively skewed with an outlier
C approximately symmetric D approximately symmetric with an outlier
E negatively skewed with an outlier
7 The value of the first quartile could be:

A 5.5 B 6.8 C 8.9 D 10.5 E 11.4
8 log10 100 equals:

A 0 B 1 C 2 D 3 E 100
9 Find the number with log equal to 2.314; give the answer to the nearest whole number.
A 2 B 21 C 206 D 231 E 20606

Chapter 1 Review 97
Review
The percentage histogram opposite displays 32%

the distribution of the log of the annual
24%
per capita CO2 emissions (in tonnes) for
Percentage
192 countries in 2011. 16%
8%
0%
−1.0 −0.5 0.0 0.5 1.0 1.5 2.0
log(CO2)
10 Australia’s per capita CO2 emissions in 2011 were 16.8 tonnes. In which column of the
histogram would Australia be located?
A −0.5 to <0.0 B 0.0 to <0.5 C 0.5 to <1.0 D 1.0 to <1.5 E 1.5 to < 2.0
11 The percentage of countries with per capita CO2 emissions of under 10 tonnes is
closest to:
A 14% B 17% C 31% D 69% E 88%
12 The following is an ordered set of 10 daily maximum temperatures (in degrees

Celsius):
22 22 23 24 24 25 26 27 28 29
The five-number summary for these temperatures is:
A 22, 23, 24, 27, 29 B 22, 23, 24.5, 27, 29 C 22, 24, 24.5, 27, 29
D 22, 23, 24.5, 27.5, 29 E 22, 24, 24.5, 27, 29
The following information relates to Questions 13 to 15.
The stem plot opposite displays the distribution of the Key: 1|5 means 15 marks
marks obtained by 25 students. 0 2
1 5 9 9 9
2 0 4 4 5 5 8 8 8
3 0 3 5 5 6 8 9
4 1 2 3 5
5
6 0
13 The median mark is:

A 20 B 27 C 28 D 29 E 30
14 The interquartile quartile range (IQR) of the marks is:

A 12 B 16.5 C 20 D 30.5 E 31.5

15 The shape of the data distribution displayed by this stem plot is best described as:
A approximately symmetric B approximately symmetric with an
outlier
C negatively skewed with an outlier D negatively skewed
E positively skewed with an outlier
A B
50 55 60 65 70 75 80 50 55 60 65 70 75 80
C D
50 55 60 65 70 75 80 50 55 60 65 70 75 80
16 The median of boxplot A is closest to:

A 5 B 53 C 54.5 D 55 E 60
17 The IQR of boxplot B is closest to:

A 9 B 20 C 25 D 65 E 75
18 The range of boxplot C is closest to:

A 4 B 13 C 20 D 31 E 80
19 The description that best matches boxplot A is:

A symmetric B symmetric with outliers
C negatively skewed D positively skewed
E positively skewed with outliers
20 The description that best matches boxplot B is:

A symmetric B negatively skewed with an outlier
21 The description that best matches boxplot D is:

A symmetric B symmetric with outliers
22 For the data represented by boxplot D, the percentage of data values greater than 65 is:
A 2.5% B 25% C 50% D 75% E 100%

Chapter 1 Review 99
Review
23 To be an outlier in boxplot D, a score must be:
A either less than 52.5 or greater than 72.5 B greater than 72.5
C either less than 55 or greater than 70 D greater than 70
E less than 55
24 The mean ( x̄) and standard deviation (s) for the following set of test marks
1 1 10 15 16 25 8 10 12
are closest to:
A 7.1, 10.9 B 7.5, 10.9 C 10.9, 7.1 D 10.9, 7.5 E 10.8, 7.5
25 It would not be appropriate to determine the mean and standard deviation of a group of
peoples’:
A ages B phone numbers C heights
D weights E family sizes
26 The median is a more appropriate measure of the centre of a distribution than the mean
when the distribution is:
A symmetric B symmetric with no C bell-shaped
outliers
D clearly skewed E normal
27 The mean length of 10 garden stakes is 180.5 cm, and the standard deviation is 2.9 cm.
If the length of each garden stake is reduced by exactly 5 cm, the mean and standard
deviation of the 10 stakes will be:
A 175.5 cm and 2.4 cm B 180.5 cm and 2.4 cm C 175.5 cm and 2.9 cm
D 175.5 cm and 3.4 cm E 185.5 and 2.9 cm
28 A student’s mark on a test is 50. The mean mark for their class is 55 and the standard
deviation is 2.5. Their standard score is:
A −2.5 B −2.0 C 0 D 2 E 2.5

Each week, a bus company makes 200 trips between two large country towns. The time
taken to make a trip between the two towns is approximately normally distributed with a
mean of 78 minutes and a standard deviation of 4 minutes.
29 The percentage of trips each week that take 78 minutes or more is:
A 16% B 34% C 50% D 68% E 84%
30 The number of trips each week that take between 70 and 82 minutes is approximately:
A 4 B 32 C 68 D 127 E 163

31 A trip that takes 71 minutes has a standardised time (z-score) of:

A −1.75 B −1.5 C −1.25 D 1.5 E 1.75
32 A standardised time for a trip is z = −0.25. The actual time (in minutes) is:
A 77 B 77.25 C 77.75 D 78.25 E 79
33 The time of a bus trip has a standardised time of z = 2.1. This time is:
A very much below average B just below average C around average
D just above average E very much above average
34 The table shows the time taken to run one kilometre Runner Time(mins)
(in minutes) by three runners. To be invited to join
Albie 7.5
the athletics team the standardised score for their time
needs to be no more than than 0.5. Lincoln 4.9
Wendy 8.0
If the mean running time for one kilometre is 7.0 minutes and the standard deviation is
1.2 minutes, who will be invited to join the athletics team?
A Only Wendy B Lincoln and Albie C Only Lincoln
D Albie and Wendy E All three runners
35 The diameter of bolts produced by a machine is normally distributed. If 2.5% of bolts

the bolts have a diameter of more than 4.94 mm, and 0.15% have a diameter less
than 4.84 mm, the mean and standard deviation of this distribution in millimetres are
closest to:
A mean = 4.88 standard deviation =0.02
B mean = 4.92 standard deviation = 0.01
C mean = 4.90 standard deviation = 0.001
D mean = 4.90 standard deviation = 0.1
E mean = 4.90 standard deviation = 0.02
Written response questions

1 A group of 52 teenagers were asked, Frequency
‘Do you agree that the use of marijuana
Legalise Number Percentage
should be legalised?’ Their responses are
summarised in the table. Agree 18
Disagree 26
a Construct a labelled and scaled
frequency bar chart for the data. Don’t know 8
b Complete the table by calculating the Total 52
percentages to one decimal place.

Chapter 1 Review 101
Review
c Use the percentages to construct a percentage segmented bar chart for the data.
d Write a short report describing the distribution of responses.
2 Students were asked how much they spent 10

on entertainment each month. The results
are displayed in the histogram. Use the 8
Frequency
histogram to answer the following questions.
6
a How many students:
i were surveyed? 4
ii spent from $100 to less than $105 per
2
month?
b What is the mode? 0
90 100 110 120 130 140
c How many students spent $110 or more
Amount ($)
per month?
d What percentage spent less than $100 per month?
e i What is the shape of the distribution displayed by the histogram?
ii In which interval is the median of the distribution?
iii In which interval is the upper quartile of the distribution (Q3 )?
3 The amount of weight lost in one week by 32 people who participated in a weight loss
program was recorded and displayed in the ordered stem plot below.
Weight loss (kg) key: 2|0 represents 2.0
1 5 5 7 8 9 9
2 2 2 2 3 3 4 4
2 5 6 6 7 7 8
3 0 1 3 3 4
3 5 5 5 7
4 1 2 2
4
5 0
a Describe the shape of the distribution.
b Determine the median weight loss. Give your answer to 2 decimal places.
c Find the value of the interquartile range. Give your answer to 2 decimal places.
d What percentage of this group had a weight loss of more than 3.5 kg? Give your
answer to 2 decimal places.
e Is the weight loss of 5.0 kg an outlier for this data set? Justify your answer.
4 The systolic blood pressure (measured in mmHg) for a group of 2000 people
was measured. The results are summarised in the five-number summary below:
Min = 75, Q1 = 110, M = 125, Q3 = 140, Max = 180
a Use the five-number summary to construct a simple boxplot.

b Indicate on your plot where the lower and upper fences would be, and hence if there
would be any outliers.
c Assume that the distribution of systolic blood pressure for this sample of 2000
people is approximately normally distributed, with a mean of 128 mmHg and a
standard deviation of 20 mmHg.
i Approximately what percentage of people of people have a systolic blood
pressure between 108 mmHg and 148 mmHg?
ii Suppose a person has a blood pressure three standard deviations below the mean,
what would be their actual blood pressure?
iii Of the 2000 people measured, how many could we expect to have a blood
pressure three standard deviations below the mean?
iv Of the 2000 people measured, how many actually did have a blood pressure
three standard deviations below the mean?
5 The hand span in centimetres of 200 women was recorded and displayed in the dot plot
below.
a Write down the modal hand span, in centimetres, for this group of 200 women.
b The mean hand span for this group of 200 women is 17.9 cm, and the standard
deviation is 1.1 cm. Use the information in the dot plot to determine the percentage
of women in this group who had an actual hand span more than two standard
deviations above or below the mean. Round your answer to one decimal place.
c The five-number summary for this sample of hand spans, in centimetres, is given
below:
Min =15.0, Q1 = 17.0, M = 18.0, Q3 =18.5, Max = 21.5
Use the five-number summary to construct a boxplot showing outliers.

Chapter
2
Investigating
associations between
two variables
Chapter questions
I What are bivariate data?
I What are explanatory and response variables?
I What are two-way frequency tables and how do we interpret them?
I How do we construct and interpret segmented bar charts from two-way
frequency tables?
I How do we construct and interpret parallel dot plots?
I How do we construct and interpret back-to-back stem plots?
I How do we construct and interpret parallel boxplots?
I What is a scatterplot, how is it constructed and what does it tell us?
I What do we mean when we describe the association between two numerical
variables in terms of direction, form and strength?
I What is the difference between observation and experimentation?
I What is the difference between association and causation?
In this chapter we begin our study of bivariate data, data which is recorded on two
variables from the same subject.

104 Chapter 2 Investigating associations between two variables
2A Bivariate data – Classifying the variables

Learning intentions
I To introduce bivariate data.
I To be able to classify data as categorical or numerical.
I To be able to identify explanatory and response variables.
So far you have learned how to display, describe and compare the distributions of single
variables. In the process you learned how to use data to answer questions like ‘What is the
favourite colour of prep-grade students?’ or ‘How do the weights of tuna fish vary?’ In each
case we concentrated on investigating the statistical variables individually.
However, questions like ‘Does the new treatment for headache work more quickly than
the old treatment?’, ‘Are city voters more likely to vote for the Greens party than country
voters?’ or ‘Can we predict a student’s test score from the time (in hours) they spent
studying for the test?’ cannot be answered by considering variables separately. All of these
questions relate to situations where the two variables are linked in some way (associated) so
that they vary together. The data generated in these circumstances is called bivariate data.
Analysing bivariate data requires a new set of statistical tools. Developing and applying
these tools is the subject of the next four chapters.
Categorical and numerical variables

You will recall from Chapter 1 we defined two classifications of variables, categorical and
numerical variables:
Categorical variables generate data values that are names or labels, such as favourite pet
(dog, cat, rabbit, bird) or coffee size (small, medium, large).
Numerical variables generate data values that are numbers, usually resulting from
counting or measuring, such as number of brothers (0, 1, 2, . . .) or hand span (cm).
The first step in investigating the association between two variables is to classify each
variable as either categorical or numerical. Consider again the previous questions:
Does the new treatment for headache work more quickly than the old treatment?
The two variables in this question are type of treatment, a categorical variable taking
the values ‘new’ and ‘old’, and time taken for the headache to be relieved, a numerical
variable, measured in minutes. Thus, investigation of a question like this can be classified
as investigating the association between a categorical variable and a numerical
variable.

2A Bivariate data – Classifying the variables 105
Are city voters more likely to vote for the Greens party than country voters?
This question involves two variables, place of residence, which is a categorical variable
taking the values ‘city’ and ‘country’, and vote for the Greens, which also is a categorical
variable taking the values ‘yes’ and ‘no’. Investigation of a question like this can be
classified as investigating the association between two categorical variables.
Can we predict a student’s test score (%) from time (in hours) spent studying for the test?
The variable test score is a numerical variable, as is time spent studying for the test.
Investigation of a question like this can be classified as investigating the association
between two numerical variables.
As discussed in Chapter 1, categorical variables can be further classified as nominal or
ordinal, and numerical variables can be further classified as discrete or continuous.
Example 1 Identifying associations as categorical or numerical
For each of the following questions, determine if they involve investigating associations
between
one numerical variable and one categorical variable or
two categorical variables or
two numerical variables.
a Are younger people (age measured in years) more likely to believe in astrology
(measured as ‘yes’ or ‘no’) than older people?
b Do people who weigh more (weight measured in kg) tend to have higher blood
pressure (blood pressure measured in mmHg)?
c Are people who have a driver’s licence (measured as ‘yes’ or ‘no’) more likely to be in
favour of lowering the driving age (measured as ‘yes’ or ‘no’)?
Solution
a One numerical variable (age) and one categorical variable (belief in astrology)
b Two numerical variables (weight and blood pressure)
c Two categorical variables (have a driver’s licence and support for lowering the driving
age)
Identifying response and explanatory variables

When investigating associations between variables, it is helpful to think of one of the
variables as the explanatory variable. The other variable is then called the response
variable. We use the explanatory variable to explain changes that might be observed in the
response variable.
For example, the question, ‘Are city voters more likely to vote for the Greens party than
country voters?’, suggests that knowing a person’s place of residence might be useful in

explaining voting preference. In this situation place of residence is the explanatory variable
and vote for Greens is the response variable.
It is important to be able to identify the explanatory and response variables before you
explore the association between them. Consider the following examples.
Example 2 Identifying the response and explanatory variables
We wish to investigate the question, ‘Does the time it takes a student to get to school
depend on their mode of transport?’ The variables here are time and mode of transport.
Which is the response variable (RV) and which is the explanatory variable (EV)?
In asking the question in this way we are suggesting EV: mode of transport
that a student’s mode of transport might explain the RV: time
differences we observe in the time it takes students to
get to school.
Example 3 Identifying the response and explanatory variables
Can we predict people’s height (in cm) from their wrist measurement? The variables in
this investigation are height and wrist measurement. Which is the response variable (RV)
and which is the explanatory variable (EV)?
Since we wish to predict height from wrist circumfer- EV: wrist measurement
ence, we are using wrist measurement as the predictor RV: height
or explanatory variable. Height is then the response
variable.
It is important to note that, in Example 3, we could have asked the question the other way
around; that is, ‘Can we predict people’s wrist measurement from their height?’ In that case
height would be the explanatory variable, and wrist measurement would be the response
variable. The way we ask our statistical question is an important factor when there is no
obvious explanatory variable.
Response and explanatory variables

When investigating the association between two variables the explanatory
variable (EV) is the variable we expect to explain or predict the value of the response
variable (RV).
Note: The explanatory variable is sometimes called the independent variable (IV) and the response variable
the dependent variable (DV).

2A 2A Bivariate data – Classifying the variables 107
Exercise 2A
Identifying variables as categorical or numerical

Example 1 1 For each of the following questions, determine if they involve investigating associa-
tions between:
one numerical and one categorical variable or
two categorical variables or
two numerical variables.
a Are full-time and part-time students equally likely travel to university by car?
b Do Year 11 students watch more hours of television each week than Year 12
students?
c Do countries with higher household incomes ($) tend to have lower infant mortality
rates (deaths/1000 births)?
d Is there a relationship between attitude to gun control and country of birth?
Identifying explanatory and response variables

Example 2 2 For each of the following situations identify the explanatory variable (EV)
Example 3 and response variable (RV). In each situation the variable names are italicised.
a We wish to investigate whether a fish’s toxicity can be predicted from its colour. We
want to be able to predict toxicity from colour.
b The relationship between weight loss and type of diet is to be investigated.
c We wish to investigate the relationship between a used car’s age and its price.
d It is suggested that the cost of heating in a house depends on the type of fuel used.
e The relationship between the price of a house and its location is to be investigated.
3 The following pairs of variables are related. In each case identify which is likely
to be the explanatory variable and which is the response variable, and the level of
measurement of each variable (categorical or numerical). The variable names are
italicised.
a exercise level (1 = light, 2 = moderate, 3 = a lot) and age (years).
b years of education (years) and salary level ($ per annum).
c comfort level (0 = uncomfortable, 1 = comfortable) and temperature (◦ C).
d time of year (summer, autumn, winter, spring) and incidence of hay fever (1 = never,
2 = sometimes, 3 = regularly).
e age group (less than 25, 25 - 40, more than 40) and musical taste (classical, rock,
rap, country, indie, dance, jazz).
f AFL team supported and state of residence.

108 Chapter 2 Investigating associations between two variables 2A

4 Respondents to a survey question "How concerned are you about climate change?"
were asked to select from the following responses:
1 = not at all, 2 = a little, 3 = moderately, 4 = extremely
The data which was collected in response to this question is:
A nominal B ordinal C discrete
D continuous E numerical
5 The variables weight (light, medium, heavy) and height (less than 160cm, 160-175cm,
over 175cm) are:
A both nominal variables
B both ordinal variables
C a nominal and an ordinal variable respectively
D an ordinal and a nominal variable respectively
E both continuous variables
6 Researchers believe that reaction time might be lower in cold temperatures. They
devise an experiment where reaction time in seconds is measured at three different
temperature levels (1 = less than 8◦ C, 2 = from 8◦ C to 18 ◦ C, 3 = more than 18◦ C).
The response variable, and its classification are:
A reaction time, categorical B temperature, categorical
C reaction time, numerical D temperature, numerical
E temperature, ordinal
2B Investigating associations between categorical

variables
Learning intentions
I To be able to summarise data from two categorical variables using two-way frequency
tables.
I To be able to appropriately percentage two-way frequency tables.
I To be able to use a percentaged two-way frequency table to identify and describe an
association between two categorical variables.
If two variables are related or linked in some way, we say they are associated. To begin the
investigation of an association between two categorical variables we create a contingency
table or a two-way frequency table. It is called a two-way frequency table because it is
summarising data from two variables.

2B Investigating associations between categorical variables 109
Constructing a two-way frequency table

It has been suggested that city and country people have differing attitudes to gun control;
that is, that support for gun control depends on where a person lives. How might we
investigate this relationship? Suppose we ask a sample of three people about their attitude to
gun control, and we also record their residence. The resulting data for the three people might
look like this:
Subject no. Residence Attitude to gun control

1 City For
2 Country For
3 City Against
The first thing to note is that these two variables, attitude to gun control (for or against)
and residence (city or country), are both categorical variables. Categorical data are usually
presented in the form of a frequency table.
Suppose we continue until we have interviewed a sample Residence Frequency
of 100 people, and we find that there are 58 who live in the
Country 58
country and 42 who live in the city. We can present this
result in a frequency table as shown to the right. City 42
Total 100
From this table, we can see that there were more country
than city people in our sample.
Suppose also when we record the attitude to Attitude to gun control Frequency
gun control, we might have 62 ‘for’ and 38
For 62
‘against’ gun control. Again, we could present
these results in a frequency table as shown to Against 38
the right. Total 100
From this table, we can see that more people in the sample were for gun control than against
gun control. However, we cannot tell from the information contained in the tables whether
attitude to gun control depends on the residence of the person. To do this we need to
construct a two-way frequency table, which gives both the attitude to gun control and the
residence for each person in the sample.
We begin by counting the number of people in the sample who are:
from the country and for gun control
from the city and for gun control
from the country and against gun control
from the city and against gun control.

Suppose again from our sample of 100 people we find the following frequencies:
32 country people are for gun control
30 city people are for gun control
26 country people are against gun control
12 city people are against gun control.
Explanatory and response variables in two-way frequency tables

Before we set up the two-way frequency table, we need to decide which is the explanatory
variable and which is the response variable of the two variables. Since we think that a
person’s attitude to gun control might depend on their place of residence, but not the other
way around, then:
residence is the explanatory variable (EV)
attitude to gun control is the response variable (RV).
In two-way frequency tables, it is conventional to let the categories of the response variable
label the rows of the table and the categories of the explanatory variable label the columns
of the table. Following this convention, we can create the following two-way frequency
table.
Residence
Attitude to gun control Country City
For 32 30
Against 26 12
To complete the table, it is usual to calculate the row and column sums, as shown below.
Residence
Attitude to gun control Country City Total
For 32 30 62 Row sum
Against 26 12 38 Row sum
Total 58 42 100
Column sum Column sum
The shaded regions in the table are called the cells of the table. It is the numbers in these
cells that we look at when investigating the relationship between the two variables.
Example 4 Constructing a two-way frequency table
The following data were obtained when a sample of ten Year 9 students were asked
if they intended to go to university (university). The gender of the student was also
recorded.

Student Gender University Student Gender University

1 Female Yes 6 Male Yes
2 Male Yes 7 Female Yes
3 Female No 8 Male No
4 Female Yes 9 Female No
5 Male No 10 Female Yes
Create a two-way frequency table from these data.

1 We first need to identify the It is possible that a student’s intention to go
explanatory variable and the response to university may depend on their gender,
variable. but not the other way around. Thus, gender
is the explanatory variable and university is
the response variable.
2 Create the table showing the values Gender
of Gender labelling the columns, and
University Male Female
University labelling the rows.
Yes
No
3 Consider Student 1, who is female and Gender

indicated yes to go to university. Place
a mark in the corresponding cell of the
table. Yes |
No
4 Go through the data set one person at a Gender

time, placing a mark in the appropriate
cell for each person.
Yes || ||||
No || ||
5 Finally, tally the marks in each cell, Gender

and then calculate the row and columns
University Male Female Total
sums. Make sure the total adds to the
number of students in the sample. Yes 2 4 6
No 2 2 4
Total 4 6 10

Consider again the two-way frequency table created to investigate the association between
place of residence and attitude to gun control. This table tells us that more country people
are in favour of gun control than city people. But is this just due to the fact that there were
more country people in the sample? To help us answer this question we need to express the
frequencies in each cell as percentage frequencies.
Percentaged two-way frequency table

When the two-way frequency table has been constructed so that the values of the
explanatory variable label the rows, then we calculate column percentages to help us
investigate the association. This will give us the percentage of country and the percentage of
city people for and against gun control, which can then be compared.
Column percentages are determined by dividing each of the cell frequencies by the relevant
column sums. Thus, the percentage of:
32
country people who are for gun control is: × 100 = 55.2%
58
26
country people who are against gun control is: × 100 = 44.8%
58
30
city people who are for gun control is: × 100 = 71.4%
42
12
city people who are against gun control is: × 100 = 28.6%
42
Note: Unless small percentages are involved, it is usual to round percentages to one decimal place in tables.
Residence
Attitude to gun control Country City
For 55.2% 71.4%
Against 44.8% 28.6%
Total 100.0% 100.0%
Using percentages to identify relationships between

variables
Calculating the values in the table as percentages enables us to compare the attitudes of city
and country people on an equal footing. From the table, we see that 55.2% of country people
in the sample were for gun control compared to 71.4% of the city people. This means that
the city people in the sample were more supportive of gun control than the country people.
This reverses what the frequencies showed.
The fact that the percentage of ‘country people for gun control’ differs from the percentage
of ‘city people for gun control’ indicates that a person’s attitude to gun control depends on
their residence. Thus, we can say that the variables attitude to gun control and residence are
associated.

If the variables attitude to gun control and residence were not associated, we would expect
approximately equal percentages of country people and city people to be ‘for’ gun control.
Finding a single row in the two-way frequency distribution in which percentages are clearly
different is sufficient to identify a relationship between the variables.
We could have also arrived at this conclusion by focusing our attention on the percentages
‘against’ gun control. We might report our findings as follows.
Report
In this sample of 100 people, a higher percentage of city people were for gun control than
country people: 71.4% to 55.2%. This indicates that a person’s attitude to gun control is
associated with their place of residence.
We will now consider a two-way percentage frequency table that shows no evidence of a
relationship. Consider the following table that summarises responses to the question ‘Should
mobile phones be banned in cinemas?’ These responses were obtained from 100 students
in Year 10 and Year 12 – we are interested in investigating whether there is an association
between these variables.
Year level
Should mobile phones Year 10 Year 12
be banned in cinemas?
Yes 87.9% 86.8%
No 12.1% 13.2%
Total 100.0% 100.0%
When we look across the first row of the table, we see that the percentages in favour are very
similar. In this case, we might report our findings as follows.
Report
In this sample of 100 Year 10 and Year 12 students, we see that the percentage of Year 10
and Year 12 students in support of banning mobile phones in cinemas is similar: 87.9%
to 86.8%. This indicates that a person’s support for banning mobile phones in cinemas is
not associated with their year level.

Example 5 Identifying and describing an association from a percentaged

two-way table (2 × 2)
Are males and females in Year 9 equally likely to indicate an intention to go to

university? Data from interviews with 200 Year 9 students are summarised in the
following table. Write a brief report addressing this question and quoting appropriate
percentages.
Gender
University Male Female Total
Yes 50 54 104
No 55 41 96
Total 105 95 200
1 Determine the column Gender
percentages and complete the
table as shown.
Yes 47.6% 56.8%
No 52.4% 43.2%
Total 100.0% 100.0%
2 Select an appropriate row to We can see from the top row that a greater
compare the male and female proportion of females than males (56.8% compared
percentages. with 47.6%) were intending to go to university.
3 Construct a report. Report: In this sample of 200 Year 9 students, a
greater proportion of females than males (56.8%
compared with 47.6%) were intending to go to
university. There is an association between gender
and intention to go to university.
Two-way frequency tables for categorical variables taking more than two
values
The table below displays the smoking status for a group of adults (smoker, past smoker,
never smoked) by educational level (Year 9 or less, Year 10 or 11, Year 12, university). This
is still a two-way frequency table (because involves two variables), each of these variables
can take three values, and so we call this a 3×3 table.

Education level (%)

Smoking status Year 9 or less Year 10 or 11 Year 12 University
Smoker 34.0 31.7 26.5 18.4
Past smoker 36.0 33.8 30.9 28.0
Never smoked 30.0 34.5 42.6 53.6
Total 100.0 100.0 100.0 100.0
Again, we look for an association between variables by comparing the percentages across
one of the rows. The following report has been prepared using the percentages in the
‘Smoker’ row.
Report
From Table 3.3 we see that the percentage of smokers steadily decreases with education
level, from 33.9% for Year 9 or below to 18.4% for university. This indicates that
smoking is associated with level of education.
Example 6 Identifying and describing associations from a percentaged two-way

table (3 × 3)
A survey was conducted with 1000 Age group (%)

males under 50 years old. As part of
Interest in Under 18 19–25 26–35 36–50
this survey, they were asked to rate their sport years years years years
interest in sport as ‘high’, ‘medium’, or
High 56.5 50.2 40.7 35.0
‘low’. Their age group was also recorded
Medium 30.1 34.4 36.8 44.7
as ‘under 18’,‘19–25’, ‘26–35’ and
‘36–50’. The results are displayed in the Low 13.4 15.4 22.5 20.3
table. Total 100.0 100.0 100.0 100.0
a Which is the explanatory variable, interest in sport or age group?

b Is there an association between interest in sport and age group? Write a brief response
quoting appropriate percentages.
a Age is a possible explanation for the Age group is the EV.
level of interest in sport, but interest in
sport cannot explain age.

b If we look across all rows, we can There is an association between the level
see that the percentages are different of interest in sport and age. A high level
for each age group. Select one row to of interest in sport is seen to decrease
compare and discuss – here we have steadily across the age categories from
chosen ‘high’. 56.5% for under 18 years, 50.2% for 19–25
years, 40.7% for 26–35 years to, at its
lowest, 35% for 36–50 years.
The segmented bar chart

A visual display which which can be used to display the information in a two-way frequency
table is a segmented bar chart. A segmented bar chart consists of separate bars for each
value of the explanatory variables, with each bar separated into parts (segments) that show
the percentage for each value of the response variable.
The following segmented bar chart below that has been constructed from the table
displaying the smoking status of adults (smoker, past smoker, never smoked) by level of
education (Year 9 or less, Year 10 or 11, Year 12, university).
100
Smoking status
80 Never smoked
Past smoker
Percentage
60
Smoker
40
20
0
Yr 9 or less Yr 10 or 11 Yr 12 University
Level of education
Example 7 Constructing a segmented bar chart
Construct a segmented bar chart to display the association interest in sport and age group
displayed in the table in Example 6.
Solution
1 Since age group is the EV, this variable will label the horizontal axis.
2 The vertical axis should be scaled from 0% to 100%, in intervals of 10%.
3 There will be a bar for each value of age group, that is, a bar for each column of the
table.
4 Mark off, and colour, with each value of interest in sport assigned in the same colour.
5 Add a Key showing which colour has been assigned to each value of interest in sport.

100%
90%
80%
70%
Interest in sport
60%
50% Low
40% Medium
30% High
20%
10%
0%
under 18 18-25 26-35 36-50
Age group (years)
The percentaged segmented bar chart allows an easier visual comparison of the percentages
than does the percentaged two-way table, and can be used to investigate the association
between two categorical variables, as shown in the following example.
Example 8 Identifying and describing associations from a segmented bar chart
The percentaged segmented bar chart below shows the association between preferred
holiday (country or coast) and age group (under 40, 40 or over) for a sample of 800
visitors to a travel website.
100%
90%
80%
70%
Preferred holiday
60% country
50% coast
40%
30%
20%
10%
0%
under 40 40 and over
Age group
Does the percentaged segmented bar chart support the contention there is an association
between preferred holiday and age group?
If we look across all rows, we can see that There is an association between the
the heights of the segments, and thus the holiday preference and age. Those aged
percentages, are different for each age under forty are more likely to choose a
group. Select one row to compare and coastal holiday (75%) than those aged
discuss – here we have chosen ‘coast’. forty or over (60%).

118 Chapter 2 Investigating associations between two variables 2B
Skill-
sheet Exercise 2B
Constructing a two-way frequency table

Example 4 1 The following data were obtained when a sample of 20 Year 12 students were asked
if they intended to go to university (university). The gender of the student was also
recorded.
Student Gender Intends to go Student Gender Intends to go
No. to university No. to university
1 F Yes 11 F Yes
2 M Yes 13 M Yes
3 F No 13 F No
4 F Yes 14 F Yes
5 M No 15 M No
6 M Yes 16 M Yes
7 F Yes 17 F Yes
8 M No 18 M No
9 F No 19 F No
10 F Yes 20 F Yes
a Identify which is the explanatory and which is the response variable.

b Create a two-way frequency table from the data, with the values of the explanatory
variable labelling the columns.
Example 5 2 The following data were obtained when a sample of 30 adults were asked if they
supported reducing university fees. They were also classified by their age group: 17–18
years, 19–25 years, or 26 years or more. The results are given in the table below.
Age group Reduce fees Age group Reduce fees Age group Reduce fees
17–18 Yes 26 or more Yes 26 or more No
19–25 Yes 17–18 Yes 19–25 Yes
26 or more No 19–25 Yes 17–18 No
17–18 Yes 17–18 Yes 26 or more Yes
19–25 Yes 17–18 Yes 17–18 No
26 or more Yes 26 or more No 26 or more Yes
17–18 Yes 19–25 Yes 19–25 Yes
19–25 No 26 or more Yes 17–18 Yes
26 or more No 17–18 No 19–25 No
19–25 No 17–18 Yes 26 or more Yes

2B 2B Investigating associations between categorical variables 119
a Identify which variable is the explanatory variable and which is the response
variable.
b Create a two-way frequency table from these data, with the values of the
explanatory variable labelling the columns.
c Calculate the column percentages for the table.
Using two-way tables to identify associations between two categorical variables

Example 6 3 A survey was conducted with 242 university Drinks Enrolment status (%)
students. For this survey, data were collected
alcohol Full-time Part-time
on the students’ enrolment status (full-time,
Yes 80.5 81.8
part-time) and whether or not each drinks
alcohol (‘Yes’ or ‘No’). Their responses are No 19.5 18.2
summarised in the table opposite. Total 100.0 100.0
a Which variable is the explanatory variable?

b Is there an association between drinking alcohol and enrolment status? Write a brief
report quoting appropriate percentages.
4 The table opposite was constructed from data Gender

collected to see if handedness (left, right) was
Handedness Male Female
associated with gender (male, female).
Left 22 16
a Which variable is the response variable?
Right 222 147
b Convert the table to percentages by
calculating column percentages.
c Is handnesss is associated with gender? Write a brief explanation using appropriate
percentages.
5 A survey was conducted with 59 students Course (%)

studying Business and 51 students studying
Exercise Business Arts
Arts at university to determine whether
Rarely 28.8 39.2
they exercise, ‘regularly’, ‘sometimes’ or
‘rarely’. Their responses are summarised in Sometimes 52.5 54.9
the percentaged two-way frequency table. Regularly 18.6 5.9
Total 99.9 100.0
a Which is the explanatory variable?

b Is the variable exercise nominal or ordinal?
c What percentage of Arts students exercised sometimes?
d Is there an association between how regularly these students exercise and their
course? Write a brief response quoting appropriate percentages.

Example 7 6 It was suggested that students in Dr Evans’ mathematics class would achieve higher
grades than students in Dr Smith’s mathematics class. The following table shows the
results for each class that year.
Class
Exam grade Dr Evans Dr Smith Total
Fail 2 3 5
Pass 11 20 31
Credit or above 5 9 14
Total 18 32 50
a Construct a percentaged two-way frequency table.

b Construct a percentaged segmented bar chart.
c Write a brief report on the association between teacher and grade.
Example 8 7 Are those people who are satisfied with their job more likely to be satisfied with their
life? Data collected from a survey of 200 adults are summarised in the following
percentaged segmented bar chart.
100%
90%
80%
70%
Satisfaction with life
60% dissatisfied
50% satisfied
40%
30%
20%
10%
0%
satisfied dissatisfied
Satisfaction with job
Does the data support the contention that people who are satisfied with their job are
more likely to be satisfied with their life? Write a brief report quoting appropriate
percentages.
8 Researchers predicted that using a special pillow would be more effective in curing
snoring than treatment with drugs. The association between the outcome of treatment
and type of treatment is shown in the following percentaged segmented bar chart.

2B 2B Investigating associations between categorical variables 121
100%
90%
80%
70%
Treatment outcome 60%
50% no cure
40% partial cure
30% complete cure
20%
10%
0%
drug pillow
Type of treatment
a Identify which variable is the explanatory variable and which is the response
variable.
b Does the data support the contention the special pillow is more effective at
treating snoring than the drug treatment? Write a brief report quoting appropriate
percentages.
9 As part of the General Social Survey conducted in the US, respondents were asked to
say whether they found life exciting, pretty routine or dull. Their marital status was
also recorded as married, widowed, divorced, separated or never married. The results
are organised into a table as shown.
Marital status (%)

Attitude to life Married Widowed Divorced Separated Never
Exciting 47.6 33.8 46.7 45.9 52.3
Pretty routine 48.7 54.3 47.6 44.6 44.4
Dull 3.7 11.9 6.7 9.5 3.2
Total 100.1 100.0 100.0 100.0 99.9
a What percentage of widowed people found life ‘dull’?

b What percentage of people who were never married found life ‘exciting’?
c What is the likely explanatory variable in this investigation?
d Is the variable attitude to life nominal or ordinal?
e Does the information you have been given support the contention that a person’s
attitude to life is related to their marital status? Justify your argument by quoting
appropriate percentages.


Use the following information to answer Questions 10-12
The data in the following table was collected to investigate the association between tertiary
qualifications and happiness.
Tertiary qualification
Happy with life Yes No Total
Yes 116 138 254
No 12 34 46
Total 128 172 300
10 The percentage of participants in the study who do not have a tertiary education is
closest to:
A 57.3% B 80.2% C 54.3% D 11.3% E 19.8%
11 Of those people in the study who did not have a tertiary education, the percentage who
are happy with their lives is closest to:
A 57.3% B 80.2% C 54.3% D 11.3% E 19.8%
12 The data in the table supports the contention that there is an association between
tertiary qualifications and happiness because:
A 84.7% of people are happy.
B more people without a tertiary qualification are happy than people with a tertiary
qualification.
C 90.6% of people with a tertiary qualification are happy, compared to 80.2% of those
without a tertiary qualification.
D 54.3% of happy people do not have a tertiary qualification.
E 57.3% of people do not have a tertiary qualification, compared to 42.7% who do.
2C Investigating the association between a

numerical and a categorical variable
Learning intentions
I To be able to use parallel dot plots to identify and describe the association between a
numerical variable and a categorical variable for small data sets.
I To be able to use back-to-back stem plots to display and describe the association
between a numerical variable and a categorical variable for small data sets.
I To be able to use parallel boxplots to display the association between a numerical
variable and a categorical variable which can take two or more values.

2C Investigating the association between a numerical and a categorical variable 123
In the previous section, we learned how to identify and describe associations between two
categorical variables. In this section, we will learn to identify and describe associations
between a numerical variable and a categorical variable. Suppose, for example, we wish to
investigate the association between attendance at a revision class, and test score. Here we
can actually identify two variables. One is the variable test score, a numerical variable, and
the other is the variable attended revision class, which is a categorical variable taking the
values ‘yes’ or ‘no’.
The outcome of such an investigation will be a brief written report that compares the
distribution of the numerical variable across two or more groups, the number of groups
equal to the number of values which the categorical variable can take. The starting point for
these investigations will be, as always, a graphical display of the data. Here our options are
parallel dot plots, back-to-back stem plots or the parallel boxplots.
Using a graphical display of the data, as well as the values of the relevant summary statistics,
we can compare the distributions of the numerical variable for each value of the categorical
variable according to:
shape centre spread
If any of these are noticeably different for differing values of the categorical variable we
will conclude that the two variables are associated. Because it is often difficult to clearly
identify the shape of a distribution with a small amount of data, we usually confine ourselves
to comparing centre and spread, using the medians and IQRs, when using dot plots and
back-to-back stem plots.
Using parallel dot plots and back-to-back stem plots to

identify and describe associations
For small data sets, parallel dot plots and back-to-back stem plots are ideal displays for
identifying and describing associations between a numerical and a categorical variable.
Example 9 Using a parallel dot plot to identify and describe associations
The parallel dot plots below display the distribution of the number of sit-ups performed
by 15 people before and after they had completed a gym program.
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54
Number of sit-ups before gym program
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54
Number of sit-ups after gym program
Do the parallel dot plots support the contention that the number of sit-ups performed is
associated with completing the gym program? Write a brief explanation that compares the
distributions.

Solution
Here the numerical variable sit-ups is the response variable and the categorical variable
gym program, taking the values ‘before’ and ‘after’, is the explanatory variable.
1 Locate the median number of sit-ups performed before and after the gym program,
M = 26 and M = 32 sit-ups respectively.
2 Determine the IQR of sit-ups performed before and after the gym program, IQR = 6
and IQR = 10 sit-ups respectively.
3 There are reasonable difference in both the median and IQR of the number of sit-ups
performed before and after the gym program, evidence that the number of sit-ups
performed is associated with completing the gym program. Report your conclusion,
backed up by a brief explanation.
The median number of sit-ups performed after attending the gym program (M = 32) is
higher than the median number of sit-ups performed before attending the gym program
(M = 26). The variability in the number of sit-ups has also increased from IQR = 6 to
IQR = 10. Thus we can conclude that the number of sit-ups performed is associated
with completing the gym program.
Example 10 Using a back-to back stem plot to identify and describe associations
The back-to-back stem plot below displays the distribution of life expectancy (in years)
for the same 13 countries in 1970 and 2010.
1970 2010
2|5 = 52 years 6|7 = 67 years
8 3
4
9 4 2 5 8
9 9 9 8 7 7 1 0 6 9
4 7 1 2 4 4 6 8 9 9
8 0 0 0
Do the back-to-back stem plots support the contention that life expectancy has changed
between these two time periods?
Solution
Here the numerical variable life expectancy is the response variable and the variable year
is the explanatory variable. While year can be considered a numerical variable, because
it is only taking two values (1970 and 2010) we are treating it as a categorical variable in
this example.
1 Determine the median life expectancies for 1970 and 2010. You should find them to be
67 and 76 years, respectively.

2C Investigating the association between a numerical and a categorical variable 125
2 Determine the quartiles, and hence the values of the IQR for 1970 and 2010. You
should find them to be 12.5 years and 8 years respectively.
3 These differences in median and IQR between 1970 and 2010 are sufficient to
conclude that the distribution of life expectancy had changed over this time period.
Report your conclusion, supported by a brief explanation.
Report: There is an association between year and life expectancy. The median life
expectancy has increased between 1970 and 2010, from M = 67 years to M = 76
years. Over the same period life expectancy has also become less variable (IQR in
1970 = 12.5 years; IQR in 2010 = 8 years).
Using parallel boxplots to identify and describe

associations
The statistical tool most commonly used for investigating associations between a numerical
and a categorical variable is the parallel boxplot. In a parallel boxplot, there is one
boxplot for each category of the categorical variable. Associations can then be identified
by comparing the way in which the distribution of the numerical variable changes between
categories in terms of shape, centre and spread. We should also mention outliers when
describing the distributions.
Example 11 Comparing distributions across two groups using parallel boxplots
Use the following parallel boxplots to compare the pulse rates (in beats/minute) for a
group of 70 male students and 90 female students.
Female
Male
40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120

Resting pulse rate
Solution 0.5
Here the numerical variable resting pulse rate is the response variable and the categorical
variable gender is the explanatory variable.
1 Compare the medians: The median for females is about 72, which is higher than that
for males, which is about 65.
2 Compare the spread: The IQR for females is 15, which is more than the IQR for
males, which is 10.

3 Compare the shape: Both distributions are approximately symmetric.

4 Locate any outliers. There are two outliers for the males, one at 40 and one at 120.
5 Write the report comparing the distributions.
Report
There is an association between resting pulse rate and gender. On average, the resting
pulse rate for males is lower (median: male = 65, female = 72) and less variable than that
for females (IQR: male = 10, female = 15). The distributions of resting pulse rates for
both male and female students were approximately symmetric. One male was found to
have an extremely low pulse rate of 40, while another had an extremely high pulse rate of
120.
Example 12 Comparing distributions across more than two groups using parallel
boxplots
Use the parallel boxplots below to compare the salary distribution for workers in a certain
industry across four different age groups: 20–29 years, 30–39 years, 40–49 years and
50–65 years.
50−65 years
40−49 years
30−39 years
20−29 years
40 50 60 70 80 90 100 110 120 130 140

Salary ($’000)
Solution
Here the numerical variable salary is the response variable and the categorical variable
age group is the explanatory variable.
1 Compare the medians: The median salary increased from $64 000 for 20−29 year-olds
to $72 000 for 50−65 year-olds.
2 Compare the IQRs: The IQR increased from around $12 000 for 20−29-year-olds to
around $20 000 for 50−65-year-olds.
3 Comparing the shapes: The shape of the distribution of salaries changes with the age
group, from symmetric to positively skewed.
4 Locate the outliers: There are no outliers in the 20-29 and 30-39 age group. Outliers
also begin to appear at $110 000 for the 40-49 age group, and at $119 000, $126 000
and $140 000 for the 50-65 age group.
5 Write the report comparing the distributions.

2C 2C Investigating the association between a numerical and a categorical variable 127
Report
In this industry there is an association between salary and age group. The median salaries
increase across the age groups, from $64 000 for 20−29 year-olds to $72 000 for 50−65
year-olds. The salaries also became more variable, with the IQR increasing from around
$12 000 for 20−29-year-olds to around $20 000 for 50−65-year-olds.The shape of the
distribution of salaries changes with age group, from symmetric for 20−29-year-olds, to
progressively more positively skewed as age increases. There are no outliers in the 20-29
and 30-39 age group. Outliers also begin to appear at $110 000 for the 40-49 age group,
and at $119 000, $126 000 and $140 000 for the 50-65 age group.
Skill-
sheet Exercise 2C
Example 9 1 Data was collected to compare the the number of days spent away from home (number
of days away) by 21 tourists from each of Japan and Australia (country of origin). The
data collected is displayed in the parallel dot plots below.
Japanese
0 5 10 15 20 25 30 35 40 45 50
Australian
0 5 10 15 20 25 30 35 40 45 50
Number of days away
a Identify each of the variables and classify each as categorical or numerical.

b Use the parallel dot plots to compare the distributions of number of days away for
the two different nationalities.
Example 10 2 The back-to-back stem plot shown compares the Females Males
distribution of the age patients (in years) admitted 9 0
5 0 1 3 6
to a small hospital during one week, and their 7 2 1 4 5 6 7
gender. 7 1 3 4
3 0 4 0 7
a Classify each variable as categorical or 0 5
numerical. 6
9 7
b Do the back-to-back plots support the 0 4 = 40 years 4 0 = 40 years
contention that the age of the patients is
associated with their gender? Write a brief
explanation that compares these distributions in
terms of centre and spread.

128 Chapter 2 Investigating associations between two variables 2C
3 The following back-to-back stem plot displays the distributions of the number of hours
per week spent online by a group of students, and their year level.
Number of hours per week online

Year 10 Year 11
1|2 = 21 hours 8 0 0 5 6 7 1|6 = 16 hours
3 1 1 2 4
8 7 6 5 1 5 5 6 7
4 4 3 1 1 0 0 2 0 1 2 3 3 4
8 5 2 7
a Classify each of the variables as categorical or numerical.
b Use the stem plots to compare these distributions in terms of centre and spread.
Draw an appropriate conclusion about the association between year level and the
number of hours students spend online each week.
Example 11 4 The parallel boxplots show the distribution of ages of 45 men and 38 women when first
married.
women
(n = 38)
men
(n = 45)
10 20 30 40 50
Age at marriage (years)
a Identify each of the variables and classify as categorical or numerical.

b Use the boxplots to compare these distributions in terms of shape, centre and spread
and draw an appropriate conclusion about the association between gender and the
age when first married.
5 The parallel boxplots female

show the distribution of (n = 21)
pulse rates of 21 females male
(n = 22)
and 22 males.
60 70 80 90
Pulse rate (beats per minute)
a Identify each of the variables and classify as categorical or numerical.
b Use the boxplots to compare these distributions, and draw an appropriate conclusion
about the association between gender and pulse rate.

2C 2C Investigating the association between a numerical and a categorical variable 129
Example 12 6 The parallel boxplots show the

high
distribution of the lifetime (in
hours) of three differently priced
batteries (low, medium, high). medium
a The two variables displayed
here are battery lifetime and low
battery price (low, medium,
high). Which is the numerical
10 20 30 40 50 60
and which is the categorical
Lifetime (hours)
variable?
b Do the parallel boxplots support the contention that battery lifetime depends on
price? Write a brief explanation.
Use the following information to answer Questions 7 and 8
The data in the following boxplots was collected to investigate the association between
smoking and the birth weight of babies.
Smoker
Non-smoker
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Weight of baby at birth (lbs)
Use the information in the boxplots to answer the following questions.
7 Which of the following statements is true:

A 75% of babies born to non-smokers weigh more than the lightest 50% of babies
born to smokers.
B 50% of babies born to non-smokers weigh more than the heaviest 25% of babies
born to smokers.
C 25% of the babies born to smokers weigh less than all the babies born to
non-smokers
D All of the babies born to non-smokers weigh more than the heaviest 75% of the
babies born to smokers
E The range of baby weights for smokers is less than the range of baby weights for
non smokers.

130 Chapter 2 Investigating associations between two variables 2C
8 The information in the boxplots supports the contention that there is an association
between smoking and weight of baby at birth because:
A The IQRs of birthweight for both groups are approximately the same.
B The median birthweight for smokers is more than the median birthweight for
non-smokers.
C The IQRs of birthweight for both groups are very different.
D The median birthweight for smokers is less than the median birthweight for
non-smokers.
E Both distributions are approximately symmetric.
2D Investigating associations between two

numerical variables
Learning intentions
I To be able to introduce the scatterplot for displaying data from two numerical
variables.
I To be able to construct a scatterplot using a CAS calculator.
In this section, we will learn to identify and describe associations between two numerical
variables. Suppose, for example, we wish to investigate the association between university
participation rate (the EV) and average hours worked (the RV) in nine countries. The
starting point for this investigation is again a graphical display of the data. Here our options
are to construct a scatterplot. The data for 9 countries are shown below.
Participation rate (%) 26 20 36 1 25 9 30 3 55

Hours worked 35 43 38 50 40 50 40 53 35
The first step in investigating an association between two numerical variables is to construct
a visual display of the data, which we call a scatterplot.
The scatterplot
A scatterplot is a plot which enables us to display bivariate data when both of the
variables are numerical.
In a scatterplot, each point represents a single case.
When constructing a scatterplot, it is conventional to use the vertical or y-axis for the
response variable (RV) and the horizontal or x-axis for the explanatory variable (EV).
The scatterplot below left shows the point for a country for which the university partici-
pation rate is 26% and average hours worked is 35, and the scatterplot below right is the
completed scatterplot when each of the remaining countries are plotted.

2D Investigating associations between two numerical variables 131
55 55
50 50
Hours worked
Hours worked
45 45
40 40
(26, 35)
35 35
30 30
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Participation rate (%) Participation rate (%)
CAS 1: How to construct a scatterplot using the TI-Nspire CAS

Construct a scatterplot for the set of test scores given below.
Treat test 1 as the explanatory (i.e. x) variable.
Test 1 10 18 13 6 8 5 12 15 15
Test 2 12 20 11 9 6 6 12 13 17
Steps
1 Start a new
document by
pressing / + N .
2 Select Add Lists &
Spreadsheet. Enter
the data into lists
named test1 and
test2.
3 Press / + I and select Add Data & Statistics.
4 a Click on Click to add variable on the x-axis
and select the explanatory variable test1.
b Click on Click to add variable on the y-axis and
select the response variable test2. A scatterplot is
displayed. The plot is scaled automatically.
CAS 1: How to construct a scatterplot using the ClassPad

Construct a scatterplot for the set of test scores given below.
Treat test 1 as the explanatory (i.e. x) variable.

132 Chapter 2 Investigating associations between two variables 2D
Test 1 10 18 13 6 8 5 12 15 15
Test 2 12 20 11 9 6 6 12 13 17
Steps
and enter the data into the columns
named test1 and test2.
2 Tap to open the Set
StatGraphs dialog box and
complete as given below.
Draw: select On.
Type: select Scatter ( ¤ ).
XList: select main\test1( ¤ ).
YList: select main\test2( ¤ ).
Freq: leave as 1.
Mark: leave as square.
3 Tap in the toolbar at the top of
the screen to plot the scatterplot in
the bottom half of the screen.
4 To obtain a full-screen plot, tap
from the icon panel.
Note: If you have more than one graph
on your screen, tap the data screen, select
StatGraph and turn off any unwanted
graphs.
Exercise 2D
1 The scatterplot opposite has been 850
constructed to investigate the association 825
between the airspeed (in km/h) of
800
Airspeed (km/h)
commercial aircraft and the number of

775
passenger seats.
Use the scatterplot to answer the following 750
questions. 725
a Which is the explanatory variable? 700
b What type of variable is airspeed?
c How many aircraft were investigated? 100 150 200 250 300 350 400 450
Number of seats
d What was the airspeed of the aircraft
ISBN 978-1-009-11041-9
that has 300 seats? © Peter Jones et al 2023 Cambridge University Press
2D 2D Investigating associations between two numerical variables 133
2
Minimum temperature (x) 17.7 19.8 23.3 22.4 22.0 22.0
Maximum temperature (y) 29.4 34.0 34.5 35.0 36.9 36.4
The table above shows the maximum and minimum temperatures (in ◦ C) during a hot
week in Melbourne.
a Enter the data into your calculator, naming the variables mintemp and maxtemp.
b Construct a scatterplot with minimum temperature as the EV.
3
Balls faced 29 16 19 62 13 40 16 9 28 26 6
Runs scored 27 8 21 47 3 15 13 2 15 10 2
The table above shows the number of runs scored and the number of balls faced
by batsmen in a 1-day international cricket match. Use a calculator to construct an
appropriate scatterplot.
4
Temperature (◦C) 0 10 50 75 100 150
Diameter (cm) 2.00 2.02 2.11 2.14 2.21 2.28
The table above shows the changing diameter of a metal ball as it is heated. Use a
calculator to construct an appropriate scatterplot, with temperature as the EV.
5
Number in theatre 87 102 118 123 135 137
Time (minutes) 0 5 10 15 20 25
The table above shows the number of people in a movie theatre at 5-minute intervals
after the advertisements started. Use a calculator to construct an appropriate scatterplot.

6 For which one of the following pairs of variables would it be appropriate to construct a
scatterplot?
A eye colour (blue, green, brown, other) and hair colour (black, brown, blonde, other)
B test score and sex (male, female)
C political party preference (Labor, Liberal, Other) and age in years
D age in years and blood pressure in mmHg
E height in cm and sex (male, female)

2E How to interpret a scatterplot

Learning intentions
I To be able to use a scatterplot to identify an association between two variables.
I From the the scatterplot, be able to classify an association according to:
B Direction, which may be positive or negative.
B Form, which may be linear or non-linear.
B Strength, which may be weak, moderate or strong.
What features do we look for in a scatterplot to help us identify

and describe any associations present? First we look to see if
there is a clear pattern in the scatterplot. In the scatterplot
opposite, there is no clear pattern in the points. The points are
randomly scattered across the plot, so we conclude that there is
no association.
For the three examples below, there is a clear (but different) pattern in each set of points, so
we conclude that there is an association in each case.
Having found a clear pattern, we need to be able to describe these associations clearly, as
they are obviously quite different. The three features we look for in the pattern of points are
direction, form and strength. Having found a clear pattern, there are several things we look
for in the pattern of points. These are:
direction and outliers (if any) form strength.
Direction and outliers

The scatterplot of height against age for a group of 205
footballers (shown opposite) is just a random scatter
195
of points. This suggests that there is no association
Height (cm)
between the variables height and age for this group 185
of footballers. However, there is a possible outlier for
height; a footballer who is 201 cm tall. 175
165
16 18 20 22 24 26 28 30 32
Age (years)

2E How to interpret a scatterplot 135
In contrast, there is a clear pattern in the scatterplot 100

of weight against height for these footballers (shown
90
Weight (kg)
opposite). The two variables are associated. If the
points in the scatterplot trend upwards as we go from 80
left to right we say there is a positive association
between the variables. In this example the positive 70
association means that taller players tend to be
60
heavier. In this scatterplot, there are no outliers. 170 180 190 200 210
Height (cm)
Likewise, the scatterplot of working hours against 50

university participation rates for 15 countries shows
a clear pattern. The two variables are associated. 45
Hours worked
If the points in the scatterplot trend downwards as
40
we go from left to right we say there is a negative
association between the variables. In this example 35
the negative association means that countries with
university participation rate tend to work fewer hours. 30
0 10 20 30 40 50 60
In this scatterplot, there are no outliers. Participation rate (%)
In general terms, we can classify the direction of an association as follows.
Direction of an association
Two variables have a positive association when the value of the response variable
tends to increase as the value of the explanatory variable increases.
Two variables have a negative association when the value of the response variable
tends to decrease as the value of the explanatory variable increases.
Two variables have no association when there is no consistent change in the value of
the response variable when the values of the explanatory variable increase.
Example 13 Direction of association
Classify each of the following scatterplots as exhibiting positive, negative or no

association. Where there is an association, describe the direction of the association
in terms of the variables in the scatterplot and what it means in terms of the variables
involved.
a 70 b 25 c 180
Reaction time (min)
60
Height daughter
20
Diameter (cm)
50 170
40 15
160
30 10
20 150
5
10
0 0 0
0 1 2 3 4 5 6 0 20 22 24 26 28 30 32 34 36 0 150 160 170 180
Dose (mg) Age (years) Height mother

a There is a clear pattern in the The direction of the association is negative.
scatterplot. The points in the scatterplot Reaction times tend to decrease as the drug
trend downwards from left to right. dose increases.
b There is no pattern in the scatterplot of There is no association between diameter
diameter against age. and age.
c There is a clear pattern in the The direction of the association is
scatterplot. The points in the scatterplot positive. Taller mothers tend to have taller
trend upwards from left to right. daughters.
Form
The next feature that interests us in an association is its general form. Do the points in a
scatterplot tend to follow a linear pattern or a curved pattern? If the scatterplot has a linear
form then we say that the association between the variables is linear.
For example, both of the scatterplots below can be described as having a linear form; that
is, the scatter in the points can be thought of as scattered around a straight line. (The dotted
lines have been added to the graphs to make it easier to see the linear form.)
30 55
25
Average working hours
50
Velocity (m/s)
20
45
15
40
10
5 35
0 30
0 1 2 3 4 5 0 10 20 30 40 50 60
Time (s) University participation (%)
By contrast, consider the scatterplot opposite, plotting 5
performance level against time spent practising a task. There 4

Performance level
is an association between performance level and time spent 3

practising, but it is clearly non-linear.
2
This scatterplot shows that while level of performance on
1
a task increases with practice, there comes a time when the
0
performance level will no longer improve substantially with 0 1 2 3 4 5 6 7 8 9 10
Time spent practising
extra practice.
While non-linear relationships exist (and we must always check for their presence by
examining the scatterplot), many of the relationships we meet in practice are linear or can
be made linear by transforming the data (a technique you will meet in Chapter 4). For this
reason we will restrict ourselves to the analysis of scatterplots with linear forms for now.
In general terms, we can describe the form of an association as follows.
2E How to interpret a scatterplot 137
Form
A scatterplot is said to have a linear form when the points tend to follow a straight line.
A scatterplot is said to have a non-linear form when the points tend to follow a curved
line.
Example 14 Form of an association
Classify the form of the association in each of scatterplot as linear or non-linear.

a 180 b 14
12
Weight loss (kg)

170 10
Height daughter
8
160 6
4
150 2
0
0
0 150 160 170 180 2 3 4 5 6 7
Height mother Number of weeks on a diet
a There is a clear pattern. The association is linear.
The points in the scatterplot can be imagined to
be scattered around a straight line.
b There is a clear pattern. The association is non-linear.
The points in the scatterplot can be imagined to
be scattered around a curved line rather than a
straight line.
Strength
The strength of an association is a measure of how much scatter there is in the scatterplot.
When there is a strong association between the variables, there is only a small amount of
scatter in the plot, and a pattern is clearly seen.
Strong positive association Strong positive association Strong negative association
As the amount of scatter in the plot increases, the pattern becomes less clear. This indicates
that the association is less strong. In the examples below, we might say that there is a
moderate association between the variables.
Moderate positive association Moderate positive association Moderate negative association
As the amount of scatter increases further, the pattern becomes even less clear. The
scatterplots below are examples of weak association between the variables.
Weak positive association Weak positive association Weak negative association
In general terms, we can describe the strength of an association as follows.
Strength
An association is classified as:
Strong if the points on the scatterplot tend to be tightly clustered about a trend line.
Moderate if the points on the scatterplot tend to be broadly clustered about a trend
line.
Weak if the points on the scatterplot tend to be loosely clustered about a trend line.
When no pattern can be seen we say that there is no association.
Example 15 Strength of an association
Classify the strength of the association in each of the following scatterplots.

25
a b
40
20
30
Y Y 15
20
10
10
5
5 10 15 20 25 0 5 10 15 20
X X
a The points are loosely clustered. The association is weak.
b The points are tightly clustered. The association is strong.

2E 2E How to interpret a scatterplot 139
Exercise 2E
Assessing the direction of an association from the variables
1 For each of the following pairs of variables:
a Indicate whether you expect an association to exist between the variables.
b If associated, say which variable you would expect to be the EV and which would
be the RV, and whether you would expect the variables to be positively or negatively
associated.
i intelligence and height ii level of education and salary level
iii salary and tax paid iv frustration and aggression
v population density and distance from the city centre
vi time using social media and time spent studying
Using a scatterplot to assess the direction, form and strength of an association

Example 13 2 For each of the following scatterplots, state whether the variables appear to be
associated. If the variables appear to be associated:
Example 14 a Describe the association in terms of its direction (positive/negative), form
(linear/non-linear) and strength (strong/moderate/weak).
Example 15 b Write a sentence describing the direction of the association in terms of the variables
in the scatterplot.
i 160 ii 110
Lung cancer mortality
140
Aptitude test score
120 100
100
80 90
60
0 0
0 80 100 120 140 0 8 10 12 14 16 18 20
Smoking rate Age (months)
iii 14 iv 20
Calf measurement
15
12
CO level
10
10
5
0 0
0 100 150 200 250 300 350 400 0 30 40 50
Age (years)

2F Strength of a linear relationship: the correlation

coefficient
Learning intentions
I To introduce Pearson’s correlation coefficient r as a measure of the strength of a linear
association between two variables.
I To be able to use technology to determine the value of Pearson’s correlation
coefficient r.
I To be able to classify the strength of a linear association as weak, moderate or strong
based on the value of Pearson’s correlation coefficient r.
The strength of a linear association is an indication of how closely the points in the
scatterplot fit a straight line. If the points in the scatterplot lie exactly on a straight line,
we say that there is a perfect linear association. If there is no fit at all we say there is no
association. In general, we have an imperfect fit, as seen in all of the scatterplots to date.
To measure the strength of a linear relationship, a statistician called Carl Pearson
developed a correlation coefficient, r, which has the following properties.
If there is no linear If there is a perfect If there is a perfect negative

association, r = 0. positive linear linear association, r = −1.
association, r = +1.
r=0 r = +1 r = −1
If there is a less than perfect linear association, then the correlation coefficient, r, has a value
between –1 and +1, or –1 < r < +1. The scatterplots below show approximate values of r
for linear associations of varying strengths.
r = −0.7 r = +0.5 r = −0.3 r = +0.9

2F Strength of a linear relationship: the correlation coefficient 141
Pearson’s correlation coefficient

The Pearson’s correlation coefficient, r:
measures the strength of a linear association, with larger values indicating stronger
relationships
has a value between –1 and +1
is positive if the direction of the linear association is positive.
is negative if the direction of the linear association is negative.
is close to zero if there is no association.
Calculating the correlation coefficient

Pearson’s correlation coefficient, r, gives a numerical measure of the degree to which the
points in the scatterplot tend to cluster around a straight line.
Formally, if we call the two variables we are working with x and y, and we have n
observations, then r is given by:
P
(x − x̄)(y − ȳ)
r=
(n − 1)s x sy
In this formula, x̄ and s x are the mean and standard deviation of the x-values, and ȳ and sy
are the mean and standard deviation of the y-values.
Calculating r using the formula (optional)

In practice, you can always use your calculator to determine the value of the correlation
coefficient. However, to understand what is involved when you use your calculator, it is best
that you know how to calculate the correlation coefficient from the formula first.
How to calculate the correlation coefficient using the formula

Use the formula to calculate the correlation coefficient, r, for the following data.
x 1 3 5 4 7 x̄ = 4, s x = 2.236
y 2 5 7 2 9 ȳ = 5, sy = 3.082
Give the values rounded to two decimal places.

Steps
1 Write down the values of the means, x̄ = 4 s x = 2.236
standard deviations and n.
ȳ = 5 sy = 3.082 n = 5

2 Set up a table like that shown x (x − x̄) y (y − ȳ) (x − x̄) × (y − ȳ)

opposite to calculate
P 1 −3 2 −3 9
(x − x̄)(y − ȳ).
3 −1 5 0 0
5 1 7 2 2
4 0 2 −3 0
7 3 9 4 12
Sum 0 0 23
∴ (x − x̄)(y − ȳ) = 23
P
P
3 Write down the formula for r. (x − x̄)(y − ȳ)
r=
Substitute the appropriate values and evaluate, (n − 1)s x sy
rounding the answer to two decimal places. 23
∴r=
(5 − 1) × 2.236 × 3.082
= 0.834... = 0.83 (2 d.p.)
CAS 2: How to calculate the correlation coefficient using the TI-Nspire

CAS
The following data show the per capita income (in $’000) and the per capita carbon
dioxide emissions (in tonnes) of 11 countries.
Determine the value of Pearson’s correlation coefficient rounded to two decimal places.
Income (000) 8.9 23.0 7.5 8.0 18.0 16.7 5.2 12.8 19.1 16.4 21.7
CO2 (tonnes) 7.5 12.0 6.0 1.8 7.7 5.7 3.8 5.7 11.0 9.7 9.9
Steps
1 Start a new document
by pressing / + N.
2 Select Add Lists &
Spreadsheet.
Enter the data into
lists named income
and co2.
3 Press / + I and select Add Calculator.
Using the correlation matrix command: type in corrmat(income, co2) and press
·.
Alternatively: Press k 1 C to access the Catalog, scroll down to corrMat( and press
·. Complete the command by typing in income, co2 and press ·.
The value of the correlation coefficient is r = 0.8342 . . . or 0.83 (2 d.p.)

2F Strength of a linear relationship: the correlation coefficient 143
CAS 2: How to calculate the correlation coefficient using the ClassPad
The following data show the per capita income (in $’000) and the per capita carbon
dioxide emissions (in tonnes) of 11 countries.
Determine the value of Pearson’s correlation coefficient rounded to two decimal places.
Income ($’000) 8.9 23.0 7.5 8.0 18.0 16.7 5.2 12.8 19.1 16.4 21.7
CO2 (tonnes) 7.5 12.0 6.0 1.8 7.7 5.7 3.8 5.7 11.0 9.7 9.9
Steps
.
2 Enter the data into the columns:
Income in List1
CO2 in List2.
3 Select Calc>Regression>Linear
Reg from the menu bar.
4 Press .
5 Tap OK to confirm your selections.
The value of the correlation
coefficient is
r = 0.818 . . . or 0.82 (to 2 d.p.).
Classifying the strength of a linear association

Pearson’s correlation coefficient, r, can be used to classify the strength of a linear
associations as follows:
0.75 ≤ r ≤1 strong positive association
0.5 ≤ r < 0.75 moderate positive association
0.25 ≤ r < 0.5 weak positive association
−0.25 < r < 0.25 no association
−0.5 < r ≤ −0.25 weak negative association
−0.75 < r ≤ −0.5 moderate negative association
−1 ≤ r ≤ −0.75 strong negative association

144 Chapter 2 Investigating associations between two variables 2F
Example 16 Classifying the strength of a linear association
Classify the strength of each of the following linear associations using the previous table:
a r = 0.35 b r = −0.507
c r = 0.992 d r = −0.159
a The value 0.35 is more than 0.25 and weak, positive
less than 0.5. That is, 0.25 ≤ r < 0.5
b The value –0.507 is more than –0.75 moderate, negative
and less than –0.5. That is, –0.75 < r ≤
–0.5
c The value 0.992 is more than 0.75 and strong, positive
less than 1. That is, 0.75 ≤ r ≤1
d The value –0.159 is more than –0.25 no association
and less than 0.25. That is, –0.25 < r <
0.25
Warning!
If you use the value of the correlation coefficient as a measure of the strength of an
association, you should ensure that:
1 the variables are numeric
2 the association is linear
3 there are no outliers in the data (the correlation coefficient can give a misleading
indication of the strength of the linear association if there are outliers present)
Skill-
sheet Exercise 2F
Basic ideas
1 The scatterplots of three sets of related variables are shown.
Scatterplot A Scatterplot B Scatterplot C

a For each scatterplot, describe the association in terms of strength, direction, form
and outliers (if any).
2F 2F Strength of a linear relationship: the correlation coefficient 145
b For which of these scatterplots would it be inappropriate to use the correlation

coefficient, r, to give a measure of the strength of the association between the
variables? Give reasons.
Calculating r using the formula (optional)

2 Use the formula to calculate the correlation coefficient, r, correct to two decimal
places.
x 2 3 6 3 6 x̄ = 4, s x = 1.871
y 1 6 5 4 9 ȳ = 5, sy = 2.915
Calculating r using a CAS calculator

3 a The table below shows the maximum and minimum temperatures during a
heat-wave. The maximum and minimum temperature each day are linearly
associated. Use your calculator to show that r = 0.818, correct to three decimal
places.
Day Sunday Monday Tuesday Wednesday Thursday Friday

Maximum (◦C) 29.4 34.0 34.5 35.0 36.9 36.4
Minimum (◦C) 17.7 19.8 23.3 22.4 22.0 22.0
b This table shows the number of runs scored and balls faced by batsmen in a cricket
match. Runs scored and balls faced are linearly associated. Use your calculator to
show that r = 0.8782, correct to four decimal places.
Batsman 1 2 3 4 5 6 7 8 9 10 11
Runs scored 27 8 21 47 3 15 13 2 15 10 2
Balls faced 29 16 19 62 13 40 16 9 28 26 6
c This table shows the hours worked and university participation rate (%) in six
countries. Hours worked and university participation rate are linearly associated.
Use your calculator to show that r = −0.6727, correct to four decimal places.
Country Australia Britain Canada France Sweden US

Hours worked 35.0 43.0 38.2 39.8 35.6 34.8
Participation rate (%) 26 20 36 25 37 55
Classifying the strength of the association based on the value of r

Example 16 4 Use the guidelines on page 143 to classify the strength of the linear associations for
each of the linear associations in Question 3.

2G The coefficient of determination

Learning intentions
I To be able to calculate the value of the coefficient of determination.
I To be able to use the coefficient of determination to assess the strength of the
association in terms of the explained variation.
If two variables are associated, it is possible to estimate the value of one variable from
that of the other. For example, people’s weights and heights are associated. Thus, given a
person’s height, we can roughly predict their weight. The degree to which we can make such
predictions depends on the value of r. If there is a perfect linear association (r = 1) between
two variables, we can make an exact prediction.
For example, when you buy cheese by the gram there is an exact association between the
weight of the cheese and the amount you pay (r = 1). At the other end of the scale, there is
no association between an adult’s height and their IQ (r ≈ 0). So knowing an adult’s height
will not enable you to predict their IQ any better than guessing.
The coefficient of determination

The degree to which one variable can be predicted from another linearly related variable
is given by a statistic called the coefficient of determination.
The coefficient of determination is calculated by squaring the correlation coefficient:

coefficient of determination = r2
Example 17 Calculating the coefficient of determination
If the correlation between weight and height is r = 0.8, find the value of the coefficient of
determination. Express your answer as a percentage.
Solution
The coefficient of determination = r2 = 0.82 = 0.64 = 64%
Note: We have converted the coefficient of determination into a percentage (64%) as this is the most useful
form when we come to interpreting the coefficient of determination.
We now know how to calculate the coefficient of determination, but what does it tell us?
Interpreting the coefficient of determination

The coefficient of determination (as a percentage) tells us the variation in the response
variable that is explained by the variation in the explanatory variable.

2G The coefficient of determination 147
Example 18 Interpreting the coefficient of determination
In the previous example we found the coefficient of determination between height and
weight to be 0.64 (or 64%). Interpret this value in terms of the variables weight and
height.
Solution
The coefficient of determination tells us that 64% of the variation in people’s weight is
explained by the variation in their height.
What do we mean by ‘explained’?

If we take a group of people, their weights and heights will vary. One explanation is that
taller people tend to be heavier and shorter people tend to be lighter. The coefficient of
determination tells us that 64% of the variation in people’s weights can be explained by the
variation in their heights. The rest of the variation (36%) in their weights will be explained
by other factors, such as diet, lifestyle, build. We could say that 36% of the variation in
weight is NOT explained by the variation in height.
Example 19 Calculating and interpreting the coefficient of determination
The level of carbon monoxide (CO) in the air measured at the roadside, and the traffic
volume at the same location are linearly related, with r =+0.985. Determine the value
of the coefficient of determination, write it in percentage terms and interpret. In this
relationship, traffic volume is the explanatory variable.
Solution
The coefficient of determination is:
r2 = (0.985)2 = 0.9702
Written as a percentage: 0.9702 × 100 = 97.0% rounded to one decimal place.
Therefore, 97.0% of the variation in carbon monoxide levels in the air can be explained
by the variation in traffic volume.
Clearly, traffic volume is a very good predictor of carbon monoxide levels in the air. Thus,
knowing the traffic volume enables us to predict carbon monoxide levels with a high degree
of accuracy. This is not the case with the next example.
Example 20 Calculating and interpreting the coefficient of determination
Scores on tests of verbal and mathematical ability are linearly related with correlation
coefficient r =+0.275. Determine the value of the coefficient of determination, write it
in percentage terms, and interpret. In this relationship, verbal ability is the explanatory
variable.

148 Chapter 2 Investigating associations between two variables 2G
Solution
The coefficient of determination is:
r2 = (0.275)2 = 0.0756
Written as a percentage: 0.0756 × 100 = 7.6% rounded to one decimal place.
Therefore, only 7.6% of the variation observed in scores on the mathematical ability test
can be explained by the variation in scores obtained on the verbal ability test.
Clearly, scores on the verbal ability test are not good predictors of the scores on the
mathematical ability test; 92.4% of the variation in mathematical ability is explained by
other factors.
Given the value of the coefficient of determination we can reverse the calculation and find
the value of the correlation coefficient. However, since the square root of a number can be
positive or negative, we need more information to be able to do this correctly, such as a
scatterplot.
Example 21 Calculating the correlation coefficient from the coefficient of

determination
For the relationship described by this scatterplot, the

coefficient of determination = 0.5210.
Determine the value of the correlation coefficient, r, rounded
to four decimal places.
1 Since we know the value of the coefficient of r2 = 0.5210
√
determination (= r2 ), we need to find the square ∴ r = ± 0.5210 = ±0.7218
root of this value to find r.
2 There are two solutions, one positive and the Scatterplot indicates a negative
other negative. Use the scatterplot to decide association.
which applies.
3 Write down your answer. ∴ r = −0.7218
Exercise 2G
Calculating the coefficient of determination from r

Example 17 1 For each of the following values of r, calculate the value of the coefficient of
determination and convert to a percentage (correct to one decimal place).
a r = 0.675 b r = 0.345 c r = −0.567 d r = −0.673 e r = 0.124

2G 2G The coefficient of determination 149
Calculating and interpreting the coefficient of determination

Example 18 2 For each of the following, determine the value of the coefficient of determination, write
Example 19 it in percentage terms, and interpret.
a Scores on a hearing test and age (EV) are linearly related, with r = −0.611.
b Mortality rate and smoking rate (EV) are linearly related, with r = 0.716.
c Life expectancy and birth rate (EV) are linearly related, with r = −0.807.
d Daily maximum (RV) and minimum temperatures are linearly related, with
r = 0.818.
e Runs scored (RV) and balls faced by a batsman are linearly related, with r = 0.8782.
Calculating r from the coefficient of determination given a scatterplot

Example 21 3 a For the relationship described by the scatterplot shown, the
coefficient of determination, r2 = 0.8215.
Determine the value of the correlation coefficient, r (rounded to
three decimal places).
b For the relationship described by the scatterplot shown, the

coefficient of determination r2 = 0.1243.
Determine the value of the correlation coefficient, r (rounded to
three decimal places).
Use the following information to answer Questions 4 to 6
The association between the number 45

Time to complete course
of training sessions attended by

40
participants before undertaking an
(mins)
an obstacle course, and the time in 35
minutes it took them to complete

30
the course, is described by the
scatterplot shown. The coefficient of 25
1 2 3 4 5 6 7 8
determination is 0.3969. Number of training sessions
4 The value of the correlation coefficient, r (rounded to two decimal places) is closest to.
A 0.16 B 0.40 C 0.63 D –0.40 E –0.63

150 Chapter 2 Investigating associations between two variables 2G
5 The percentage of variation in time explained by the variation in the number of training
sessions is closest to:
A 39.7% B 63.0% C 15.8% D 37.0% E 60.3%
6 The percentage of variation in time NOT explained by the variation in the number of
training sessions is closest to:
A 39.7% B 63.0% C 15.8% D 37.0% E 60.3%
7 Suppose that in a certain industry the correlation between years spent studying and
income for employees is 0.73, and the correlation between age and income is 0.45.
Given this information, which one of the following statements is true?
A Older employees tend to have spent more years studying.
B The correlation between age and years spent studying is 0.32.
C Age explains a higher percentage of the variation in income than years spent
studying.
D Years spent studying explains a higher percentage of the variation in income than
age.
E Together age and years spent studying explain 100% of the variation in income.
8 Which of the following statements could be true?

A The correlation coefficient between height (in centimetres) and weight (1 = light, 2
= medium, 3 = heavy) was found to be 0.68.
B The correlation coefficient between height (in centimetres) and head circumference
(in centimetres) was found to be 1.45.
C The correlation coefficient between blood pressure (in mmHg) and weight (in kg)
was found to be -0.3, and the coefficient of determination was found to be r2 =
-0.09.
D The correlation coefficient between age (in years) and salary (in $000’s) was found
to be 0.68.
E The correlation coefficient between height (in centimetres) and head circumference
(in centimetres) was found to be 0.49, and the coefficient of determination was
found to be 70%.

2H Correlation and causality 151
2H Correlation and causality

Learning intentions
I To be able to define and differentiate the concepts of association and causation.
Recently there has been interest in the strong association between the number of Nobel
prizes a country has won and the number of IKEA stores in that country (r = 0.82). This
strong association is evident in the scatterplot below. Here country flags are used to
represent the data points.
35
30
25
10 Million Population
Nobel Laureates per
20
15
10
5 r = 0.82
0 5 10 15 20
IKEA Stores per 10 Million Population
Does this mean that one way to increase the number of Australian Nobel prize winners is to
build more IKEA stores?
Almost certainly not, but this association highlights the problem of assuming that a strong
correlation between two variables indicates the association between them is causal.
Correlation does not imply causality

A correlation tells you about the strength of the association between the variables, but no
more. It tells you nothing about the source or cause of the association.
Video
To help you with this concept, you should watch the video ‘The Question of Causation’,
which can be accessed through the link below. It is well worth 15 minutes of your time.
http://cambridge.edu.au/redirect/?id=6103

Establishing causality
To establish causality, you need to conduct an experiment. In an experiment, the value of
the explanatory variable is deliberately manipulated, while all other possible explanatory
variables are kept constant or controlled. A simplified version of an experiment is displayed
below.
Treatment 1:
Group 1 Lesson on time
series
Randomly allocate
a group of students Give test on
to two groups time series
Treatment 2:
Group 2 Lesson on
Shakespeare
In this experiment, a class of students is randomly allocated into two groups. Random
allocation ensures that both groups are as similar as possible.
Next, group 1 is given a lesson on time series (treatment 1), while group 2 is given a lesson
on Shakespeare (treatment 2). Both lessons are given under the same classroom conditions.
When both groups are given a test on time series the next day, group 1 does better than
group 2.
We then conclude that this was because the students in group 1 were given a lesson on time
series.
Is this conclusion justified?

In this experiment, the students’ test score is the response variable and the type of lesson
they were given is the explanatory variable. We randomly allocated the students to each
group while ensuring that all other possible explanatory variables were controlled by
giving the lessons under the same classroom conditions. In these circumstances, the
observed difference in the response variable (test score) can reasonably be attributed to the
explanatory variable (lesson type).
Unfortunately, it is extremely difficult to conduct properly controlled experiments,
particularly when the people involved are going about their everyday lives.
When data are collected through observation rather than experimentation, we must accept
that strong association between two variables is insufficient evidence by itself to conclude
that an observed change in the response variable has been caused by an observed change
in the explanatory variable. It may be, but unless all of the relevant variables are under our
control, there will always be alternative non-causal explanations to which we can appeal. We
will now consider the various ways this might occur.

2H Correlation and causality 153
Possible non-causal explanations for an association

Common response
Consider the following. There is a strong positive association between the number of people
using sunscreen and the number of people fainting. Does this mean that applying sunscreen
causes people to faint?
Almost certainly not. On hot and sunny days, more people apply sunscreen and more
people faint due to heat exhaustion. The two variables are associated because they are both
strongly associated with a common third variable, temperature. This phenomenon is called a
common response. See the diagram below.
causes Sunscreen
Temperature observed
association
causes Fainting
Unfortunately, being able to attribute an association to a single third variable is the exception
rather than the rule. More often than not, the situation is more complex.
Confounding variables
Statistics show that crime rates and unemployment rates in a city are strongly correlated. Can
you then conclude that a decrease in unemployment will lead to a decrease in crime rates?
It might, but other possible causal explanations could be found. For example, these data
were collected during an economic downturn. Perhaps the state of the economy caused the
problem. See the diagram below.
Unemployment
causes ? observed
Economy association
causes ?
Crime
In this situation, we have at least two possible causal explanations for the observed
association, but we have no way of disentangling their separate effects. When this happens,
the effects of the two possible explanatory variables are said to be confounded, because we
have no way of knowing which is the actual cause of the association.

154 Chapter 2 Investigating associations between two variables 2H
Coincidence
It turns out that there is a strong correlation (r = 0.99) between the consumption of
margarine and the divorce rate in the American state of Maine. Can we conclude that eating
margarine causes people in Maine to divorce?
A better explanation is that this association is purely coincidental.
Occasionally, it is almost impossible to identify any feasible confounding variables to
explain a particular association. In these cases we often conclude that the association is
‘spurious’ and it has happened just happened by chance. We call this coincidence.
Conclusion
However suggestive a strong association may be, this alone does not provide sufficient
evidence for you to conclude that two variables are causally related. Unless the association
is totally spurious and devoid of meaning, it will always be possible to find at least one
variable ‘lurking’ in the background that could explain the association.
Association (correlation) and causation

By itself, an observed association between two variables is never enough to justify the
conclusion that two variables are causally related, no matter how obvious the causal
explanation may appear to be.
Exercise 2H
1 A study of primary school children aged 5 to 11 years finds a strong positive

correlation between height and score on a test of mathematics ability. Does this mean
that taller people are better at mathematics? What common cause might counter this
conclusion?
2 There is a clear positive correlation between the number of churches in a town and
the amount of alcohol consumed by its inhabitants. Does this mean that religion is
encouraging people to drink? What common cause might counter this conclusion?
3 There is a strong positive correlation between the amount of ice-cream consumed and
the number of drownings each day. Does this mean that eating ice-cream at the beach
is dangerous? What common cause might explain this association?
4 The number of days a patient stays in hospital is positively correlated with the number
of beds in the hospital. Can it be said that bigger hospitals encourage patients to stay
longer than necessary just to keep their beds occupied? What common cause might
counter this conclusion?
5 Suppose we found a high correlation between smoking rates and heart disease across
a group of countries. Can we conclude that smoking causes heart disease? What
confounding variable(s) could equally explain this correlation?

2H 2I Which graph? 155
6 There is a strong correlation between cheese consumption and the number of people
who died after becoming tangled in their bed sheets. What do you think is the most
likely explanation for this correlation?
7 There is a strong positive correlation between the number of fire trucks attending a
house fire and the amount of damage caused by the fire. Is the amount of damage
in a house fire caused by the fire trucks? What common cause might explain this
association?

8 There is a positive correlation between the Gross Domestic Product (GDP), a
measure of a country’s wealth, and the country’s carbon dioxide emissions. From this
information it can be concluded that:
A increasing a country’s GDP will increase the carbon dioxide emissions of that
country.
B decreasing a country’s GDP will increase the carbon dioxide emissions of that
country.
C increasing a country’s carbon dioxide emissions will increase the GDP of that
country.
D countries with higher GDP also tend to have lower carbon dioxide emissions.
E countries with higher GDP also tend to have higher carbon dioxide emissions.
2I Which graph?
When investigating associations your first decision is choosing an appropriate graph to
display and understand the data you have been given. This decision depends on the type
of variables involved – that is, whether they are both categorical, one categorical and one
numerical, or both numerical.
The following guidelines might help you make your decision. They are guidelines only,
because in some instances there may be more than one suitable graph.
Type of variables
Response variable Explanatory variable Graph
Categorical Categorical Segmented bar chart.
Numerical Categorical Parallel boxplots, parallel dot plots
Numerical Categorical Back-to-back stem plot, parallel dot
(two categories only) plots or parallel boxplots
Numerical Numerical Scatterplot

156 Chapter 2 Investigating associations between two variables 2I
Exercise 2I
1 Which graphical display (parallel boxplots, parallel dot plots, back-to-back stem plot, a
segmented bar chart or a scatterplot) would be appropriate to display the relationships
between the following? There may be more than one appropriate graph.
a vegetarian (yes, no) and sex (male, female)
b mark obtained on a statistics test and time spent studying (in hours)
c number of hours spent at the beach each year and state of residence
d number of CDs purchased per year and income (in dollars)
e runs scored in a cricket game and number of ‘overs’ faced
f attitude to compulsory sport in school (agree, disagree, no opinion) and school type
(government, independent)
g income level (high, medium, low) and place of residence (urban, rural)
h number of cigarettes smoked per day and sex (male, female)
2 A back-to-back stem plot would be an appropriate graphical tool to investigate the

association between a car’s speed, in kilometres per hour, and the
A driver’s age, in years
B car’s colour (white, red, grey, other)
C car’s fuel consumption, in kilometres per litre
D average distance travelled, in kilometres
E driver’s type of licence (probationary licence, full licence)

3 The relationship between height (in centimetres) and weight (1 = light, 2 = medium, 3
= heavy) is best displayed using:
A a histogram
B segmented bar charts
C a scatterplot
D parallel boxplots
E a percentaged two-way frequency table

Chapter 2 review 157
Review
Bivariate data Bivariate data are generated when information about two variables is
Assign- recorded for each subject.
ment
Explanatory When investigating associations (relationships) between two variables,

and response the explanatory variable (EV) is the variable we expect to explain or
variables
predict the value of the response variable (RV).
Two-way Two-way frequency tables are used as the starting point for
frequency tables investigating the association between two categorical variables.
Segmented bar A segmented bar chart can be used to graphically display the
charts information contained in a two-way frequency table. It is a useful tool
for identifying relationships between two categorical variables.
Identifying Associations between two categorical variables are described by

associations comparing appropriate percentages in a percentaged two-way
between two
categorical
frequency table or percentaged segmented bar chart.
variables
Identifying Associations between a numerical and a categorical variable are

associations identified using parallel dot plots, parallel boxplots or a back-to-back
between a
numerical and
stem plot. Associations between a numerical and a categorical variable
a categorical are described by comparing the shape, centre and spread for the
variable distributions.
Scatterplots A scatterplot is used to help identify and describe an association

between two numerical variables. In a scatterplot, the response
variable (RV) is plotted on the vertical axis and the explanatory
variable (EV) is plotted on the horizontal axis.
Identifying Associations between two numerical variables are identified using a

associations scatterplot. Associations are classified according to:
between two
numerical Direction, which may be positive or negative.
variables Form, which may be linear or non-linear.
Strength, which may be weak, moderate or strong.
Correlation The correlation coefficient, r, gives a measure of the strength of a

coefficient, r linear association.

Review 158 Chapter 2 Investigating associations between two variables
The coefficient of Coefficient of determination = r2

determination The coefficient of determination gives the percentage of variation
in the response variable that can be explained by the variation in the
explanatory variable.
Correlation and A correlation between two variables does not automatically imply that
causation the association is causal. Alternative non-causal explanations for the
association include a common response to a common third variable, a
confounded variable or simply coincidence.
Skills checklist
list
2A 1 I can identify categorical and numerical variables in bivariate data.
2A 2 I can identify explanatory and response variables.
2B 3 I can construct a two-way frequency table.
2B 4 I can percentage a two-way frequency table.
2B 5 I can describe an association from a percentaged two-way frequency table.
2B 6 I can construct a segmented bar chart from a percentaged two-way frequency

table.
2B 7 I can describe an association from a percentaged segmented bar chart.
2C 8 I can use parallel dot plots to identify and describe the association between a
numerical variable and a categorical variable.

Review
2C 9 I can use back-to-back stem plots to display and describe the association
between a numerical variable and a categorical variable.
2C 10 I can use parallel boxplots to display and describe the association between a
numerical variable and a categorical variable.
2D 11 I can construct a scatterplot using a CAS calculator.
See CAS 1, and Exercise 2D Question 2
2E 12 I can classify the direction, form and strength of an association from a

scatterplot.
See Example 13, Example 14, Example 15, and Exercise 2E Question 2
2F 13 I can use technology to determine the value of the correlation coefficient r.
See CAS 2, and Exercise 2F Question 3
2F 14 I can classify the strength of a linear association as weak, moderate or strong

based on the value of the correlation coefficient r.
2G 15 I can calculate the value of the coefficient of determination.
2G 16 I can use the coefficient of determination to assess the strength of the

association in terms of the the explained variation.
2H 17 I understand that correlation does not imply causation.
See Exercise 2H Question 1

The information in the following frequency table relates to Questions 1 to 4.
Gender
Plays sport Male Female
Yes 68 79
No 34
Total 102 175
1 The variables plays sport and gender are:

A both categorical variables
B a categorical and a numerical variable, respectively
C a numerical and a categorical variable, respectively
D both numerical variables
E neither numerical nor categorical variables
2 The number of females who do not play sport is:

A 21 B 45 C 79 D 96 E 175
3 The percentage of males who do not play sport is:

A 19.4% B 33.3% C 34.0% D 66.7% E 68.0%
4 The variables plays sport and gender appear to be associated because:

A more females play sport than males
B fewer males play sport than females
C a higher percentage of females play sport compared to males
D a higher percentage of males play sport compared to females
E both males and females play a lot of sport
Questions 5 to 7 relate to the following information
Students in Year 10 and Year 12 in a certain school were asked whether they read for leisure
(read). Their responses are summarised in the percentaged segmented bar chart shown.

Review
100%
90%
80%
70%
Read for leisure

60% No
50% Yes
40%
30%
20%
10%
0%
Year 10 Year 12
Year level
5 The percentage of Year 12 students who do not read for leisure is closest to:
A 10% B 25% C 30% D 75% E 90%
6 The results could be summarised in a two-way frequency table. Which of the following
frequency tales could match the percentaged segmented bar chart?
Year Level Year Level
Read Year 10 Year 12 Read Year 10 Year 12
A B
Yes 31 45 Yes 45 11
No 47 66 No 135 99
Year Level Year Level
Read Year 10 Year 12 Read Year 10 Year 12
C D
Yes 75 90 Yes 75 25
No 25 10 No 90 10
Year Level
Read Year 10 Year 12
E
Yes 40 8
No 38 5
7 The variables read and year level appear to be associated because:

A very few students in either year level read for leisure
B 75% of Year 10 students do not read for leisure
C only 10% of Year 12 students read for leisure
D 25% of Year 10 students read for leisure, while only 10% of Year 12 students read
for leisure
E a higher percentage of Year 12 students read for leisure than Year 10 students

8 The stem plots displays the time taken (in minutes) for two groups of 12 people to
solve a complex puzzle. Before commencing the puzzle the people were divided
into two groups and assigned a different acitvity. Group A were asked to exercise
vigorously for 10 minutes, while Group B were asked to meditate for 10 minutes.
time
Group A Group B
1|2 = 21 minutes 9 0 5 6 7 1|6 = 16 minutes
3 1 1 2 4
8 6 5 1 5 5 6
4 4 3 1 2 0 1 2 3
8 5 2
The information in the stem plots supports the contention that there is an association
between time and activity because:
A The median time for Group A is more than the median time for Group B.
B The range of times for both groups are approximately equal.
C The median time for Group B is more than the median time for Group A.
D Both distributions are approximately symmetric.
E Both distributions are negatively skewed.
The information in the following parallel boxplots relates to Questions 9 and 10.
The parallel boxplots shown display the brand A

distribution of battery life (in hours) for two
brands of batteries (brand A and brand B). brand B
10 20 30 40 50 60
Hours
9 The variables battery life and brand are:
B a categorical and a numerical variable respectively
C a numerical and a categorical variable respectively
E neither a numerical nor a categorical variable

Review
10 Which of the following statements (there may be more than one) support the contention
that battery life and brand are related?
I the median battery life for brand A is clearly higher than for brand B
II battery lives for brand B are more variable than brand A
III the distribution of battery lives for brand A is symmetrical with outliers but
positively skewed for brand B
A I only B II only C III only D I and II only E I, II and III
11 The association between weight at age 21 (in kg) and weight at birth (in kg) is to be
investigated. The variables weight at age 21 and weight at birth are:
B a categorical and a numerical variable respectively
C a numerical and a categorical variable respectively
E neither numerical nor categorical variables
12 The scatterplot shows the weights at age 65

21 and weight at birth of 12 women. The Weight at 21 years (kg)
60
association is best described as a:
55
A weak positive linear
50
B weak negative linear
45
C moderate positive non-linear
40
D strong positive non-linear
1.5 2 2.5 3 3.5 4 4.5
E moderate positive linear Birth weight (kg)
13 The association between weight at age 21 and weight at birth for a group of males is
found to be positive and linear, with a correlation coefficient of r = 0.58. For males, the
percentage of variation in weight at age 21 explained by the variation in weight at birth
is closest to:
A 0.34% B 24% C 34% D 58% E 76%
14 The variables response time to a drug and drug dosage are linearly associated, with
r = −0.9. From this information, we can conclude that:
A response times are –0.9 times the drug dosage
B response times decrease with decreased drug dosage
C response times decrease with increased drug dosage
D response times increase with increased drug dosage
E response times are 81% of the drug dosage

15 The birth weight and weight at age 21 of eight women are given in the table below.
Birth weight (kg) 1.9 2.4 2.6 2.7 2.9 3.2 3.4 3.6
Weight at 21 (kg) 47.6 53.1 52.2 56.2 57.6 59.9 55.3 56.7
The value of the correlation coefficient is closest to:
A 0.536 B 0.6182 C 0.7863 D 0.8232 E 0.8954
16 The value of a correlation coefficient is r = –0.7685. The value of the corresponding

coefficient of determination is closest to:
A −0.77 B −0.59 C 0.23 D 0.59 E 0.77
Use the following information to answer Questions 17 and 18.
The correlation coefficient between heart weight and body weight in a group of mice is
r = 0.765.
17 Using body weight as the EV, we can conclude that:

A 58.5% of the variation in heart weight is explained by the variation in body weights
B 76.5% of the variation in heart weight is explained by the variation in body weights
C heart weight is 58.5% of body weight
D heart weight is 76.5% of body weight
E 58.5% of the mice had heavy hearts
18 Given that heart weight and body weight of mice are strongly correlated (r = 0.765),
we can conclude that:
A increasing the body weights of mice will decrease their heart weights
B increasing the body weights of mice will increase their heart weights
C increasing the body weights of mice will not change their heart weights
D heavier mice tend to have lighter hearts
E heavier mice tend to have heavier hearts
19 We wish to investigate the association between the variables weight (in kg) of young
children and level of nutrition (poor, adequate, good). The most appropriate graphical
display would be:
A a histogram B parallel boxplots C a segmented bar chart
D a scatterplot E a back-to-back stem plot
20 We wish to investigate the association between the variables weight (underweight,

normal, overweight) of young children and level of nutrition (poor, adequate, good).
The most appropriate graphical display would be:
A a histogram B parallel boxplots C a segmented bar chart
D a scatterplot E a back-to-back stem plot

Review
21 There is a strong linear positive correlation (r = 0.85) between the amount of garbage
recycled and salary level.
From this information, we can conclude that:
A the amount of garbage recycled can be increased by increasing people’s salaries
B the amount of garbage recycled can be increased by decreasing people’s salaries
C increasing the amount of garbage you recycle will increase your salary
D people on high salaries tend to recycle less garbage
E people on high salaries tend to recycle more garbage
22 There is a strong linear positive correlation (r = 0.95) between the marriage rate in
Kentucky and the number of people who drown falling out of a fishing boat.
From this information, the most likely conclusion we can draw from this correlation is:
A reducing the number of marriages in Kentucky will decrease the number of people
who drown falling out of a fishing boat
B increasing the number of marriages in Kentucky will increase the number of people
who drown falling out of a fishing boat
C this correlation is just coincidence, and changing the marriage rate will not affect
the number of people drowning in Kentucky in any way
D only married people in Kentucky drown falling out of a fishing boat
E stopping people from going fishing will reduce the marriage rate in Kentucky

1 One thousand drivers who had an accident during the past year were classified
according to age and the number of accidents.
Number of accidents Age < 30 Age ≥ 30

At most one accident 130 170
More than one accident 470 230
Total 600 400
a What are the variables shown in the table? Are they categorical or numerical?
b Determine the response and explanatory variables.
c How many drivers under the age of 30 had more than one accident?
d Convert the table values to percentages by calculating the column percentages.
e Use these percentages to comment on the statement: ‘Of drivers who had an
accident in the past year, younger drivers (age < 30) are more likely than older
drivers (age ≥ 30) to have had more than one accident.’

2 In order to improve their ability in French conversation a group of 50 students who

were studying French participated in a 12 weeks intensive conversation course. The
students were given a test to assess their conversation ability at the start of the course,
midway through the course, and at the end of the course. Their results in each of the
three tests are shown in the following boxplots.
100
90
Conversation test score 80
70
60
50
40
30
20
10
0 weeks 6 weeks 12 weeks
Completed weeks of course
a The two variables are Completed weeks of course and Conversation test score.
Which is numerical and which is categorical?
b Use the boxplots to compare these distributions, and draw an appropriate conclusion
about the association between the number of weeks of the course completed and the
score in the conversation test. Quote appropriate statistics in your response.
3 The data below give the hourly pay rates (in dollars per hour) of 10 production-line
workers along with their years of experience on initial appointment.
Rate ($ /h) 22.57 25.78 28.84 27.37 27.23 24.64 28.95 33.35 29.68 33.99
Experience (years) 1 1 2 2 3 4 5 6 8 12
a Determine which variable is the explanatory variable and which is the response
variable.
b Use a CAS calculator to construct a scatterplot of the data,
c Comment on direction, outliers, form and strength of any association revealed.
d Determine the value of the correlation coefficient (r) rounded to three decimal
places.
e Determine the value of the coefficient of determination (r2 ), giving your answer as a
percentage rounded to one decimal place, and interpret.

Review
4 In a study of the effects of meditation on the quality of sleep a sample of 500 people
were asked to rate the quality of their sleep as ‘good’, ‘OK’, or ‘poor’ before and
after participating in the course. Their responses are shown in the segmented bar chart
below.
100%
90%
80%
70%
Quality of sleep
60%
50% Poor
40% Ok
30% Good
20%
10%
0%
Before After
Meditation course
a What percentage of people rated the quality of their sleep as ‘good’ before they
participated in the course?
b Does the segmented bar chart support the contention that for these people their
quality of sleep is associated with participation in the course? Justify your answer
by quoting appropriate percentages.

Chapter
3
Investigating and
modelling linear
associations
Chapter questions
I What is linear regression?
I What is a residual?
I What is a least squares line of best fit?
I How do you find the equation of the least squares line using summary
statistics?
I How do you find the equation of the least squares line using technology?
I How do you interpret the intercept and slope of the least squares line?
I How do you use the equation of the least squares line to make predictions?
I How do you use the coefficient of determination in a regression analysis?
I What is a residual plot and how is it used?
I How do you report a regression analysis?
Once we identify a linear association between two numerical variables, we can

fit a linear model to the data and find its equation. This equation gives us a better
understanding of the nature of the relationship between the two variables, and we
can also use the linear model to make predictions based on this understanding of the
relationship.

3A Fitting a least squares regression line to numerical data 169
3A Fitting a least squares regression line to

numerical data
Learning intentions
I To be able to define linear regression.
I To be able to define a residual.
I To introduce the least squares line of best fit.
I To be able to find the equation of the least squares line using summary statistics.
I To be able to find the equation of the least squares line using technology.
The process of modelling an association with a straight line is known as linear regression
and the resulting line is often called the regression line.
The equation of a line relating two variables x and y is of the form
y = a + bx
where a and b are constants. When the equation is written in this form:
a represents the coordinate of the point where the line crosses the y-axis (the y-intercept)
b represents the slope of the line.
In order to summarise any particular (x, y) data set, numerical values for a and b are needed
that will ensure the line passes close to the data. There are several ways in which the values
of a and b can be found.
The easiest way to fit a line to bivariate data is to construct a scatterplot and draw the line
‘by eye’. We do this by placing a ruler on the scatterplot so that it seems to follow the
general trend of the data. You can then use the ruler to draw a straight line. Unfortunately,
unless the points are very tightly clustered around a straight line, the results you get by using
this method will differ a lot from person to person.
The more mathematical approach to fitting a straight line to data is to use the least squares
method. This method assumes that the variables are linearly related, and works best when
there are no clear outliers in the data.
Some terminology
To explain the least squares method, we need to define several terms.

170 Chapter 3 Investigating and modelling linear associations
The scatterplot shows five data points, (x1 , y1 ),

y
(x2 , y2 ), ( x3 , y3 ), (x4 , y4 ) and (x5 , y5 ).
A regression line (not necessarily the least (x5, y5)
d5
squares line) has also been drawn on the regression line
(x3, y3)
scatterplot.
(x2, y2) d3 d4
The vertical distances d1 , d2 , d3 , d4 and d5 of (x4, y4)
d2
each of the data points from the regression line
are also shown. d1
(x1, y1)
These vertical distances, d, are known as
x
residuals.
Residuals can be positive, negative or zero:
Data points above the fitted regression line have a positive residual
Data points below the fitted regression line have a negative residual
Data points on the fitted regression line have zero residual.
The least squares line

The least squares line is the line where the sum of the squares of the residuals is as small as
possible; that is, it minimises:
the sum of the squares of the residuals = d12 + d22 + d32 + d42 + d52
Why do we minimise the sum of the squares of the residuals and not the sum of the
residuals? This is because the sum of the residuals for the least squares line is always zero.
The least squares line is like the mean. It balances out the data values on either side of itself.
Some residuals are positive and some negative, and in the end they add to zero. Squaring the
residuals solves this problem.
The least squares line

The least squares line is the line that minimises the sum of the squares of the residuals.
The assumptions for fitting a least squares line to data are the same as for using the
correlation coefficient, r. These are that:
the data is numerical the association is linear there are no clear outliers.
Determining the equation of the least squares regression

line
To determine exactly the equation of the least squares regression line we need to determine
the values of the intercept (a) and the slope (b) that define the line. The mathematics
required is beyond the scope of this course, but calculus can be used to give us rules for
these values:

The equation of the least squares regression line

The equation of the least squares regression line is given by y = a + bx, where:
rsy
the slope (b) is given by b=
sx
and
the intercept (a) is then given by a = y − bx
Here:
r is the correlation coefficient
s x and sy are the standard deviations of x and y
x and y are the mean values of x and y.
In these formulas y is the response variable, and x is the explanatory variable.
Note: The formula for the slope of the least squares regression line can be used to find
the value of the correlation coefficient (r), when the slope is known.
bs x
The correlation coefficient (r) is given by r =
sy
Warning!
If you do not correctly decide which is the explanatory variable (the x-variable) and
which is the response variable (the y-variable) before you start calculating the equation of
the least squares regression line, you will get the wrong answer.
Example 1 Determining the equation of the least squares regression line using
summary statistics and the correlation coefficient
The height and weight of 11 people have been recorded, and the values of the following
statistics determined:
height weight
mean 173.3 cm 65.45 kg
standard deviation 7.444 cm 7.594 kg
correlation coefficient r = 0.8502
Use the formula to determine the equation of the least squares regression line that enables
weight to be predicted from height. Calculate the values of the slope and intercept
rounded to two decimal places.

1 Identify and write down the EV: height (x)
explanatory variable (EV) and the
RV: weight (y)
response variable (RV). Label as x and
y, respectively.
2 Write down the given information. x = 173.3 s x = 7.444
y = 65.45 sy = 7.594
r = 0.8502
3 Calculate the slope. Slope:
rsy 0.8502 × 7.594
b= =
sx 7.444
= 0.87 (rounded to two significant figures)

4 Calculate the intercept. Intercept:
a = y − bx
= 65.45 − 0.87 × 173.3
= −85 (rounded to two significant figures)
5 Use the values of the intercept and the y = −85 + 0.87x
slope to write down the least squares
or
regression line using the variable
names. weight = −85 + 0.87 × height
Example 2 Determining the correlation coefficient using the slope of the least
squares regression line
Use the following information to find the value of the correlation coefficient r, rounded to
three significant figures.
hours studied exam score

mean 5.87 68.3
standard deviation 1.34 5.42
least squares equation exam score = 52.7 + 2.45 × hours studied
1 Identify and write down the EV: hours studied (x)
explanatory variable (EV) and the
RV: exam score (y)
response variable (RV). Label as x and
y, respectively.

2 Write down the required information.

b = 2.45 s x = 1.34 sy = 5.42
3 Calculate the correlation coefficient. Correlation coefficient:

bs x 2.45 × 1.34
r= =
sy 5.42
= 0.61 (rounded to two significant figures)
CAS 1: How to determine and graph the equation of a least squares

regression line using the TI-Nspire CAS
The following data give the height (in cm) and weight (in kg) of 11 people.
Height (x) 177 182 167 178 173 184 162 169 164 170 180
Weight (y) 74 75 62 63 64 74 57 55 56 68 72
Determine and graph the equation of the least squares regression line that will enable
weight to be predicted from height. Write the intercept and slope rounded to three
Steps
into lists named height and weight, as shown.
3 Identify the explanatory variable (EV) and
the response variable (RV).
EV: height
RV: weight
Note: In saying that we want to predict weight from
height, we are implying that height is the EV.
Construct a scatterplot with height (EV) on the
horizontal (or x-) axis and weight (RV) on the
vertical (or y-) axis.
Press b>Settings and click the Diagnostics
box. Select OK to activate this feature for all
future documents. This will show the coefficient
of determination (r2 ) whenever a regression is
performed.

5 Press b>Analyze>Regression>Show Linear (a

+ bx) to plot the regression line on the scatterplot.
Note that, simultaneously, the equation of the
regression line is shown on the screen.
The equation of the regression line is:
weight = −84.8 + 0.867 × height
The coefficient of determination is r2 = 0.723, rounded to three significant figures.
CAS 1: How to determine and graph the equation of a least squares

regression line using the ClassPad
The following data give the height (in cm) and weight (in kg) of 11 people.
Height (x) 177 182 167 178 173 184 162 169 164 170 180
Weight (y) 74 75 62 63 64 74 57 55 56 68 72
Determine and graph the equation of the least squares regression line that will enable
weight to be predicted from height. Write the intercept and slope rounded to three
Steps
and enter the
data into columns labelled height
and weight.
2 Tap to open the Set
StatGraphs dialog box and
complete as shown.
3 Tap in the toolbar at the top of
the screen to plot the scatterplot in
the bottom half of the screen.

3A 3A Fitting a least squares regression line to numerical data 175
4 To calculate the equation of the

least squares regression line:
Tap Calc from the menu bar.
Tap Regression and select
Linear Reg.
Complete the Set Calculations
dialog box as shown.
Tap OK to confirm your
selections in the Set
Calculations dialog box. This
also generates the regression
results shown opposite.
Tapping OK a second time automatically plots and displays the regression line.
Note: y6 as the formula destination is an arbitrary choice.
5 Use the values of the intercept a and weight = −84.8 + 0.867 × height (to three
slope b to write the equation of the significant figures)
least squares line in terms of the The coefficient of determination is r2 =
variables weight and height. 0.723, rounded to three significant places.
Skill-
sheet Exercise 3A
Basic ideas
1 What is a residual?
2 The least-squares regression line is obtained by:

A minimising the residuals
B minimising the sum of the residuals
C minimising the sum of the squares of the residuals
D minimising the square of the sum of the residuals
E maximising the sum of the squares of the residuals.
3 Write down the three assumptions we make about the association we are modelling
when we fit a least squares line to bivariate data.
Using a formula to calculate the equation of a least square line

Example 1 4 A least squares line y = a + bx is calculated for a set of bivariate data.
a Write down the explanatory variable for this least squares line.
b Given the following information, determine the equation of the least squares line,
giving the values of the intercept and slope rounded to three significant figures.

176 Chapter 3 Investigating and modelling linear associations 3A
x y
mean 10.65 19.91
5 We wish to find the equation of the least squares regression line that enables pollution
level beside a freeway to be predicted from traffic volume.
a Which is the response variable (RV) and which is the explanatory variable (EV)?
b Use the formula to determine the equation of the least squares regression line that
enables the pollution level to be predicted from the traffic volume where:
traffic volume pollution level

mean 11.4 231
Write the equation in terms of pollution level and traffic volume with the intercept
and slope rounded to two significant figures.
6 We wish to find the equation of the least squares regression line that enables life
expectancy in a country to be predicted from birth rate.
b Use the formula to determine the equation of the least squares regression line that
enables life expectancy to be predicted from birth rate, where:
life expectancy birth rate

mean 55.1 34.8
correlation coefficient r = −0.810
Write the equation in terms of life expectancy and birth rate with the y-intercept and
slope rounded to two significant figures.
Using a formula to calculate the correlation coefficient from the slope

Example 2 7 The equation of a least squares line y = a + bx is calculated for a set of bivariate data.
a Write down the response variable for this least-squares line.
b Use the following information to find the value of the correlation coefficient r,
rounded to three significant figures.

3A 3A Fitting a least squares regression line to numerical data 177
x y
mean 12.51 10.65
least squares equation y = 16.72 − 0.4847x
8 The equation of the least squares regression line that enables distance travelled by a car
(in 1000s of km) to be predicted from its age (in years) was found to be:
distance = 15.62 + 11.08 × age
b Use the following information to find the value of the correlation coefficient r,
distance age
mean 78.0 5.63
9 The following questions relate to the formulas used to calculate the slope and intercept
of the least squares regression line.
a A least squares line is calculated and the slope is found to be negative. What does
this tell us about the sign of the correlation coefficient?
b The correlation coefficient is zero. What does this tell us about the slope of the least
squares regression line?
c The correlation coefficient is zero. What does this tell us about the intercept of the
least squares regression line?
Using a CAS calculator to determine the equation of the least squares line from data
10 The table shows the number of sit-ups and push-ups performed by six students.
Sit-ups (x) 52 15 22 42 34 37
Push-ups (y) 37 26 23 51 31 45
Let the number of sit-ups be the explanatory (x) variable. Use your calculator to show
that the equation of the least squares regression line is:
push-ups = 16.5 + 0.566 × sit-ups (rounded to three significant figures)
11 The table shows average hours worked and university participation rates (%) in six
countries.
Hours 35.0 43.0 38.2 39.8 35.6 34.8
Rate 26 20 36 25 37 55

178 Chapter 3 Investigating and modelling linear associations 3A
Use your calculator to show that the equation of the least squares regression line that
enables participation rates to be predicted from hours worked is:
rate = 130 − 2.6 × hours (rounded to two significant figures)
12 The table shows the number of runs scored and balls faced by batsmen in a cricket
match.
Runs (y) 27 8 21 47 3 15 13 2 15 10 2
Balls faced (x) 29 16 19 62 13 40 16 9 28 26 6
a Use your calculator to show that the equation of the least squares regression line
enabling runs scored to be predicted from balls faced is:
y = −2.6 + 0.73x
b Rewrite the regression equation in terms of the variables involved.
13 The table below shows the number of TVs and cars owned (per 1000 people) in six
countries.
Number of TVs (y) 378 404 471 354 381 624

Number of cars (x) 417 286 435 370 357 550
We wish to predict the number of TVs from the number of cars.

a Which is the response variable?
b Show that, in terms of x and y, the equation of the regression line is:
y = 61.2 + 0.930x (rounded to three significant figures).

c Rewrite the regression equation in terms of the variables involved.

14 A least squares line of the form y = a + bx is fitted to a scatterplot. Which of the
following statements is always true:
A The line will divide the data points so that there are as many points above the line as
below the line.
B The sum of the vertical distances from the line to each data point will be a
minimum.
C x is the explanatory variable and y is the response variable.
D y is the explanatory variable and x is the response variable.
E Most of the data points will lie on the line.

3A 3B Using the least squares regression line to model a relationship 179
15 The statistical analysis of the set of bivariate data involving variables x and y resulted
in the information displayed in the table below:
x y
mean 32.5 88.1
least squares equation y = −2.56 + 1.45x
Using this information the value of the correlation coefficient r for this set of bivariate
data is closest to
A 0.73 B 0.34 C 0.50 D 0.53 E 0.78
16 A retailer recorded the number of ice creams sold and the day’s maximum temperature
over 8 consecutive Saturdays one summer.
Temperature (◦ C) 22 25 36 34 21 28 41 31
Number of ice creams sold 145 155 200 198 150 179 230 180
The equation of the least squares regression line fitted to the data is closest to:
A number of ice-creams = 4.08 + 58.2× temperature
B number of ice-creams = −12.9 + 0.237× temperature
C number of ice-creams = 58.2 + 4.08× temperature
D temperature = 3.57 + 72.3× number of ice-creams
E temperature = −12.8 + 0.237× number of ice-creams
3B Using the least squares regression line to model

a relationship between two numerical variables
Learning intentions
I To be able to interpret the intercept and slope of the least squares line.
I To be able to use the equation of the least squares line to make predictions.
I To be able to use the coefficient of determination in a regression analysis.
I To be able to use a residual plot to investigate the linearity assumption.
I To be able to report a regression analysis.
Suppose, for example, that we wish to investigate the nature of the association between the
price of a secondhand car and its age. The ultimate aim is to find a mathematical model that
will enable the price of a secondhand car to be predicted from its age.
The age (in years) and price (in dollars) of a selection of secondhand cars of the same brand
and model have been collected and are recorded in a table (shown).

Age (years) Price (dollars) Age (years) Price (dollars) Age (years) Price (dollars)
1 32 500 3 22 000 5 18 400
1 30 500 4 22 000 6 6 500
2 25 600 4 23 000 7 6 400
3 20 000 4 19 200 7 8 500
3 24 300 5 16 000 8 4 200
We start our investigation of the association between price and age by constructing a
scatterplot and using it to describe the association in terms of strength, direction and form.
In this analysis, age is the explanatory variable.
From the scatterplot, we see that there is a 40000
strong, negative, linear association between 35000
the price of the car and its age. There are no 30000
25000
Price (dollars)
clear outliers. The correlation coefficient is
20000
r = −0.9643.
15000
10000
5000
0
0 1 2 3 4 5 6 7 8 9 10
Age (years)
The equation of the least squares regression line from these data is:
price = 35 100 − 3940 × age
Interpreting the slope and intercept of a regression line

Interpreting the slope and intercept of a regression line
For the regression line y = a + bx:
the slope (b) estimates the average change (increase/decrease) in the response variable
(y) for each one-unit increase in the explanatory variable (x)
the intercept (a) estimates the average value of the response variable (y) when the
explanatory variable (x) equals 0.
Note: The interpretation of the y-intercept in a data context can be meaningless when x = 0 is not within
the range of observed x-values.
Consider again the least squares regression line relating the age of a car to its price:
price = 35100 − 3940 × age
The two key values in this mathematical model are the intercept (35100) and the slope
(−3940). The interpretation of these values is discussed in the following example.

3B Using the least squares regression line to model a relationship 181
Example 3 Interpreting the slope and intercept of a regression line
The equation of a regression line that 45000

enables the price of a second-hand car to be 40000
predicted from its age is: 35000
30000
Price (dollars)
price = 35 100 − 3940 × age
25000
a Interpret the slope in terms of the 20000
variables price and age. 15000
b Interpret the intercept in terms of the 10000
variables price and age. 5000
0
0 1 2 3 4 5 6 7 8 9 10
Age (years)
a The slope predicts the average change On average, for each additional
(increase/decrease) in the price for each 1-year year of age the price of these cars
increase in the age. Because the slope is decreases by $3940.
negative, it will be a decrease.
b The intercept predicts the value of the price of On average, the price of these cars
the car when age equals 0; that is, when the car when new was $35 100.
is new.
Using the regression line to make predictions
Example 4 Using the regression line to make predictions
The equation of a regression line that enables the price of a second-hand car to be
predicted from its age is:
price = 35 100 − 3940 × age
Use this equation to predict the price of a car that is 5.5 years old.

There are two ways this can be done. 40000
One is to draw a vertical arrow at 35000
age = 5.5 up to the graph and then horizontally 30000
Price (dollars)
25000
across to the price axis as shown, to get an
20000
answer of around $14 000.
15000 (5.5, 13 430)
A more accurate answer is obtained by
10000
substituting age = 5.5 into the equation to obtain 5000
$13 430, as shown below. 0
0 1 2 3 4 5 6 7 8 9 10
Price = 35 100 − 3940 × 5.5 Age (years)
= $13 430
Interpolation and extrapolation

When using a regression line to make predictions, we must be aware that, strictly speaking,
the equation we have found applies only to the range of data values used to derive the
equation.
For example, using the equation and rounding to the nearest dollar we would predict that:
a car which is 2 years old would have a price of $27 220 (price = 35 100 − 3940 × 2)
a car which is 7 years old would have a price of $7520 (price = 35 100 − 3940 × 7)
a car which is 12 years old would have a price of $-12 180 (price = 35 100 − 3940 × 12)
This last result, −$12 180 points to one of the limitations of substituting into a regression
equation without thinking carefully. Using this regression equation, we have predicted a
negative price, which is clearly not correct.
The problem is that we are using the regression equation to make predictions well outside
the range of values used to calculate this equation. We only have data for cars which are
up to 8 years old. Without knowing that the model works equally well for cars older than
this, which we don’t, we are venturing into unknown territory and can have little faith in our
predictions.
As a general rule, a regression equation only applies to the range of values of the
explanatory variables used to determine the equation. Thus, we are reasonably safe using
the line to make predictions that lie roughly within this data range, from 1 to 8 years. The
process of making a prediction within the range of values of the explanatory variable used
to derive the regression equation is called interpolation and we can have some faith in these
predictions.
However, we must be extremely careful about how much faith we put into predictions made
outside the range of values of the explanatory variable. Making predictions outside the data
range is called extrapolation.

Interpolation and extrapolation

Predicting within the range of values of the explanatory variable is called interpolation.
Interpolation is generally considered to give a reliable prediction.
Predicting outside range of values of the explanatory variable is called extrapolation.
Extrapolation is generally considered to give an unreliable prediction.
The coefficient of determination

In the previous chapter we define the coefficient of determination as r2 , where r is the value
of the correlation coefficient. The coefficient of determination can be considered a measure
of the predictive power of a regression equation. While the association between the price of
a second-hand car and its age does not explain all the variation in price, knowing the age of a
car does give us some information about its likely price.
For a perfect relationship, the regression line explains 100% of the variation in prices. In this
case, with r = −0.964 we have the:
coefficient of determination = r2 = (−0.9643)2 = 0.930 or 93.0%
Thus, we can conclude that:

93% of the variation in price of the second-hand cars can be explained by the variation
in the ages of the cars.
In this case, the regression equation has good predictive power. As a guide, any relationship
with a coefficient of determination greater than 30% can be regarded as having good
predictive power. In practice, even much lower values of the coefficient of determination can
useful.
Example 5 Using the coefficient of determination to compare associations
In a recent study across a number of countries the correlation between educational

attainment and the amount spent on education was found to be 0.26, whilst the correlation
between educational attainment and the student : teacher ratio was found to be −0.38.
a Find the values of the coefficient of determination between educational attainment and
the amount spent on education, and student : teacher ratio respectively.
b Which of the variables, amount spent on education or student : teacher ratio is more
important in explaining the variation in educational attainment?
Solution
a educational attainment: r2 = 0.262 = 6.8%
student : teacher ratio : r2 = (−0.38)2 = 14.4%

b The variable student : teacher ratio explains 14.4% of the variation in educational
attainment, making it a more important explanatory variable than the amount spent on
education which explains only 6.8%.
The residual plot – assessing the appropriateness of

fitting a linear model to data
So far all of our analysis has been based on the assumption that the relationship between
the two variables of interest is linear. This is why it has been essential to examine the
scatterplot before proceeding with any further analyses. However, sometimes the scatterplot
is not sensitive enough to reveal the non-linear structure of a relationship. To gain more
information we need to investigate the fit of the regression line to the data, and we do this
using a residual plot.
Residuals are defined as the vertical distances between the regression line and the actual
data value.
Residual plot
A residual plot is a graph of the residuals (plotted on the vertical axis) against the
explanatory variable (plotted on the horizontal axis), where:
Residual value = actual data value − predicted data value
Remember residuals can be positive, negative or zero.

To determine the appropriateness of fitting the least squares regression line to these data we
will construct a residual plot. But first, we need to calculate the residual for each value of the
explanatory variable, in this case age.
Example 6 Calculating a residual
The actual price of the 6-year-old car is $6500. Calculate the residual when its price is
predicted using the regression equation: price = 35100 − 3940 × age
1 Write down the actual price. Actual price: $6500
2 Determine the predicted price using
the least squares regression equation: Predicted price = 35 100 − 3940 × 6
price = 35100 − 3940 × age = $ 11 460
3 Determine the residual. Residual = actual - predicted
= $6500 − $11 460
= −$4960

By completing this calculation for all 4000

data points, we can construct a residual 2000
Residual
plot. Because the mean of the residuals
0
is always zero, we will construct the
−2000
horizontal axis for the plot at zero
−4000
(indicated by the red line) as shown.
−6000
1 2 3 4 5 6 7 8 9
Age (years)
What are we looking for in a residual plot?

The residual plot is used to check the linearity assumption required for a linear regression.
The scatterplot below shows a relationship that is clearly linear. When a line is fitted to the
data, the resultant residual plot appears to be a random collection of points roughly spread
around zero (the horizontal red line in the residual plot).
6 1
5
y=6–x .5
4
Residual
y 3 0 x
2
−.5
1
0 −1
0 1 2 3 4 5 6 0 1 2 3 4 5 6
x
By contrast, the relationship shown in the following scatterplot is clearly non-linear. Fitting
a straight line to the data results in the residual plot shown. While there is some random
behaviour, there is also a clearly identifiable curve shown in the scatterplot.
5 1
4 .5
Residual
3
y 0
2 x
−.5
1
0 −1
0 1 2 3 4 5 6 7 8 9 10 0 2 4 6 8 10
x

In summary, if a residual plot shows evidence of some sort of systematic behaviour (a

pattern), then it is likely that the underlying relationship is non-linear. However, if the
residual plot appears to be a random collection of points roughly spread around zero, then
we can be happy that our original assumption of linearity was reasonable and that we
have appropriately modelled the data. From a visual inspection, it is difficult to say with
certainty that a residual plot is random. It is easier to see when it is not random. For present
purposes, it is sufficient to say that a clear lack of a pattern in a residual plot is an indication
of randomness.
Example 7 Interpreting a residual plot
Which of the following residual plots would call into question the assumption of linearity
in a regression analysis? Give reasons for your answers.
1 1
.5 .5
Residual
Residual
0 0 x
x
−.5 −.5
−1 −1
0 2 4 6 8 10 0 2 4 6 8 10
A B
1 1
.5 .5
Residual
Residual
0 0 x
x
−.5 −.5
−1 −1
0 2 4 6 8 10 0 2 4 6 8 10
C D

Examine each plot, looking for a pattern Plot A – residuals look random, so linearity
or structure in the residual. assumption is met.
Plot B – there is a clear curve in the
residuals, the linearity assumption is not
met.
Plot C – residuals look random, so
linearity assumption is met.
Plot B – there is a clear curve in the
residuals, the linearity assumption is not
met.
Performing a regression analysis

A full regression analysis involves all of the following analyses, the results of which are
collated in a report.
Performing a regression analysis

To carry out a regression analysis involves several processes, which include:
constructing a scatterplot to investigate the nature of an association
calculating the correlation coefficient to indicate the strength of the relationship
determining the equation of the regression line
interpreting the coefficients of the y-intercept (a) and the slope (b) of the least squares
regression line y = a + bx
calculating and interpreting the coefficient of determination
using the regression line to make predictions
calculating residuals and using a residual plot to test the assumption of linearity
writing a report on your findings.
Reporting the results of a regression analysis

The final step is to construct a report which brings together all of the analyses which have
been described in this section, as shown in the following example.

188 Chapter 3 Investigating and modelling linear associations 3B
Example 8 Reporting the results of a regression analysis
Construct a report to describe the association between the price and age of secondhand
cars.
Solution
From the scatterplot we see that there is a strong negative, linear association between the
price of a second hand car and its age, r = −0.964. There are no obvious outliers.
The equation of the least squares regression line is: price = 35 100 − 3940 × age.
The slope of the regression line predicts that, on average, the price of these
second-hand cars decreased by $3940 each year.
The intercept predicts that, on average, the price of these cars when new was $35 100.
The coefficient of determination indicates that 93% of the variation in the price of these
second-hand cars is explained by the variation in their age.
The lack of a clear pattern in the residual plot confirms the assumption of a linear
association between the price and the age of these second-hand cars.
Skill-
sheet Exercise 3B
Some basics
1 Use the line on the scatterplot opposite to 100
determine the equation of the regression line in
80
terms of the variables, mark and days absent.
Give the intercept correct to the nearest whole 60
Mark
number and the slope correct to one decimal

40
place.
20
0
0 1 2 3 4 5 6 7 8
Days absent
Interpreting the intercept and slope of a regression line

Example 3 2 The equation of a regression line that enables hand span (in cm) to be predicted from
height (in cm) is:
hand span = 2.9 + 0.33 × height
a Write down the value of the intercept, and interpret this value in this context of the
variables in the equation.
b Write down the value of the slope, and interpret this value in this context of the

3B 3B Using the least squares regression line to model a relationship 189
3 The following regression equation can be used to predict a company’s weekly sales ($)
from their weekly online advertising expenditure ($).
sales = 575 + 4.85 × expenditure
Using the regression line to make predictions

Example 4 4 For children between the ages of 36 and 60 months, the equation relating their height
(in cm) to their age (in months) is:
height = 72 + 0.40 × age
Use this equation to predict the height (to the nearest cm) of a child with the following
ages. In each case indicate whether you are interpolating or extrapolating.
a 20 months old b 50 months old c 65 months old
5 When preparing between 25 and 100 meals, a hospital’s cost (in dollars) is given by the
equation:
cost = 487.50 + 6.70 × meals
Use this equation to predict the cost (to the nearest dollar) of preparing the following
meals. Are you interpolating or extrapolating?
a 0 meals b 80 meals c 110 meals
6 For males of heights from 150 cm to 190 cm tall cm, the equation relating a son’s
height (in cm) to his father’s height (in cm) is:
son’s height = 83.9 + 0.525 × f ather’s height
Use this equation to predict (to the nearest cm) the adult height of a male whose father
is the following heights. State, with a reason, how reliable your predictions are in each
case.
a 170 cm tall b 200 cm tall c 155 cm tall
Using the coefficient of determination to compare associations

Example 5 7 A teacher found the correlation between her students’ scores on an IQ test (IQ) and
their final examination score in Year 12 (exam score) is 0.45, whilst the correlation
between the average number of hours they spend each week studying mathematics
(hours) and their final examination score in Year 12 (exam score) is 0.65.
a Determine the value of the coefficient of determination between exam score and IQ,
expressed as a percentage rounded to one decimal place.
b Determine the value of the coefficient of determination between exam score and
hours, expressed as a percentage rounded to one decimal place.

c Which of the variables, IQ or hours is more important in explaining the variation in

exam score?
Calculating a residual
Example 6 8 The equation of a regression line that enables hand span to be predicted from height is:
hand span = 2.9 + 0.33 × height
a Using this equation, show that the predicted hand span of a person who is 160 cm is
55.7 cm.
b This person has an actual hand span of 58.5 cm. Show that the residual value for
this person is 2.8 cm.
9 For a 100 km trip, the equation of a regression line that enables fuel consumption of a
car (in litres) to be predicted from its weight (kg) is:
fuel consumption = −0.1 + 0.01 × weight
a Use this equation to predict (to one decimal place) the fuel consumption of a car
which weighs 980kg.
b This car has an actual fuel consumption of 8.9 litres. What is the residual value for
this for this data point?
10 From the scatterplot shown determine (to the y

nearest whole number) the residual values when 10
the value of x is equal to: 9
8
a 1 7
b 3 6
5
c 8 4
3
2
1
0 x
0 1 2 3 4 5 6 7 8 9 10
Interpreting a residual plot

Example 7 11 Each of the following residual plots has been A 4.5
constructed after a least squares regression line 3.0
residual
has been fitted to a scatterplot. Which of the 1.5

residual plots suggest that the use of a linear 0.0
model to fit the data was inappropriate? Why? −1.5
0 2 4 6 8 10 12 14
B C
3.0 3
1.5
residual
residual
0
0.0
−1.5 −3
−3.0
0 2 4 6 8 10 12 0 2 4 6 8 10 12 14
12 In an investigation of the association between the food energy content (in calories) and
the fat content (in g) in a standard-sized packet of chips, the least squares regression
line was found to be:
energy content = 27.8 + 14.7 × fat content r2 = 0.7569
c Interpret the value of the coefficient of determination in terms of the variables in
energy content and fat content.
d Use this equation to predict the energy content of a packet of chips which contains 8
grams of fat.
e If the actual energy content of a packet of chips containing 8 grams of fat is 132
calories, what is the value of the residual?
13 In an investigation of the association between the success rate (%) of sinking a putt and
the distance from the hole (in cm) of amateur golfers, the least squares regression line
was found to be:
success rate = 98.5 − 0.278 × distance r2 = 0.497
a Write down the slope of this regression equation and interpret.
b Use the equation to predict the success rate when a golfer is 90 cm from the hole.
c At what distance (in metres) from the hole does the regression equation predict an
amateur golfer to have a 0% success rate of sinking the putt?
d Calculate the value of r, rounded to three decimal places.
e Write down the value of the coefficient of determination in percentage terms and
interpret.
14 The scatterplot opposite shows the pay 24
rate (dollars per hour) paid by a company 23
22
to workers with different years of work
Pay rate ($)
21
experience. Using a calculator, the equation 20
of the least squares regression line is found 19
18
to have the equation: 17
16
y = 18.56 + 0.289x with r = 0.967
15
0 2 4 6 8 10 12
Experience (years)
a Is it appropriate to fit a least squares regression line to the data? Why?
b Work out the coefficient of determination.
c What percentage of the variation in a person’s pay rate can be explained by the
variation in their work experience?

d Write down the equation of the least squares line in terms of the variables pay rate
and years of experience.
e Interpret the y-intercept in terms of the variables pay rate and years of experience.
What does the y-intercept tell you?
f Interpret the slope in terms of the variables pay rate and years of experience. What
does the slope of the regression line tell you?
g Use the least squares regression equation to:
i predict the hourly wage of a person with 8 years of experience
ii determine the residual value if the person’s actual hourly wage is $21.20.
h The residual plot for this regression 0.5
0.4
analysis is shown opposite. Does 0.3
0.2
Residual
the residual plot support the initial 0.1
assumption that the relationship 0
−0.1
between pay rate and years of −0.2
−0.3
experience is linear? −0.4
Explain your answer. Experience (years)
15 The scatterplot opposite shows scores on a 5
hearing test against age. In analysing the data, a
4
Hearing test score
statistician produced the following statistics:

coefficient of determination: r2 = 0.370
3
least squares line: y = 4.9 − 0.043x
a Determine the value of Pearson’s correlation 2
coefficient, r, for the data.
b Interpret the coefficient of determination 0
25 30 35 40 45 50 55 60
in terms of the variables hearing test score
Age (years)
and age.
c Write down the equation of the least squares line in terms of the variables hearing
test score and age.
d Write down the slope and interpret.
e Use the least squares regression equation to:
i predict the hearing test score of a person who is 20 years old
ii determine the residual value if the person’s actual hearing test score is 2.0.
f Use the graph to estimate the value of the residual for the person aged:
i 35 years ii 55 years.

g The residual plot for this regression analysis is

0.5
shown opposite.
residual
Does the residual plot support the initial 0.0
assumption that the relationship between hearing −0.5
test score and age is essentially linear? Explain −1.2
your answer. 0 2 4 6 8 10 12
age
Reporting the results of a residual analysis

Example 8 16 In a study of the effectiveness of a pain relief
55
response time
drug, the response time (in minutes) was
measured for different drug doses (in mg). A least 40
squares regression analysis was conducted to 25
enable response time to be predicted from drug 10

dose. The results of the analysis are displayed. 0.0 1.0 2.0 3.0 4.0 5.0 6.0
drug dose
6
residual
Regression equation: y = a + bx 0
a = 55.8947
b = −9.30612 −6
r2 = 0.901028
r = −0.949225 0.0 1.0 2.0 3.0 4.0 5.0 6.0
drug dose
Use this information to complete the following report. Call the two variables drug dose
and response time. In this analysis drug dose is the explanatory variable.
Report
From the scatterplot we see that there is a strong relationship between
response time and :r= . There are no obvious outliers.
The equation of the least squares regression line is:
response time = + × drug dose
The slope of the regression line predicts that, on average, response time
increases/decreases by minutes for a 1-milligram increase in drug
dose.
The y-intercept of the regression line predicts that, on average, the response time
when no drug is administered is minutes.
The coefficient of determination indicates that, on average, % of the
variation in is explained by the variation in .
The residual plot shows a , calling into question the use of a linear equation
to describe the relationship between response time and drug dose.

17 A regression analysis was conducted to investigate

27.5
the nature of the association between femur (thigh
radius
bone) length and radius (the short thicker bone 26.5
in the forearm) length in 18-year-old males. The 25.5

bone lengths are measured in centimetres. The 24.5
results of this analysis are reported below. In this 42.543.544.545.546.547.5
femur
investigation, femur length was treated as the
independent variable.
0.15
residual
Regression equation y = a + bx 0.00
a = −7.24946
0.15
b = 0.739556
−0.30
r2 = 0.975291
r = 0.987568 42.5 43.5 44.5 45.5 46.5 47.5
femur
Use the format of the report given in the previous question to summarise findings of
this investigation. Call the two variables femur length and radius length.

18 The scatterplot shows the weight 120
(in kg) and waist measurement 115
110
(in cm) for a group of people. A 105
100
Weight(kg)
least squares line had been fitted 95

90
to the scatterplot with waist as the 85
80
explanatory variable. The equation 75
70
of the least squares line is closest 65
60
to: 70 75 80 85 90 95 100 105 110 115 120
Waist(cm)
A weight = 60.0 + 1.10 × waist B waist = 60.0 + 0.91 × weight

C weight = 70.0 + 1.10 × waist D weight = −3.70 + 0.91 × waist
E weight = −17.0 + 1.10 × waist
19 The table below shows the life expectancy in years and the percentage of government
expenditure which is spent on health (health) in 10 countries.
Health 17.3 10.3 4.7 6.0 20.1 6.0 13.2 7.7 10.1 17.5
Life expectancy (years) 82 76 68 69 83 75 76 76 75 75
A least squares line which enables a country’s life expectancy to be predicted from
their expenditure on health is fitted to the data. The number of times that a country’s
predicted life expectancy is greater than their actual life expectancy is:
A 3 B 4 C 5 D 6 E 7

3B 3C Conducting a regression analysis using data 195
20 In a study of the association between the length in cm and weight in grams of a certain
species of fish the following least squares line was obtained:
weight = −329 + 23.3 × length
Which one of the following is a conclusion that can be made from this least squares
line?
A On average, the weight of the fish increased by 23.3 grams for each centimetre
increase in length.
B On average, the length of the fish increased by 23.3 cm for each one gram increase
in weight.
C On average, the weight of the fish decreased by 329 grams grams for each
centimetre increase in length.
D The equation cannot be correct as the weight of the fish can never be negative.
E The weight of the fish in grams can be determined by subtracting 305.7 from their
length.
3C Conducting a regression analysis using data

In your statistical investigation project you will need to be able to conduct a full regression
analysis from data. This section is designed to help you with this task.
CAS 2: How to conduct a regression analysis using the TI-Nspire CAS

This analysis is concerned with investigating the association between life expectancy (in
years) and birth rate (in births per 1000 people) in 10 countries.
Birth rate 30 38 38 43 34 42 31 32 26 34
Steps
1 Write down the explanatory variable EV: birth
(EV) and response variable (RV). Use RV: life
the variable names birth and life.
2 Start a new document by pressing
/ + N.
Select Add Lists & Spreadsheet.
Enter the data into the lists named birth
and life, as shown.

3 Construct a scatterplot to investigate

the nature of the relationship between
life expectancy and birth rate.
4 Describe the association shown by the There is a strong, negative, linear

scatterplot. Mention direction, form, relationship between life expectancy and
strength and outliers. birth rate. There are no obvious outliers.
5 Find and plot the equation of the least
squares regression line and r2 value.
Note: Check if Diagnostics is activated
using b>Settings.
6 Generate a residual plot to test the

linearity assumption.
Use / + (or click on the page tab)
to return to the scatterplot.
To hide the residual plot press
b>Analyze>Residuals>Hide
Residual Plot.
7 Use the values of the intercept and Regression equation:
slope to write the equation of the least life = 105.4 − 1.445 × birth
squares regression line. Also write
the values of r and the coefficient of Correlation coefficient: r = −0.8069
determination. Coefficient of determination: r2 = 0.651
CAS 2: How to conduct a regression analysis using the ClassPad

The data for this analysis are shown below.
Birth rate (per thousand) 30 38 38 43 34 42 31 32 26 34

Steps
1 Write down the explanatory variable EV: birth
(EV) and response variable (RV). Use RV: life
the variable names birth and life.
3C Conducting a regression analysis using data 197
2 Enter the data into lists as shown.
3 Construct a scatterplot to investigate

the nature of the relationship between
life expectancy and birth rate.
a Tap and complete the Set

Calculations dialog box as
shown.
b Tap to view the scatterplot.
4 Describe the association shown by the There is a strong negative, linear

scatterplot. Mention direction, form, association between life expectancy and
strength and outliers. birth rate. There are no obvious outliers.
5 Find the equation of the least
squares regression line and
generate all regression statistics,
including residuals.
a Tap Calc in the toolbar.
Tap Regression and select Linear Reg.
b Complete the Set Calculations dialog box as shown.
Note: Copy Residual copies the residuals to list3, where they
can be used later to create a residual plot.
c Tap OK in the Set Calculation box to

generate the regression results.

198 Chapter 3 Investigating and modelling linear associations 3C
d Write down the key results. Regression equation:

life = 105.4 − 1.445 × birth
Correlation coefficient:
r = −0.8069
Coefficient of determination:
r 2 = 0.651
6 Tapping OK a second time automatically
plots and displays the regression line
on the scatterplot.
To obtain a full-screen plot, tap from the
icon panel.
7 Generate a residual plot to test the
linearity assumption.
Tap and complete the Set
Calculations dialog box as shown.
Tap to view the residual plot.
Inspect the plot and write your The random residual plot suggests
conclusion. linearity.
Note: When you performed a regression analysis earlier, the residuals were calculated automatically
and stored in list3. The residual plot is a scatterplot with list3 on the vertical axis and birth on the
horizontal axis.
Exercise 3C
1 The table below shows the scores obtained by nine students on two tests. We want to
be able to predict test B scores from test A scores.
Test A score (x) 18 15 9 12 11 19 11 14 16

Test B score (y) 15 17 11 10 13 17 11 15 19
Use your calculator to perform each of the following steps of a regression analysis.
a Construct a scatterplot. Name the variables test a and test b.
b Determine the equation of the least squares line along with the values of r and r2 .

3C 3C Conducting a regression analysis using data 199
c Display the regression line on the scatterplot.

d Obtain a residual plot.
2 The table below shows the number of careless errors made on a test by nine students.
Also given are their test scores. We want to be able to predict test score from the
number of careless errors made.
Test score 18 15 9 12 11 19 11 14 16
Careless errors 0 2 5 6 4 1 8 3 1
Use your calculator to perform each of the following steps of a regression analysis.
a Construct a scatterplot. Name the variables score and errors.
b Determine the equation of the least squares line along with the values of r and r2 .
Write answers rounded to three significant figures.
c Display the regression line on the scatterplot.
d Obtain a residual plot.
3 How well can we predict an adult’s weight from their birth weight? The weights of 12
adults were recorded, along with their birth weights. The results are shown.
Birth weight (kg) 1.9 2.4 2.6 2.7 2.9 3.2 3.4 3.4 3.6 3.7 3.8 4.1
Adult weight (kg) 47.6 53.1 52.2 56.2 57.6 59.9 55.3 58.5 56.7 59.9 63.5 61.2
a In this investigation, which would be the RV and which would be the EV?
b Construct a scatterplot.
c Use the scatterplot to:
i comment on the association between adult weight and birth weight in terms of
direction, outliers, form and strength
ii estimate the value of Pearson’s correlation coefficient, r.
d Determine the equation of the least squares regression line, the coefficient of
determination and the value of Pearson’s correlation coefficient, r. Write answers
e Interpret the coefficient of determination in terms of adult weight and birth weight.
f Interpret the slope in terms of adult weight and birth weight.
g Use the regression equation to predict the weight of an adult with a birth weight of:
i 3.0 kg ii 2.5 kg iii 3.9 kg.
Give answers correct to one decimal place.
h It is generally considered that birth weight is a ‘good’ predictor of adult weight. Do
you think the data support this contention? Explain.
i Construct a residual plot and use it to comment on the appropriateness of assuming
that adult weight and birth weight are linearly associated.

Review 200 Chapter 3 Investigating and modelling linear associations
Bivariate data Bivariate data are data in which each observation involves recording
Assign- information about two variables for the same person or thing. An
ment
example would be the heights and weights of the children in a
preschool.
Linear regression The process of fitting a line to data is known as linear regression. The
association can then be described by a rule of the form y = a + bx
In this equation:
y is the response variable
x is the explanatory variable
a is the y-intercept
b is the slope of the line.
Residuals The vertical distance from a data point to the straight line is called a
residual: residual value = data value − predicted value.
Least squares The least squares method is one way of finding the equation of a
method regression line. It minimises the sum of the squares of the residuals. It
works best when there are no outliers.
Determining The equation of the least squares regression line is given by y = a + bx,
the values of a where:
and b from the rsy
formulas the slope (b) is given by b=
sx
and
the intercept (a) is then given by a = y − bx
Here:
r is the correlation coefficient
x and y are the mean values of x and y.
Determining the The value of the correlation coefficient r is given by

value of r when b bs x
is known r=
sy
Here:
b is the slope of the least squares line

Review
Interpreting the For the regression line y = a + bx:
intercept and
the slope (b) tells us on average the change in the response variable
slope
(y) for each one-unit increase or decrease in the explanatory variable
(x).
the intercept (a) tells us on average the value of the response variable
(y) when the explanatory variable (x) equals 0.
Consider for example the regression line
cost = 1.20 + 0.06 × number o f pages
The slope of the regression line tells us that on average the cost of a
textbook increases by 6 cents ($0.06) for each additional page.
The intercept of the line tells us that on average that a book with no
pages costs $1.20 (this might be the cost of the cover).
Making The regression line y = a + bx enables the value of y to be predicted for

predictions a given value of x by substitution into the equation. For example, using
the previous equation
cost = 1.20 + 0.06 × number o f pages
predicts that the cost of a 100-page book is:
cost = 1.20 + 0.06 × 100 = $7.20
Interpolation and Predicting within the range of the values of the explanatory variable is
extrapolation called interpolation, and will give a reliable prediction.
Predicting outside the range of the values of the explanatory variable is
called extrapolation, and will give an unreliable prediction.
Coefficient of The coefficient of determination (r2 ) gives a measure of the predictive

determination power of a regression line. For example, for the regression line above,
the coefficient of determination is 0.81.
From this we conclude that 81% of the variation in the cost of a
textbook can be explained by the variation in the number of pages.
Residual plots Residual plots can be used to test the linearity assumption by plotting
the residuals against the EV.
A residual plot that appears to be a random collection of points
clustered around zero supports the linearity assumption.
A residual plot that shows a clear pattern indicates that the association is
not linear.

Skills checklist
list
3A 1 I can determine the equation of the least squares regression line using the
formulas.
3A 2 I can determine the correlation coefficient from the slope of the least squares
regression line using the formula.
3A 3 I can determine the equation of the least squares regression line using a CAS
calculator.
See CAS 1, and Exercise 3A Question 10
3B 4 I can interpret the slope and intercept of a regression line.
3B 5 I can use the regression line to make predictions.
3B 6 I can use the coefficient of determination to compare associations.
3B 7 I can calculate residual values.
3B 8 I can interpret a residual plot.
3B 9 I can write a report based on a regression analysis.
3C 10 I can use a CAS calculator to generate all of the analyses required for a
regression analysis.
See CAS 2, and Exercise 3C Question 1

Review
1 When using a least squares line to model a relationship displayed in a scatterplot, one
key assumption is that:
A there are two variables B the variables are related
C the variables are linearly related D r2 > 0.5
E the correlation coefficient is positive
2 For the least squares regression line y = −1.2 + 0.52x:

A the y-intercept = −0.52 and slope = −1.2
B the y-intercept = 0 and slope = −1.2
C the y-intercept = 0.52 and slope = −1.2
D the y-intercept = −1.2 and slope = 0.52
E the y-intercept = 1.2 and slope = −0.52
3 If the equation of a least squares regression line is y = 8 − 9x and r2 = 0.25:

A r = −0.5 B r = −0.25 C r = −0.0625 D r = 0.25 E r = 0.50
4 Given that b = 1.328, s x = 1.871 and sy = 3.391, the correlation coefficient, r, is closest
to:
A 0.357 B 0.598 C 0.733 D 0.773 E 1.33
5 The association between the number of time errors

errors made in a task, and the time spent
mean 8.00 34.5
practicing the task (in minutes) was found to
be approximately linear, and the values of standard deviation 2.40 12.5
the following statistics were determined: correlation coefficient r = −0.236
The equation of the least squares line that enables errors to be predicted from time is
given by
A errors = 52.2 − 1.23 × time B errors = 10.1 − 0.99 × time
C errors = 24.7 − 1.23 × time D errors = 32.6 + 0.24 × time
E errors = 44.3 − 1.23 × time
6 The speed at which a car is travelling speed distance

(in km/hr), and the distance (in metres)
mean 90.5 52.7
taken by the car to come to a stop when
the brakes are applied, were recorded standard deviation 1.124 1.349
over speeds from 60km/hr to 120km/hr. correlation coefficient r = 0.948
The association was found to be approximately linear, and the values of the statistics
shown were determined.

On average, for each additional km/hr of speed, the distance taken to come to a stop
A decreased by 1.14 metres B decreased by 0.79 metres
C increased by 1.14 metres D increased by 0.95 metres
E increased by 0.79 metres
7 The least squares regression line y = 8 − 9x predicts that, when x = 5, the value of y is:
A −45 B −37 C 37 D 45 E 53
8 A least squares regression line of the form x 25 15 10 5

y = a + bx is fitted to the data set shown.
y 10 10 15 25
The equation of the line is:
A y = −0.69 + 24.4x B y = 24.4 − 0.69x C y = 24.4 + 0.69x
D y = 28.7 − x E y = 28.7 + x
9 A least squares regression line of the form y 30 25 15 10

y = a + bx is fitted to the data set shown.
x 40 20 30 10
The equation of the line is:
A y = 1 + 0.5x B y = 0.5 + x C y = 0.5 + 7.5x
D y = 7.5 + 0.5x E y = 30 − 0.5x
10 Using a least squares regression line, the predicted value of a data value is 78.6. The
residual value is –5.4. The actual data value is:
A 73.2 B 84.0 C 88.6 D 94.6 E 424.4
11 The equation of the least squares line plotted on the y
scatterplot opposite is closest to: 10
9
A y = 8.7 − 0.9x 8
7
B y = 8.7 + 0.9x 6
C y = 0.9 − 8.7x 5
4
D y = 0.9 + 8.7x 3
2
E y = 8.7 − 0.1x 1
0 x
0 1 2 3 4 5 6 7 8 9 10
12 The equation of the regression line plotted on the y

scatterplot opposite is closest to: 10
9
A y = −14 + 0.8x 8
7
B y = 0.8 + 14x 6
C y = 2.5 + 0.8x 5
4
D y = 14 − 0.8x 3
2
E y = 17 + 1.2x 1
0 x
20 22 24 26 28 30
Review
Weight (in kg) can be predicted from height (in cm) from the regression line:
weight = −96 + 0.95 × height, with r = 0.79
13 Which of the following statements relating to the regression line is false?

A The slope of the regression line is 0.95.
B The explanatory variable in the regression equation is height.
C The least squares line does not pass through the origin.
D The intercept is 96.
E The equation predicts that a person who is 180 cm tall will weigh 75 kg.
14 This regression line predicts that, on average, weight:

A decreases by 96 kg for each 1 centimetre increase in height
B increases by 96 kg for each 1 centimetre increase in height
C decreases by 0.79 kg for each 1 centimetre increase in height
D decreases by 0.95 kg for each 1 centimetre increase in height
E increases by 0.95 kg for each 1 centimetre increase in height
15 Noting that the value of the correlation coefficient is r = 0.79, we can say that:
A 62% of the variation in weight can be explained by the variation in height
B 79% of the variation in weight can be explained by the variation in height
C 88% of the variation in weight can be explained by the variation in height
D 79% of the variation in height can be explained by the variation in weight
E 95% of the variation in height can be explained by the variation in weight
16 A person of height 179 cm weighs 82 kg. If the regression equation is used to predict
their weight, then the residual will be closest to:
A –8 kg B 3 kg C 8 kg D 9 kg E 74 kg
The scatterplot shows the association 100
between a student’s mark on a test, 90
80
and the number of days absent during the term. 70
60
Mark
50
40
30
20
10
0
0 1 2 3 4 5 6 7 8
Days absent

17 The median mark for this group of students is closest to:

A 55 B 60 C 67 D 70 E 72
18 The median days absent for this group of students is closest to:
A 2 B 3 C 4 D 55 E 62.5
19 The coefficient of determination between mark and days absent is r2 = 0.5.

The correlation coefficient is closest to:
A −0.7 B −0.25 C 0.25 D 0.5 E 0.7
20 There were two students who were absent for 2 days that term. The values of the
residuals for these students are
A 0 B 10 C 60 and 80 D −10 and 10 E −10
21 Using the graph of the least squares line, we predict that a student who is absent for 4
days would receive a mark of about:
A 48 B 51 C 62 D 65 E 67
22 The table below shows the weight in grams and the length in cm for a certain species of
fish.
Length(cm) 13.5 14.3 16.3 17.5 18.4 19.0 19.0 19.8 21.2 23.0
Weight(gm) 55 60 90 120 150 140 170 145 200 273
A least squares line which enables a a fish’s weight to be predicted from their length is
fitted to the data. The number of times that the fish’s predicted weight is greater than
their actual weight is:
A 3 B 4 C 5 D 6 E 7
23 The value of the correlation coefficient r for these data is equal to 0.965. The
percentage of variation in fish weight which is not explained by the length of the fish is
closest to:
A 96.5% B 93.1% C 9.3% D 6.9% E 3.5%

1 The table below shows the age (in years), the number of seats, and the airspeed (in
km/h), of eight aircraft.
Age 3.5 3.7 4.7 4.9 5.1 7.3 8.7 8.8

number of seats 405 296 288 258 240 193 188 148
airspeed 830 797 774 736 757 765 760 718

Review
a Determine to the nearest whole number:
i the median age of these aircraft.
ii the mean and standard deviation of the airspeed of these aircraft.
To investigate the association between the number of seats and airspeed, a least squares
line is fitted to the data. The response variable in this investigation is airspeed.
b Determine and write down the equation of the least squares line in terms of number
of seats and airspeed. Round the intercept and slope to 3 significant figures.
c Determine and write down the percentage of variation in the airspeed that is
explained by the number of seats. Write the answer rounded to 1 decimal place.
2 In an investigation of the relationship between the hours of sunshine (per year) and
days of rain (per year) for 25 cities, the least squares regression line was found to be:
hours o f sunshine = 2850 − 6.88 × days o f rain, with r2 = 0.484
Use this information to complete the following sentences.
a In this regression equation, the explanatory variable is .
b The slope is and the intercept is .
c The regression equation predicts that a city that has 120 days of rain per year will
have hours of sunshine per year.
d The slope of the regression line predicts that the hours of sunshine per year will
by hours for each additional day of rain.
e r= , correct to three significant figures.
f % of the variation in sunshine hours can be explained by the variation in
.
g One of the cities used to determine the regression equation had 142 days of rain and
1390 hours of sunshine.
i The regression equation predicts that it has hours of sunshine.
ii The residual value for this city is hours.
h Using a regression line to make predictions within the range of data used to
determine the regression equation is called .
3 The cost of preparing meals, in dollars, and

550
the number of meals prepared are plotted
in the scatterplot shown. A least squares 500
line has been fitted to the data which 450
Cost ($)
enables the cost of the meals prepared to

400
be predicted from the number of meals
prepared. 350
300
25 30 35 40 45 50 55 60 65 70 75 80
Number of meals

a Which is the response variable?

b Describe the association in terms of strength, direction and form.
The equation of the least squares line that relates the cost of preparing meals to the
number of meals produced is:
cost = 222.48 + 4.039 × number o f meals
c i Use the equation to predict the cost of preparing 21 meals. Round the answer to
the nearest cent.
ii In making this prediction, are you interpolating or extrapolating?
d Write down:
i the intercept of the regression line and interpret in terms of cost and the number
of meals prepared.
ii the slope of the regression line and interpret in terms of cost and the number of
meals prepared.
e When the number of meals prepared was 50, the cost of preparation was $444.
Show that, when the least squares line is used to predict the cost of preparing 50
meals, the residual is $19.57, to the nearest cent.
4 We wish to find the equation of the least squares regression line that will enable height
(in cm) to be predicted from femur (thigh bone) length (in cm).
a Which is the RV and which is the EV?
b Use the summary statistics femur length height
shown to determine the
mean 24.246 166.092
equation of the least
squares regression line standard deviation 1.873 10.086
that will enable height to correlation coefficient r = 0.9939
be predicted from femur
length.
Write the equation in terms of height and femur length. Give the slope and intercept
c Interpret the slope of the regression equation in terms of height and femur length.
d Determine the value of the coefficient of determination and interpret in terms of
height and femur length.
Arm span is also associated with height. A least squares regression line that can be
used to model this association is:
height = 0.498 + 0.926× arm span
In determining this equation, the arm span height
summary statistics displayed in the table
mean 169.615 166.092
were also calculated.

Review
e Determine the percentage of the variation in height explained by the variation in
arm span. Write the answer as a percentage rounded to one decimal place.
5 The scatter plot shows the height (in cm) of

a group of 10 children plotted against their 140
age (in years). The data used to generate this
Height (cm)
scatterplot is shown below. 120
100
80
0 2 4 6 8 10 12
Age (years)
Height (cm) 86.5 95.5 103.0 109.8 116.4 122.4 128.2 133.8 139.6 145.0
Age (years) 2 3 4 5 6 7 8 9 10 11
The task is to determine the equation of a least squares regression line that can be used
to predict height from age.
a In this analysis, which would be the RV and which would be the EV?
b Use the scatter plot to describe the association between age and height in terms of
strength and direction.
c Use your calculator to confirm that the equation of the least squares regression
line is: height = 76.64 + 6.366 × age and r = 0.9973.
d i Use the regression line to show that the predicted height of a one-year old is 83.0
cm, rounded to 3 significant figures.
ii In making this prediction are you extrapolating or interpolating?
e Interpret the slope of the least squares line in terms of height and age.
f Determine the percentage of the variation in height of these children explained by
their age. Round your answer to 1 decimal place.
g Use the least squares regression equation to:
i predict the height of the 10-year-old child in this sample
ii determine the residual value for this child.
h i Confirm that the residual plot for this
0.15
analysis is shown opposite.
residual
0.00
ii Explain why this residual plot suggests that
–0.15
a linear equation is not the most appropriate
–0.30
model for this association.
2 3 4 5 6 7 8 9 1011
age

6 The heart rate (in beats/minute) was measured and recorded for a group of 13 students.
The students then completed the same set of exercises and their heart rate measured
again immediately on completion. The scatterplot below shows the students’ heart
rate after exercise plotted against their heart rate before exercise, with a least squares
regression line fitted. Also shown is the residual plot for this line.
Heart rate after exercise

145
140
(beats/min)
135
130
125
120
115
60 65 70 75 80 85 90 95 100
Heart rate before exercise (beats/min)
10
8
6
4
Residual
2
0
−2
−4
−6
−8
−10
60 65 70 75 80 85 90 95 100
Heart rate before exercise (beats/min)
a Describe the association between heart rate before exercise and heart rate after
exercise in terms of strength, direction and form.
The equation of the least squares line is:
heart rate after exercise = 85.671 + 0.561 × heart rate before exercise
b i Use the equation to predict heart rate after exercise when heart rate before
exercise is 100 beats/minute. Round to the nearest whole number.
ii Are you extrapolating or interpolating?
c The person with a heart rate of 122 beats/minute after exercise had a heartbeat of 76
beats/minute before exercise. If the least squares line is used to predict this person’s
heart rate after exercise, determine the value of the residual. Give your answer
rounded to one decimal place.
d i What assumption about the form of the association can be tested using a residual
plot?
ii Referring to the residual plot, explain why this assumption is satisfied.

Chapter
4
Data transformation
Chapter questions
I What is a squared transformation and when is it used?
I What is a log transformation and when is it used?
I What is a reciprocal transformation and when is it used?
I How do I interpret a least squares line fitted to transformed data?
I How do I use a least squares line fitted to transformed data for
prediction?
I How do I use a residual plot to assess the effectiveness of a data
transformation?
I How do I use the coefficient of determination to assess the effectiveness of
a data transformation?
You may recall from your study of Variation in General Mathematics 12 that
a non-linear association could be transformed into a linear association using
data transformation. The transformations introduced were the squared, log and
reciprocal transformations. In this chapter we consider the effect of each of these
three transformations when applied to one axis only (either x or y, but not both), using
them to linearise scatterplots. This is the first step towards solving problems involving
non-linear associations.

212 Chapter 4 Data transformation
4A The squared transformation

Learning intentions
I To be able to apply a squared transformation to either x or y.
I To be able to fit a least squares regression line to the transformed data.
I To be able to use the least squares regression line fitted to the transformed data for
prediction.
The squared transformation is a stretching transformation. It works by stretching out

the upper end of the scale on either the x- or y-axis. The effects of applying the x2 and y2
transformations (separately) to a scatterplot are illustrated graphically below.
Transformation Outcome Graph

x2 Spreads out the high x-values y
relative to the lower x-values,
leaving the y-values unchanged.
This has the effect of straightening
out curves like the one shown
opposite.
x
2
y Spreads out the high y-values y
relative to the lower y-values,
leaving the x-values unchanged.
This has the effect of straightening
out curves like the one shown
opposite.
The following example shows how the x-squared transformation works in practice.
Example 1 Applying the x-squared transformation
A base jumper leaps from the top of a cliff, 1560 metres above the valley floor. The
scatterplot below shows the height (in metres) of the base jumper above the valley floor
every second, for the first 10 seconds of the jump.

4A The squared transformation 213
A scatterplot shows that there is a strong negative 1600

association between the height of the base jumper 1500
above the ground and time.
Height (metres)
1400
a Apply a squared transformation to the
variable time, and determine the least squares 1300
regression line for the transformed data. 1200
b Use the least squares equation to predict to the
1100
nearest metre the height of the base jumper
after 3.4 seconds. 1000
0 1 2 3 4 5 6 7 8 9 10
Time (seconds)
Solution
a Applying the squared transformation involves changing the scale on the time axis to
time2 .
From the plot opposite we see that the 1600
association between height and time2 is now 1500
linear.
Height (metres)
1400
Now that we have a linearised scatterplot, 1300
we can use a least squares line to model the
1200
association between height and time2 .
1100
The equation of this line is:
height = 1560 − 4.90 × time2 1000
0 10 20 30 40 50 60 70 80 90 100
Time2
b Like any regression line, we can use its equation to make predictions. After 3.4
seconds, we predict that the height of the base jumper is:
height = 1560 − 4.90 × 3.42 = 1503 m (to nearest m)
The next example shows how the y-squared transformation works in practice.
Example 2 Applying the y-squared transformation
In a study of the effectiveness of fertiliser on the yield of strawberry plants, differing

amounts of liquid fertiliser (in mL) were given to groups of plants, and their average yield
(in kg) measured.

A scatterplot shows that there is a strong positive 6

association between the fertiliser and yield.
5
yield(kg)
a Apply a squared transformation to the variable yield,
4
and determine the least squares regression line for the
3
transformed data.
b Use the least squares equation to predict the yield of a 2
0 2 4 6 8 10
plant given 6.5 mL of fertiliser, giving your answer to fertiliser(ml)
1 decimal place.
Solution
a Applying the y-squared transformation involves changing the scale on the y-axis to
yield2 .
From the plot opposite we see that the association 30
between yield2 and fertiliser is now linear.

Now that we have a linearised scatterplot, we can use 20
yield 2
a least squares line to model the association between
10
yield2 and fertiliser.
The equation of this line is: 0
yield2 = 4.45 + 2.29 × f ertiliser 0 2 4 6
fertiliser(ml)
8 10
b Using this equation, when we predict that:

yield2 = 4.45 + 2.29 × 6.5 = 19.34
√
and yield = 19.34 = ±4.4
Looking at the scatterplot we can see that only the positive value of the square root
makes sense, so our prediction is 4.4 kg.
Performing a data transformation is quite computationally intensive, but your CAS

calculator is well suited to the task.
CAS 1: Using the TI-Nspire CAS to perform a squared transformation

The table shows the height (in m) of a base jumper for the first 10 seconds of her jump.
Time 0 1 2 3 4 5 6 7 8 9 10
Height 1560 1555 1540 1516 1482 1438 1383 1320 1246 1163 1070
a Construct a scatterplot displaying height (the RV) against time (the EV).
b Apply an x-squared transformation and fit a least squares line to the transformed data.
c Use the regression line to predict the height of the base jumper after 3.4 seconds.

4A The squared transformation 215
Steps
1 Start a new document by pressing / + N .
Enter the data into lists named time and height, as
shown.
3 Name column C as timesq (short for ‘time squared’).

4 Move the cursor to the formula cell below timesq.
Enter the expression = time^2 by pressing = , then
typing time^2. Pressing · calculates and displays
the values of timesq.

Construct a scatterplot of height against time. Let time
be the explanatory variable and height the response
variable. The plot is clearly non-linear.

Construct a scatterplot of height against time2 .
The plot is now linear.
7 Press b>Analyze>Regression>Show Linear (a +

bx) to plot the line on the scatterplot with its equation.
Note: The x in the equation on the screen corresponds to the
transformed variable time2 .
8 Write down the regression equation in height = 1560 − 4.90 × time2

terms of the variables height and time2 .
9 Substitute 3.4 for time in the equation height = 1560 − 4.90 × 3.42 = 1503 m
to find the height after 3.4 seconds.

CAS 1: Using the CASIO Classpad to perform a squared transforma-

tion
The table shows the height (in m) of a base jumper for the first 10 seconds of her jump.
Time 0 1 2 3 4 5 6 7 8 9 10
Height 1560 1555 1540 1516 1482 1438 1383 1320 1246 1163 1070
a Construct a scatterplot displaying height (the RV) against time (the EV).
b Apply an x-squared transformation and fit a least squares line to the transformed data.
c Use the regression line to predict the height of the base jumper after 3.4 seconds.
Steps
1 In the Statistics application enter the data into lists named
time and height.
2 Name the third list timesq (short for time squared).
3 Place the cursor in the calculation cell at the bottom of the
third column and type time^2. This will calculate the values
of time2 .
Let time be the explanatory variable (x) and height the
response variable (y).
4 Construct a scatterplot of height against time.

StatGraphs dialog box as
shown.
Tap to view the scatterplot.
The plot is clearly non-linear.
5 Construct a scatterplot of height against time2 .

shown.
Tap to view the scatterplot.
The plot is now clearly linear.

4A 4A The squared transformation 217
6 Fit a regression line to the

transformed data.
Go to Calc, Regression, Linear
Reg.
Complete the Set Calculation
dialog box as shown and tap OK.
Note: The ‘x’ in the linear equation
corresponds to the transformed variable
time2 .
Tap OK a second time to plot and
display the regression line on the
scatterplot.
7 Write down the equation in terms of height = 1560 − 4.90 × time2 .
height and time2 .
8 Substitute 3.4 for time in the equation. height = 1560 − 4.90 × 3.42 = 1503 m
Skill-
sheet Exercise 4A
The x-squared transformation: some prerequisite skills

1 Evaluate y in the following expression, rounded to one decimal place.
a y = 7 + 8x2 when x = 1.25 b y = 7 + 3x2 when x = 1.25
c y = 24.56 – 0.47x2 when x = 1.23 d y = –4.75 + 5.95x2 when x = 4.7
The x-squared transformation: calculator exercises

Example 1 2 The scatterplot opposite was constructed from the y
data in the table below. 20
x 0 1 2 3 4
15
y 16 15 12 7 0
10
5
0 x
0 1 2 3 4
a Linearise the scatterplot by applying an x-squared transformation and fit a least
squares line to the transformed data.
b Give its equation.
c Use the equation to predict the value of y when x = –2.

218 Chapter 4 Data transformation 4A
3 The scatterplot opposite was constructed from the y

50
x 1 2 3 4 5
40
y 3 9 19 33 51 30
From the scatterplot, the association between 20
10
y and x is non-linear.
5
0 x
1 2 3 4 5
a Linearise the scatterplot by applying an x-squared transformation and fit a least
b Give its equation.
c Use the equation to predict the value of y when x = 6.
The y-squared transformation: some prerequisite skills

4 Evaluate y in the following expression. Give the answers rounded to one decimal place.
a y2 = 16 + 4x when x = 1.57 b y2 = 1.7 – 3.4x when x = 0.03
c y2 =16 + 2x when x = 10 (y > 0) d y2 = 58 + 2x when x = 3 (y < 0)
The y-squared transformation: calculator exercises

Example 2 5 The scatterplot opposite was constructed from the 6.0
data in the table below. 5.0
x 0 2 4 6 8 10 4.0
y 1.2 2.8 3.7 4.5 5.1 5.7 3.0

2.0
From the scatterplot, the association between y and x
1.0
is non-linear. 0
0 2 4 6 8 10
a Linearise the scatterplot by applying a y-squared transformation and fit a least
b Give its equation. Write the coefficient, rounded to two significant figures.
c Use the equation to predict the value of y when x = 9. Give the answer rounded to
one decimal place.
Applications of the squared transformation

6 The table gives the diameter (in m) of five different umbrellas and the number of
people each umbrella is designed to keep dry. A scatter plot is also shown.

4A 4A The squared transformation 219
Diameter Number 5
Number of people
0.50 1 4
0.70 2 3
0.85 3
2
1.00 4
1
1.10 5
0.2 0.4 0.6 0.8 1 1.2
Diameter (metres)
a Apply the squared transformation to the variable diameter and determine the least
squares regression line for the transformed data. Diameter is the EV.
Write the slope and intercept of this line, rounded to one decimal place, in the
spaces provided.
number = + × diameter2
b Use the equation to predict the number of people who can be sheltered by an
umbrella of diameter 1.3 m. Give your answer rounded to the nearest person.
7 The time (in minutes) taken for a local anaesthetic to take effect is associated with
to the amount administered (in units). To investigate this association a researcher
collected the following data.
Amount 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Time 3.7 3.6 3.4 3.3 3.2 3.0 2.9 2.7 2.5 2.3 2.1
The association between the variables amount and time is non-linear as can be seen
from the scatterplot below. A squared transformation applied to the variable time will
linearise the scatterplot.
a Apply the squared transformation 4
to the variable time and fit a least
3.5
Time (minutes)
squares regression line to the

transformed data. Amount is the EV. 3
Write the equation of this line with 2.5

the slope and intercept rounded to
2
two significant figures.
1.5
b Use the equation to predict the time
for the anaesthetic to take effect when 0.4 0.6 0.8 1 1.2 1.4 1.6
Amount (units)
the dose is 0.4 units. Give the answer

220 Chapter 4 Data transformation 4A
The following information relates to Questions 8 and 9
A student uses the data in the table below to construct the scatterplot shown:
x y 300
1 264 250
2 239 200
3 234 y 150
4 208 100
5 182 50
6 164
0
7 98 0 1 2 3 4 5 6 7 8
x
8 49
8 A squared transformation is applied to x to linearise the association. A least squares

line is fitted to the transformed data, with x2 as the explanatory variable.
The equation of this least squares line is closest to
A y = 310 − 29.1x2
B y = 10.2 − 0.032x2
C y = 1.80 − 0.106x2
D y = 263 − 3.26x2
E y = 79.9 − 0.303x2
9 A y2 transformation could also be used to linearise this association. A least squares line
is fitted to the transformed data, with y2 as the response variable, and the equation of
the least squares line is
y2 = 79973 − 9533.4x
Using this equation, the value of y when x = 4 is closest to:
A 205 B 208 C 247 D 531 E 41839
10 The association between the cost of a certain precious stone (in $) and its weight (in
mg) is non-linear. A squared transformation was applied to the explanatory variable
weight, and a least squares line fitted to the transformed data. The equation of the least
squares line is:
cost = 2370 + 0.238 × weight2
Using this equation, the cost of a precious stone weighing 75mg is closest to:
A $2389 B $3709 C $2689 D $7995 E $177,768

4B The log transformation 221
4B The log transformation

Learning intentions
I To be able to apply a log10 transformation to either x or y.
prediction.
You will recall from Chapter 1 that the shape of a highly skewed single variable distribution
could be changed to become more symmetric by changing the scale from x to log10 x.
When applied to bivariate data, the effect of the logarithmic transformation is to again
to compress the upper end of the scale on either the x- or the y-axis, potentially linearising
a non-linear association. The effect of applying the log10 x and log10 y transformations
(separately) to a scatterplot are illustrated graphically below.

log10 x Compresses the higher x-values y
relative to the lower x-values, leaving
the y-values unchanged.
This has the effect of straightening out
curves like the one shown.
x
log10 y Compresses larger y values relative to y
the smaller y values
This has the effect of straightening out
curves like the one shown.
Following the normal convention, log x means log10 x.
Example 3 Applying the log x transformation
The general wealth of a country, often measured by its Gross Domestic Product (GDP),
is one of several variables associated with lifespan in different countries. However, the

association is not linear, as can be seen in the scatterplot below which plots lifespan (in
years) against GDP per person (in dollars) for 13 different countries.
The scatterplot shows that there is a strong positive 83
association between the lifespan and GDP. 81
79
Lifespan (years)
a Apply a log transformation to the variable GDP, 77
and determine the least squares regression line 75
for the transformed data. 73
71
b Use the least squares equation to predict the
69
lifespan of a country with a GDP of $20 000 67
per person, giving your answer rounded to one
decimal place. 0 10000 20000 30000 40000 50000
GDP
Solution
a Applying the log x transformation involves changing the scale on the x-axis to
log(GDP).
When we make this change, we see that the 83
association between the variables lifespan and 81
Lifespan (years) 79
log (GDP) is linear. See the plot opposite.
77
Note: On the plot, when log (GDP) = 4, the actual GDP is
75
104 or $10 000.
73
We can now fit a least squares line to model the 71
association between the variables lifespan and 69
log(GDP). 67
2.5 3 3.5 4 4.5

log (GDP)
lifespan = 54.3 + 5.59 × log(GDP)
b Using this equation, for a country with a GDP of $20 000, the lifespan is predicted as:
lifespan= 54.3 + 5.59 × log 20 000 = 78.3 years (to one decimal place)
Example 4 Applying the log y transformation
The numbers of cases of a very infectious disease were recorded over a 12 day period.
The association is not linear, as can be seen in the scatterplot below which plots cases
against days.

The scatterplot shows that there is a strong 7000

positive association between the number of 6000
case and day. 5000
cases
a Apply a log transformation to the variable 4000
cases, and determine the least squares 3000
1000
b Use the least squares equation to predict the
0
cases on day 13. 0 2 4 6 8 10 12
day
Solution
a Applying the log y transformation involves changing the scale on the y-axis to
log(cases).
When we make this change, we see from the 4
plot the association between the variables
log(cases) and day is linear.
log(cases)
3
Note: On the plot, when log (cases) =3, the actual
number of cases is 103 or 1000.
We can now fit a least squares line to 2
model the association between the variables
log(cases) and day. 1
0 2 4 6 8 10 12
day
log (cases)= 1.046 + 0.227 × day
b Using this equation, on day 13 the number of cases is predicted as:
log (cases)= 1.046 + 0.227 × 13 = 3.997
To find the number of cases we use the calculator to evaluate 103.997 = 9931 cases (to
the nearest whole number).

CAS 2: Using the TI-Nspire CAS to perform a log transformation

The table shows the lifespan (in years) and GDP (in dollars)
Lifespan GDP
of people in 12 countries. The association is non-linear.
80.4 36 032
Using the log x transformation:
linearise the data, and fit a regression line to the 79.8 34 484
transformed data (GDP is the EV) 79.2 26 664
write its equation in terms of the variables lifespan and 77.4 41 890
GDP rounded to three significant figures. 78.8 26 893
use the equation of the regression line to predict the 81.5 25 592
lifespan in a country with a GDP of $20 000, rounded to 74.9 7 454
one decimal place.
72.0 1 713
77.9 7 073
70.3 1 192
73.0 631
68.6 1 302
Steps
1 Start a new document by pressing / + N .
Enter the data into lists named lifespan and gdp.
3 Name column C as lgdp (short for log (GDP)).
Now calculate the values of log (GDP) and store
them in the list named lgdp.
4 Move the cursor to the formula cell below the lgdp
heading.
We need to enter the expression = log(gdp).
To do this, press = then type in log(gdp). Pressing
· calculates and displays the values of lgdp.

Construct a scatterplot of lifespan against GDP. Let
GDP be the explanatory variable and lifespan the
response variable. The plot is clearly non-linear.


Construct a scatterplot of lifespan against log(GDP).
The plot is now clearly linear.
7 Press b>Analyze>Regression>Show Linear (a +

bx) to plot the line on the scatterplot with its equation.
Note: The x in the equation on the screen corresponds to the
transformed variable log (GDP).
8 Write the regression equation in lifespan = 54.3 + 5.59 × log (GDP)

terms of the variables lifespan and
log (GDP).
9 Substitute 20 000 for GDP in the lifespan = 54.3 + 5.59 × log 20 000
equation to find the lifespan of people = 78.3 years
in a country with GDP of $20 000.
CAS 2: Using the CASIO Classpad to perform a log transformation

The table shows the lifespan (in years) and GDP (in dollars)
Lifespan GDP
of people in 12 countries. The association is non-linear.
80.4 36 032
Using the log x transformation:
linearise the data, and fit a regression line to the 79.8 34 484
transformed data (GDP is the EV) 79.2 26 664
write its equation in terms of the variables lifespan and 77.4 41 890
GDP rounded to three significant figures. 78.8 26 893
use the equation to predict the lifespan in a country with a 81.5 25 592
GDP of $20 000 rounded to one decimal place. 74.9 7 454
72.0 1 713
77.9 7 073
70.3 1 192
73.0 631
68.6 1 302

Steps
1 In the Statistics application enter the data into lists named
Lifespan and GDP.
2 Name the third list logGDP.
third column and type log (GDP).
Let GDP be the explanatory variable (x) and lifespan the
4 Construct a scatterplot of lifespan

against log (GDP).
shown.
Tap to view the
scatterplot.
The plot is linear.
5 To find the least squares

regression equation and fit a
regression line to the transformed
data.
Go to Calc, Regression,
Linear Reg.
Complete the Set
Calculation dialog box
as shown and tap OK.
This generates the
regression results.
Note: The x in the linear equation
corresponds to the transformed
variable log (GDP).
Tap OK a second time to plot and display the regression line on the scatterplot.
6 Write the equation in terms of lifespan lifespan = 54.3 + 5.59 × log (GDP)
and log (GDP).
7 Substitute 20 000 for GDP in the lifespan = 54.3 + 5.59 × log 20 000
equation. = 78.3 years

4B 4B The log transformation 227
Skill-
sheet Exercise 4B
The log x transformation: some prerequisite skills

1 Evaluate the following expressions rounded to one decimal place.
a y = 5.5 + 3.1 log 2.3 b y = 0.34 + 5.2 log 1.4
c y = –8.5 + 4.12 log 20 d y = 196.1 – 23.2 log 303
The log x transformation: calculator exercise

9
x 5 10 150 500 1000 8
7
y 3.1 4.0 7.5 9.1 10.0 6
5
From the scatterplot, it is clear that the association 4
between y and x is non-linear. 3
0 x
0 200 400 600 800 1000
a Linearise the scatterplot by applying a log x transformation and fit a least squares
line to the transformed data.
b Write down the equation, with the coefficients rounded to one significant figure.
3 The scatterplot opposite was constructed from the y

14
x 10 44 132 436 981
12
y 15.0 11.8 9.4 6.8 5.0 10
8
From the scatterplot, it is clear that the association
6
between y and x is non-linear.
4 x
0 200 400 600 800 1000
a Linearise the scatterplot by applying a log x transformation and fit a least squares
line to the transformed data.
The log y transformation: some prerequisite skills

4 Find the value of y in the following, rounded to one decimal place if not exact.
a log y = 2 b log y = 2.34
c log y = 3.5 + 2x where x = 1.25 d log y = –0.5 + 0.024x where x = 17.3

228 Chapter 4 Data transformation 4B
The log y transformation: calculator exercise

x 0.1 0.2 0.3 0.4 0.5 80
y 15.8 25.1 39.8 63.1 100.0 60

40
From the scatterplot, it is clear that the
20
association between y and x is non-linear.
0
a Linearise the scatterplot by applying a log y x
0.1 0.2 0.3 0.4 0.5
transformation and fit a least squares line to
the transformed data.
c Use the equation to predict the value of y when x = 0.6, rounded to one decimal place.
Applications of the log transformation

6 The table below shows the level of performance level achieved by 10 people on
completion of a task. Also shown is the time spent (in minutes) practising the task. In
this situation, time is the EV. The association between the level and time is non-linear
as seen in the scatterplot.
Time Level 4
0.5 1 3.5
3
1 1.5
Level
2.5
1.5 2 2
2 3 1.5
3 3 1
0.5
4 3.5 0
5 4 0 1 2 3 4 5 6 7
Time (minutes)
6 3.5
7 3.9
7 3.6
A log transformation can be applied to the variable time to linearise the scatterplot.
a Apply the log transformation to the variable time and fit a least squares line to the
transformed data. log (time) is the EV.
Write the slope and intercept of this line, rounded to two significant figures, in the
spaces provided.
level = + × log (time)
b Use the equation to predict the level of performance (rounded to one decimal place)
for a person who spends 2.5 minutes practising the task.

4B 4B The log transformation 229
7 The table below shows the number of internet users signing up with a new internet
service provider for each of the first nine months of their first year of operation.
A scatterplot of the data is also shown.
Month Number 120
100
1 24
Number
80
2 32 60
40
3 35 20
4 44 0
0 1 2 3 4 5 6 7 8 9
5 60 Month
6 61
7 78
8 92
9 118
The association between number and month is non-linear.
a Apply the log transformation to the variable number and fit a least squares line to
the transformed data. Month is the EV.
Write the slope and intercept of this line, rounded to four significant figures, in the
spaces provided.
log (number) = + × month
b Use the equation to predict the number of internet users after 10 months. Give
answer to the nearest whole number.

8 A student uses the data in the table below to construct the scatterplot shown.
x y
6.0
5 3.45
44 4.41
5.0
94 4.64 y
187 5.03
4.0
791 5.65
1350 5.81 3.0
1960 5.97 0 500 1000 1500 2000 2500
x
2345 6.06
A log transformation is applied to x to linearise the association. A least squares line is

fitted to the transformed data, with log x as the explanatory variable.

230 Chapter 4 Data transformation 4B

A y = 4.43 + 0.001 log x
B y = −3570 + 861 log x
C y = 2.78 + 0.976 log x
D y = −2.85 + 1.03 log x
E y = 0.642 + 0.00724 log x
9 The association between the power of a car (in horsepower) and the time it takes to
accelerate from 0 to 100 km/hr (in seconds) is non-linear. A log transformation was
applied to the explanatory variable horsepower, and a least squares line fitted to the
transformed data. The equation of the least squares line is:
time = 42.7 − 13.9 × log(horsepower)
Using this equation, the time it would take for a car with 180 horsepower to accelerate
from 0 to 100km/hr is closest to:
A 29.5 seconds B 65 seconds C 28.8 seconds D 11.3 seconds E 11.4 seconds
10 The price of shares in a newly formed technology company price has increased
non-linearly since the company was formed 12 months ago. A log transformation was
applied to the maximum share price each month (share price), and a least squares line
fitted to the transformed data, with month as the explanatory variable. The equation of
the least squares line is:
log(shareprice) = 1.39 + 0.050 × month
Using this equation, the maximum monthly share price in month 14 is closest to:
A $123.03 B $2.09 C $8.08 D $20.16 E $25.25
4C The reciprocal transformation

Learning intentions
I To be able to apply a reciprocal transformation to either x or y.
prediction.
The reciprocal transformation is a stretching transformation that compresses the upper end
of the scale on either the x- or y-axis.

4C The reciprocal transformation 231
The effect of applying a reciprocal y transformation to a scatterplot is as follows:

1 Compresses larger x values relative y
x to the smaller data values, but to
a greater extent than log x. This
has the effect of straightening out
curves like the one shown opposite.
Note that values of x less than one
become greater than 1, and values
of x greater than 1 become less x
than 1, so that the order of the data
values is reversed.
1 Compresses larger values of y rel- y

y ative to lower values of y. This
has the effect of straightening out
curves like the one shown oppo-
site. Again, the order of the data is
reversed.
The following example shows how the 1/x transformation works in practice.
Example 5 Applying the reciprocal (1/x) transformation
After embarking on a new healthy eating and exercise plan, Ben recorded his weekly
weight loss over a 10 week. The association is not linear, as can be seen in the scatterplot
below which plots weekly weight loss in kg against length of diet in weeks.
The scatterplot shows that there is a strong negative 6
weekly weight loss (kg)
association between weekly weight loss and length 5

of diet. 4
a Apply a reciprocal transformation to the variable 3
length of diet, and determine the least squares 2
b Use the least squares equation to predict the 0
0 2 4 6 8 10
weekly weight loss in week 11, giving your length of diet (weeks)
answer to one decimal place.

Solution
a Applying the 1/x transformation involves changing the scale on the x-axis to 1/(length
of diet).
When we make this change, we see from 6
weekly weight loss (kg)

the plot the association between the 5
variables weekly weight loss and 1/(length
4
of diet) is linear.
3
We can now fit a least squares line to
model the association between the variables 2
weekly weight loss and 1/(length of diet). 1
0
0 0.2 0.4 0.6 0.8 1.0
1/(length of diet)
5.09
weekly weight loss= 0.13 +
length of diet
b Using this equation, in week 11 the weight loss predicted is:
5.09
weekly weight loss= 0.13 + = 0.6 kg
11
The following example shows how the 1/y transformation works in practice.
Example 6 Applying the reciprocal (1/y) transformation
A homeware company makes rectangular sticky 3.5

Width (cm)
labels with a variety of lengths and widths. 3

2.5
The scatterplot opposite displays the width (in cm) 2
and length (in cm) of eight of the sticky labels. 1.5
3.5 4 4.5 5 5.5 6 6.5 7
Length (cm)
The scatterplot shows that there is a strong negative association between the width of the
sticky labels and their lengths, but it is clearly non-linear.
a Apply a reciprocal transformation to the variable width, and determine the least
squares regression line for the transformed data.
b Use the least squares equation to predict the width of a sticky label which is 5 cm long,
giving your answer to two decimal places.

Solution
a Applying the 1/y transformation involves changing the scale on the y-axis from width
to 1/(width).
When we make this change, we see from the 0.6
scatterplot that the association between 1/width 0.55
and length is linear. 0.5
1/width
We can now fit a least squares line to model the 0.45
0.4
association between 1/width and length.
0.35
0.3
1/width = 0.015 + 0.086 × length 0.25
0.2
3.5 4 4.5 5 5.5 6 6.5 7
Length (cm)
b For a sticky label of length 5 cm, we would predict that:
1/width = 0.015 + 0.086 × 5 = 0.445

1
or width = = 2.25 cm
0.445
CAS 3: Using the TI-Nspire CAS to perform a reciprocal transforma-

tion
The table shows the length (in cm) and width (in cm) of eight sizes of sticky labels.
Length 6.8 5.6 4.6 4.2 3.5 4.0 5.0 5.5

Width 1.8 2.0 2.5 3.0 3.5 2.6 2.0 1.9
Using the 1/y transformation:
linearise the data, and fit a regression line to the transformed data (length is the EV)
write its equation in terms of the variables length and width
use the equation to predict the width of a sticky label with a length of 5 cm.
Steps
Enter the data into lists named length and width.
3 Name column C as recipwidth (short for 1/width).
Calculate the values of recipwidth.
Move the cursor to the formula cell below the
recipwidth heading. Type in =1/width. Press ·
to calculate the values of recipwidth.


Construct a scatterplot of width against length.
Let length be the explanatory variable and
width the response variable. The plot is clearly
non-linear.

Construct a scatterplot of recipwidth (1/width)
against length. The plot is now clearly linear.
6 Press b>Analyze>Regression>Show Linear

(a + bx) to plot the line on the scatterplot with its
equation.
Note: The y in the equation on the screen corresponds to
the transformed variable 1/width.
7 Write down the regression equation in 1/width = 0.015 + 0.086 × length

terms of the variables width and length.
8 Substitute 5 cm for length in the 1/width = 0.015 + 0.086 × 5 = 0.445
equation. Thus width = 1/0.445 = 2.25 cm (to 2 d.p.)
CAS 3: Using the CASIO Classpad to perform a reciprocal transforma-

tion
The table shows the length (in cm) and width (in cm) of eight sizes sticky labels.
Length 6.8 5.6 4.6 4.2 3.5 4.0 5.0 5.5

Width 1.8 2.0 2.5 3.0 3.5 2.6 2.0 1.9
Using the 1/y transformation:
linearise the data, and fit a regression line to the transformed data. Length is the EV.
write its equation in terms of the variables length and width.
use the equation to predict the width of a sticky label with length of 5 cm.

Steps
1 Open the Statistics application and enter the data into lists
named length and width.
2 Name the third list recwidth (short for reciprocal width).
third column and type 1/width. This will calculate all the
reciprocal values of the width.
Let length be the explanatory variable (x) and width the
4 Construct a scatterplot of
1/width against length.
Tap and complete
the Set StatGraphs
dialog box as shown.
Tap to view the
scatterplot.
The plot is now clearly
linear.
5 Fit a regression line to the
transformed data.
Go to Calc, Regression,
Linear Reg.
Complete the Set
Calculation dialog box as
shown and tap OK.
This generates the
regression results.
Note: The y in the linear
equation corresponds to the
transformed variable 1/width;
that is 1/y.
Tap OK a second time to plot and display the line on the scatterplot.
6 Write down the equation in terms of 1/width = 0.015 + 0.086 × length

the variables width and length.
7 Substitute 5 cm for length in the 1/width = 0.015 + 0.086 × 5 = 0.445
equation. Thus width = 1/0.445 = 2.25 cm (to 2 d.p.)

236 Chapter 4 Data transformation 4C
Skill-
sheet Exercise 4C
The reciprocal (1/x) transformation: some prerequisite skills

1 Evaluate the following expressions rounded to one decimal place.
22 2.3
a y=6+ when x = 3 b y = 4.9 − when x = 1.1
x x
7.95 223.5
c y = 8.97 − when x = 1.97 d y = 102.6 + when x = 1.08
x x
The reciprocal (1/x) transformation: calculator exercise

Example 5 2 The scatterplot opposite was constructed from the data in the y
table below. 60
x 2 4 6 8 10 40
y 60 30 20 15 12
20
From the scatterplot, it is clear that the association between
0 x
y and x is non-linear. 0 2 4 6 8 10
a Linearise the scatterplot by applying a 1/x transformation and fit a least squares line
to the transformed data.
b Write down its equation.
The reciprocal (1/y) transformation: some prerequisite skills

3 Find the value of y in the following, rounded to two decimal places.
1 1
a = 3x when x = 2 b = 6 + 2x when x = 4
y y
1 1
c = –4.5 + 2.4x when x = 4.5 d = 14.7 + 0.23x when x = 4.5
y y
The reciprocal (1/y) transformation: calculator exercise

Example 6 4 The scatterplot opposite was constructed from the data in the y
table below. 1
x 1 2 3 4 5
y 1 0.5 0.33 0.25 0.20 0.5
From the scatterplot, it is clear that the association between y

0 x
and x is non-linear. 0 1 2 3 4 5
a Linearise the scatterplot by applying a 1/y transformation and fit a least squares line
to the transformed data.
b Write down its equation.
c Use the equation to predict the value of y when x = 0.25.

4C 4C The reciprocal transformation 237
Applications of the reciprocal transformation

5 The table shows the horsepower of 10 cars and their fuel consumption. From the
scatterplot, it is clear that the association between horsepower and fuel consumption is
non-linear.
160
Fuel consumption Horsepower
150
5.2 155 140
130
Horsepower
7.3 125 120
12.6 75 110
100
7.1 110 90
6.3 138 80
70
10.1 88 60
10.5 80 4 6 8 10 12 14 16
Fuel consumption
14.6 70 (km/litre)
10.9 100
7.7 103
a Apply the reciprocal transformation to the variable fuel consumption and fit a least
squares line to the transformed data. Horsepower is the RV.
Write the intercept and slope of this line in the boxes provided, rounded to three
1
horsepower = + ×
f uel consumption
b Use the equation to predict the horsepower of a car with a fuel consumption of
9 km/litre.
6 Ten students were given an opportunity to practise a complex matching task as often as
they liked before they were assessed. The number of times they practised the task and
the number of errors they made when assessed are given in the table.
16
Times Errors
14
1 14 12
2 9 10
Errors
2 11 8
6
4 5
4
5 4 2
6 4 0
0 2 4 6 8 10
7 3 Times
7 3
9 2

238 Chapter 4 Data transformation 4C
a Apply the reciprocal transformation to the variable errors and determine the least
squares regression with the number of times the task was practiced as the EV.
Write the intercept and slope of this line in the boxes provided, rounded to two
1
= + × times
errors
b Use the equation to predict the number of errors made when the task is practised
six times.

7 A student used the data in the table below to construct the scatterplot shown
x y .60
8 0.58 .50
11 0.43 .40
y
14 0.39 .30
22 0.24 .20
26 0.19 .10
35 0.13
.00
41 0.10 0 5 10 15 20 25 30 35 40 45 50
x
50 0.12
A reciprocal transformation is applied to y to linearise the association. A least squares

line is fitted to the transformed data, with 1/y as the response variable. The equation of
this least squares line is closest to
1
A = 0.198 + 0.196x
y
1
B = −0.546 − 0.011x
y
4.665
C y = 0.013 +
x
1
D = 0.013 + 4.665x
y
E y = 0.546 − 0.011x
8 The association between score on a problem solving test (score) and the number of
attempts a person has at the test (attempts) is non-linear. A reciprocal transformation
was applied to the explanatory variable attempts, and a least squares line fitted to the
transformed data. The equation of the least squares line is:
1
score = 50 − 22.8 ×
attempts

4C 4D Choosing and applying the appropriate transformation 239
Using this equation, the score that a person achieves on their fourth attempt is closest
to:
A 6.8 B 27.2 C 55.7 D 41.2 E 44.3
9 The price of shares in a newly formed technology company price has increased non-
linearly since the company was formed 12 months ago. A reciprocal transformation
was applied to the maximum share price each month (share price), and a least squares
line fitted to the transformed data, with month as the explanatory variable. The
equation of the least squares line is:
1
= 0.0349 − 0.00215 × month
shareprice
A $2.18 B $28.78 C $208.33 D 0.48 cents E 48 cents
4D Choosing and applying the appropriate

transformation
Learning intentions
I To be able to use the circle of transformations to determine which transformations may
help linearise a non-linear association.
I To be able to use a residual plot to assess the effectiveness of a data transformation.
I To be able to use the coefficient of determination to assess the effectiveness of a data
transformation.
The types of scatterplots that can be transformed by the squared, log or reciprocal
transformations can be fitted together into what we call the circle of transformations.
The circle of transformations

Possible transformations Possible transformations
y2
log x y2
x2
1
x
log y log y
1 1
y y
log x
x2
1
x

The purpose of the circle of transformations is to guide us in our choice of transformation to

linearise a given scatterplot.
There are two things to note when using the circle of transformations:
1 In each case, there is more than one type of transformation that might work.
2 These transformations only apply to scatterplots with a consistently increasing or
decreasing trend.
The advantage of having alternatives is that in practice, we can always try each of them to
see which gives us the best result. How do we decide which transformation is the best? The
best transformation is the one which results in the best linear model. To choose the best
linear model we will consider for each transformation applied:
The residual plot, in order to evaluate the linearity of the transformed association.
The value of the coefficient of determination, r2 .
This procedure is illustrated in the following example.
Example 7 Choosing the best transformation
The scatterplot shows the age (in years) and 40

Age of tree (years)
diameter at a height of 1.5 metre (in cm) for

a sample of 19 trees of the same species. 30
Use an appropriate transformation to find
20
a regression model which allows the age of
this species of tree to be predicted from its 10
diameter.
0
0 5 10 15 20
Diameter (cm)
Solution
The scatterplot has a consistently increasing trend so the circle of transformations applies.
Comparing the scatterplot to those in the circle of transformations we see that the x2 , 1/y
and log x transformations all have the potential to linearise this scatterplot. All of these
transformations have been applied in turn, and the resulting scatterplots and residual plots
are shown in the following table.

4D Choosing and applying the appropriate transformation 241
Transformation Scatterplot Residual plot

40 6
x2
Residual (x squared)
Age of tree (years)
4
30 2
0
20 –2
–4
10
–6
0 –8
0 100 200 300 400 0 100 200 300 400
Diameter squared Diameter squared
1 .3
y 0.8
Residual (1/y)
.2 0.6
0.4
1/(age)
.2
0.2
.1 0.0
–.02
.1
–.04
.0 –.06
0 5 10 15 20 0 5 10 15 20
Diameter (cm) Diameter (cm)
log y 1.8 .2
1.6
Residual(log y)
1.4 .1
log(age)
1.2
.0
1.0
.8 –.1
.6
.4 –.2
0 5 10 15 20 0 5 10 15 20
Diameter (cm) Diameter (cm)
Applying each of these transformations in turn we can see from the residual plots that
both the x2 and the log y transformations have been quite effective in linearising the
association between the age of the tree and its diameter. There still seems to be a curve in
the residual plot after the the 1/y transformation so that has been less effective.
To further help to choose the best transformation we can compare the values of r2 , the
coefficient of determination.
For the x2 transformation, r2 = 92.7%
For the 1/y transformation, r2 = 75.7%
For the log y transformation, r2 = 90.2%
Both the x2 and log y transformations have a very high explanatory power, and
either would seem to be acceptable. When more than one transformation is doing a

242 Chapter 4 Data transformation 4D
reasonable job of linearising the association, and they have similar value of r2 then the
transformation which is easier to interpret in terms of the variables is preferred. In this
case diameter2 makes more sense in that it tells us that the age of the tree relates to
the cross sectional area of the tree. The log transformation does not have an equivalent
meaningful interpretation.
We can now fit a least squares line to model the association between age and diameter2
age = 5.098 + 0.091 × diameter2
At this stage you might find it helpful to use the interactive ‘Data transformation’ (accessible
through the Interactive Textbook) to see how these different transformations can be used to
linearise scatterplots.
Exercise 4D
Example 7 1 The scatterplots below are non-linear. For each, identify the transformations x2 , log x,
1/x, y2 , log y, 1/y or none that might be used to linearise the plot.
a 5 b 5
4 4
3 3
2 2
1 1
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
c 5 d 5
4 4
3 3
2 2
1 1
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10

4D 4D Choosing and applying the appropriate transformation 243
2 The data below gives the yield in kilograms and length in metres of 12 commercial
potato plots.
yield(kg) 346 1798 152 86 436 968

length(m) 12.1 27.4 8.3 5.5 15.7 21.5
yield(kg) 686 257 2435 287 1850 1320
length(m) 19.5 9.0 34.2 14.7 31.9 25.3
a Construct a scatterplot showing the association between yield in kilograms (the RV)
and length of the plot in metres (the EV).
b Fit a least squares regression line to the data. Write down the equation in terms of
the variables in the question, giving the values of the intercept and slope rounded to
four significant figures.
c Construct a residual plot, and comment on whether the linearity assumption has
been met.
d Use the circle of transformations to select which transformations could be
considered in order to linearise the association.
e Using an appropriate transformation, recommend a regression model for the
association between yield and length of the plot. Write down the equation in terms
of the transformed variables, giving the values of the intercept and slope rounded to
f What is the value of r2 for the recommended model? Give your answer as a
percentage rounded to one decimal place.
3 In order to investigate the association between the average number of cigarettes per day
per smoker (smoking) and the cost of cigarettes in $ per cigarette (cost) for a group of
countries the following data was collected.
cost ($ ) 0.67 0.75 0.80 0.92 1.00 1.08 1.17 1.25 1.30 1.40
smoking 16.7 15.5 14.8 13.4 12.5 12.0 11.1 10.9 10.3 9.5
a Construct a scatterplot showing the association between smoking (in cigarettes/day)
(the RV) and cost ($/cigarette) (the EV).
c Construct a residual plot, and comment on whether the linearity assumption has
been met.
association between smoking and cost. Write down the equation in terms of the

244 Chapter 4 Data transformation 4D
transformed variables, giving the values of the intercept and slope rounded to four
4 The following data shows the population density in people per hectare (density) and the
distance from the centre of the city in km (distance) for a large city.
density 307.58 294.67 283.93 270.82 234.93 175.08 101.56 49.80
distance 0 2 4 6 8 10 12 14
a Construct a scatterplot showing the association between the population density in

people per hectare (the RV) and distance from the centre of the city in km (the EV).
c Construct a residual plot, and comment on the whether linearity assumption has
been met.
association between density and distance. Write down the equation in terms of the
transformed variables, giving the values of the intercept and slope rounded to four

Review
Data In regression analysis, data transformation involves changing the scale

transformation on either the x- or y axis in order to linearise an association prior to
Assign-
ment
fitting a least squares line.
Squared The squared transformation stretches out the upper end of the scale on
transformation an axis.
Logarithmic The logarithmic transformation compresses the upper end of the scale
transformation on an axis.
Reciprocal The reciprocal transformation compresses the upper end of the scale
transformation on an axis but to a greater extent than the log transformation.
Residual plots Residual plots are used to assess the effectiveness of a data transforma-
tion.
Coefficient of The transformation that results in a linear association (as assessed by

determination the residual plot) and which has the highest value of the coefficient of
determination is generally the preferred transformation.
The circle of The circle of transformations provides guidance in choosing

transformations the transformations that can be used to linearise various types of
scatterplots.
Skills checklist
list
4A 1 I can apply the x2 transformation.

4A 2 I can apply the y2 transformation.

4B 3 I can apply the log x transformation.

4B 4 I can apply the log y transformation.

Review 246 Chapter 4 Data transformation
4C 5 I can apply the reciprocal (1/x) transformation.

4C 6 I can apply the reciprocal (1/y) transformation.

4D 7 I can choose the best transformation to apply.
1 Select the statement that correctly completes the sentence:
‘The effect of a squared transformation is to . . . ’
A stretch the high values in the data B maintain the distance between values
C stretch the low values in the data D compress the high values in the data
E reverse the order of the data values

‘The effect of a log transformation is to . . .’
E maintain the order of the values in the data

‘The effect of a reciprocal transformation is to . . .’
E reverse the order of the values in the data
4 The association between two variables y and x, as y

shown in the scatterplot, is non-linear. In an attempt to 5
transform the association to linearity, a student would be 4
advised to: 3
2
A leave out the first four points 1
B use a y2 transformation 0 x
0 1 2 3 4 5 6 7 8 9 10
C use a log y transformation
D use a 1/y transformation
E use a least squares regression line

Review
shown in the scatterplot, is non-linear. 5
Which of the following sets of transformations could 4
possibly linearise this association? 3
2
A log y, 1/y, log x, 1/x B y2 , x2 1
C y2 , log x, 1/x D log y , 1/y, x2 0 x
0 1 2 3 4 5 6 7 8 9 10
E ax + b

shown in the scatterplot, is non-linear. 5
Which of the following transformations is most likely 4
to linearise the association? 3
2
A a 1/x transformation B a y2 transformation 1
C a log y transformation D a 1/y transformation 0 x
0 1 2 3 4 5 6 7 8 9 10
E a log x transformation
7 The following data were collected for two related variables x and y.
x 1 2 3 4 5 6 7 8 9 10 11
y 7 8.6 8.9 8.8 9.9 9.7 10.4 10.5 10.7 11.2 11.1
A scatterplot indicates a non-linear association. The data is linearised using a log
x transformation and a least squares line is then fitted. The equation of this line is
closest to:
A y = 7.52 + 0.37 log x B y = 0.37 + 7.52 log x
C y = −1.71 + 0.25 log x D y = 3.86 + 7.04 log x
E y = 7.04 + 3.86 log x
8 A student uses the data in the table below to construct the scatterplot shown
x y 300000
1 2030 250000
2 1265
3 8265 200000
4 5654
y 150000
5 6893
6 43265 100000
7 67890
8 87803 50000
9 113062
0
10 286370
0 1 2 3 4 5 6 7 8 9 10
x

A log transformation is applied to y to linearise the association. A least squares line is

fitted to the transformed data, with log y as the response variable.
A y = 2.88 + 0.256 log x
B log y = −10.1 + 3.64 log x
C log y = −69800 + 24000x
D log y = 2.88 + 0.256x
E log y = −4.84 + 4.38x
9 The association between the total weight of produce picked from a vegetable garden
and its width is non-linear. An x2 transformation is used to linearise the data.
When a least squares line is fitted to the data, its y-intercept is 10 and its slope is 5.
Assuming that weight is the response variable, the equation of this line is:
A (weight)2 = 10 + 5 × width B width = 10 + 5 × (weight)2
C width = 5 + 10 × (weight)2 D weight = 10 + 5 × (width)2
E (weight)2 = 5 + 10 × weight
10 A model that describes the association between the hours spent studying for an exam
and the mark achieved is:
mark = 20 + 40 × log (hours)
From this model, we would predict that a student who studies for 20 hours would score
a mark (to the nearest whole number) of:
A 80 B 78 C 180 D 72 E 140
11 A 1/y transformation is used to linearise a scatterplot.

The equation of a least squares line fitted to this data is:
1/y = 0.14 + 0.045x
This regression line predicts that, when x = 6, y is closest to:
A 0.17 B 0.27 C 0.41 D 2.4 E 3.7

Review
1 The table below shows the age in years (age) and the length in metres (length), for a
group of 18 dugongs. A scatterplot of the data is also shown.
14
age length age length
12
1.0 1.80 8.0 2.47
10
age (years)
1.5 1.85 8.5 2.19
1.5 1.87 9.0 2.26 8
1.5 1.77 9.5 2.40 6
2.5 2.02 9.5 2.39 4
4.0 2.27 10.0 2.41
2
5.0 2.15 12.0 2.50
0
5.0 2.26 12.0 2.32 1.6 1.8 2.0 2.2 2.4 2.6
7.0 2.35 13.0 2.43 length (metres)
!
1
The scatterplot shows that the association is clearly non-linear. A reciprocal
y
transformation can be applied to the variable age to linearise the association.
!
1
a Apply the reciprocal transformation to the data and use the transformed data
y
1
to determine the equation of a least squares line that enables to be predicted
age
from length. Write the values of the intercept and slope in the the appropriate boxes
provided. Round to four significant figures.
1
= − × length
age
b The association can also be linearised by applying a log transformation to the
variable age. When this is done, and a least squares line fitted to the transformed
data, the resulting equation is:
log(age)= −2.443 + 1.429×length
Use this equation to predict the age of a dugong with a length of 2.00 metres. Round
the answer to one decimal place.
2 The table below shows the percentage of people who can read (literacy rate) and the
gross domestic product (GDP), in dollars/person, for a selection of 14 countries. A
scatterplot of the data is also shown.
The scatterplot can be linearised by using a log x transformation.

100
GDP literacy rate
2677 72 80
Literacy rate (%)

260 35
19 904 97 60
122 24
40
18 944 99
4 500 99 20
17 539 99
1 030 73 0
0 5 10 15 20 25
19 860 99
GDP ($’000)
409 40
406 35
6651 62
22 384 99
2 436 64
a Apply the log transformation to the variable GDP, and fit a least squares line to the
transformed data. Write down its equation terms of the variables literacy rate and
log (GDP). Give the slope and intercept rounded to three significant figures.
b Verify that the log transformation has linearised the association by constructing a
residual plot.
c Use the regression equation to predict the literacy rate of a country with a GDP of
$10 000 to the nearest percent.
d Find the value of the residual when the regression equation is used to predict the
literacy rate when the GDP is equal to $19 860. Give your answer rounded to two
3 Measurements of the distance travelled (metres) and time taken (seconds) were made
on a falling body. The data are given in the table below.
time 0 1 2 3 4 5 6
distance 0 5.2 18.0 42.0 79.0 128.0 168.0
a Construct a scatterplot of the data and comment on its form.

b Determine the values of time2 and complete the table.
c Construct a scatterplot of distance against time2 .
d Fit a least squares line to the transformed data, with distance as the RV.
e Use the regression equation to predict the distance travelled in 7 seconds.
f Obtain a residual plot and comment on whether the assumption of linearity is
reasonable.

Review
4 Is the infant mortality rate in a country associated with the number of doctors in
that country? The data below gives infant mortality rate in deaths per 1000 births
(mortality) and the number of doctors per 100 000 of population (doctors) for 14
countries.
mortality 12 13 12 10 10 7 111
doctors 192 222 154 182 179 204 61
mortality 15 10 20 54 75 121 71
doctors 270 271 357 79 59 27 52
a Construct a scatterplot of mortality against doctors and use it to comment on the

association between infant mortality rate and doctor numbers.
1
b Construct a scatterplot of mortality against .
doctors
c Determine the equation of the least squares regression line which would enable
1
mortality to be predicted from . Give the values of the intercept and slope
doctors
rounded to four significant figures.
d Obtain a residual plot for the model fitted in part c and comment on the linearity.
e Determine the value of coefficient of determination for the model fitted in part c.
Give your answer as a percentage rounded to one decimal place.
f Use the regression equation to predict the infant mortality rate in a country where
there are 100 doctors per 100,000 people. Give your answer rounded to the nearest
whole number.
g Comment on the reliability of the prediction made in part f.

Chapter
5
Investigating and
modelling time series
Chapter questions
I What is time series data?
I How do we construct a time series plot?
I How do we recognise features such as trend, seasonality and irregular
fluctuations?
I How do we smooth a time series plot using moving means?
I How do we smooth a time series plot using moving medians?
I How do we calculate and interpret seasonal indices?
I How do we calculate and interpret a trend line?
I How do we make forecasts of future values?
In this chapter we will focus on a special case of numerical bivariate data, called time
series data. In time series data the explanatory variable is always a measure of time (for
example hour, day, month or year), and we are concerned with understanding how the
response variable is changing over time.

5A Time series data 253
5A Time series data

Learning intentions
I To be able to recognise time series data.
I To be able to construct a time series plot.
I To be able to recognise features in the plot such as trend, seasonality and irregular
fluctuations.
When data concerned with a variable is collected, observed or recorded at successive

intervals of time, it is referred to as time series data. An example of time series data is
Annual road accident fatalities for Australia, 1982–2021, given in the following table.
Year Fatalities Year Fatalities Year Fatalities Year Fatalities

1982 3252 1992 1974 2002 1715 2012 1300
1983 2755 1993 1953 2003 1621 2013 1187
1984 2822 1994 1928 2004 1583 2014 1150
1985 2941 1995 2017 2005 1627 2015 1209
1986 2888 1996 1970 2006 1598 2016 1293
1987 2772 1997 1767 2007 1603 2017 1225
1988 2887 1998 1755 2008 1437 2018 1135
1989 2801 1999 1764 2009 1491 2019 1195
1990 2331 2000 1817 2010 1353 2020 1095
1991 2113 2001 1737 2011 1277 2021 1127
Since time series data is just a special kind of two numerical variable example, where
the explanatory variable is time, we will begin by drawing a scatterplot of the data. In
this instance, the scatterplot is called a time series plot, with time always placed on the
horizontal axis. A time series plot differs from a normal scatterplot in that, in general, the
points will be joined by line segments in time order. An example of a time series plot, of the
road accident fatality data, is given below.
3500
3000
Fatalities
2500
2000
1500
1000
1980 1990 2000 2010 2020 2030
Year
Looking at the time series plot, we can readily see a clear trend of decreasing road fatalities,
which is good news for drivers, as this provides some evidence that the many efforts being
made to reduce the road toll across Australia have been effective.

254 Chapter 5 Investigating and modelling time series
Constructing time series plots

As previously mentioned, time series data is a special case of bivariate numerical data,
where the explanatory variable is time. Consider the variable month, which takes values such
as January, February, March and so on. For the purpose of plotting and analysing time series
data, we can consider the variable month as numerical, taking the values {1, 2, 3, . . . }. If we
had monthly data for a two year period, then the variable month would take the values {1,
2, . . . , 24}. Whether the actual value of the variable is used in the plot (January, February,
March, . . . ) or its numerical equivalent (1, 2, 3, . . . ) is used, both time series plots would be
considered correct. We can use a similar approach for the variables day, or quarter.
Example 1 Constructing a time series plot
Maximum temperature was recorded each day for a week in a certain town. Construct a
time series plot of the data.
Day Mon Tues Wed Thur Fri Sat Sun

◦
Temperature ( C) 20 21 25 36 34 25 26
1 In a time series plot, time (day Day is the EV – this will label the horizontal axis.
in this case) is always the Temperature is the RV – this will label the vertical
explanatory variable (EV) and axis.
is plotted on the horizontal
axis.
2 Determine the scales for each A horizontal scale from 0–7 with intervals of 1 for
axis. each day would be suitable.
Temperature ranges from 20–36. A vertical scale
from 15–40 with intervals of 5 would be suitable.
3 Set up the axes, and then plot 40
all seven data points as for a 35
Temp (°C)
scatterplot. 30
25
20
15
Mon Tues Wed Thur Fri Sat Sun
Day
4 Complete the graph, by 40
joining consecutive data 35
Temp (°C)
points with straight lines. 30

25
20
15
Mon Tues Wed Thur Fri Sat Sun
Days

Most real-world time series data come in the form of large data sets that are best plotted with
the aid of a spreadsheet or statistical package. The availability of the data in electronic form
via the internet greatly helps this process. However, in this chapter, most of the time series
data sets are relatively small and can be readily plotted using a CAS calculator.
CAS 1: How to construct a time series using the TI-Nspire CAS

Construct a time series plot for the data presented below, which shows the birth rate in
Australia (in births per woman) from 2011–2020.
year 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
birth 1.926 1.920 1.855 1.826 1.814 1.752 1.741 1.740 1.657 1.580
Steps
into lists named year and birth.

Construct a scatterplot of birth against year.
As is the case for a time series plot, year is the
explanatory variable and birth the response
variable.
4 To display as a connected time series plot, move

the cursor to the main graph area and press
/ + b>Connect Data Points. Press ·.
CAS 1: How to construct a time series using the ClassPad

Construct a time series plot for the data presented below, which shows the birth rate in
Australia (in births per woman) from 2011–2020.

year 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
birth 1.926 1.920 1.855 1.826 1.814 1.752 1.741 1.740 1.657 1.580
Steps
1 Open the Statistics application and enter the data into the
columns named year and birth. Your screen should look
like the one shown.
2 Tap to open the Set StatGraphs dialog box and

complete as follows.
Draw: select On.
Type: select xyLine ( ).
XList: select main/year ( ).
YList: select main/birth ( ).
Freq: leave as 1.
Mark: leave as square.
3 Tap in the toolbar at the top of the screen to display the

time series plot in the bottom half of the screen.
To obtain a full-screen display, tap from the icon panel.
Tap from the toolbar, and use and to move from
point to point to read values from the plot.

Looking for patterns in time series plots

The features we look for in a time series are:
trend cycles seasonality

structural change possible outliers irregular (random) fluctuations.
We would always expect to see irregular or random fluctuations in a time series, and it is
common to see one or more of the other features as well.
Trend
Examining a time series plot we can often see a general upward or downward movement
over time. This indicates a long-term change over time that we call a trend.
Trend
The tendency for values in a time series to generally increase or decrease over a
significant period of time is called a trend.
One way of identifying trends on a time series graph is to draw a line that ignores the
fluctuations, but which reflects the overall increasing or decreasing nature of the plot. These
lines are called trend lines.
Trend lines have been drawn on the time series plots below to indicate an increasing trend
(line slopes upwards) and a decreasing trend (line slopes downwards).
trend line
trend line
Time Time
Sometimes, different trends are apparent in a time series for different time periods.

Example 2 Identifying trends
Consider the time series plot of the Australian annual birth rates over the years from 1931
to 1990, shown below. Comment on the trend shown in the plot.
3.6
3.2 trend 1
trend 2
2.8
Birth rate
2.4
2.0 trend 3
1.6
1930 1940 1950 1960 1970 1980 1990 2000

Year
Solution
There are three distinct trends, which can be seen by drawing trend lines on the plot. Each
of these trends can be explained by changing socioeconomic circumstances.
Trend 1: Between 1940 and 1961 the birth rate in Australia grew quite dramatically.
Those in the armed services came home from the Second World War, and the economy
grew quickly. This rapid increase in the Australian birth rate during this period is known
as the ‘Baby Boom’.
Trend 2: From about 1962 until 1980 the birth rate declined very rapidly. Birth control
methods became more effective, and women started to think more about careers. This
period is sometimes referred to as the ‘Baby Bust’.
Trend 3: During the 1980s, and beyond, the birth rate continued to decline slowly for a
complex range of social and economic reasons.
Cycles
The term cycle refers to variations in time series that in general last longer than a year.
These variations may not be of a regular height and they may not repeat at regular intervals.
Cycles
Cycles are recurrent movements in a time series, generally over a period greater than one
year.

Example 3 Identifying cycles
Sunspots are darker, cooler area on the surface of the sun. The following plot shows the
sunspot activity for the period 1945 to 2016. Comment on the cycles shown in the plot.
200
150
Sunspots
100
50
0
1940 1950 1960 1970 1980 1990 2000 2010 2020
Year
Solution
The recurrent pattern in the number of sunspots can be seen clearly from the time series
plot. Looking at the plot the years of lowest sunspot activity look to be at approximately
1954, 1964, 1975, 1986, 1996, 2008.
Many business indicators, such as interest rates or unemployment figures, also vary in
cycles, but their periods are usually less regular.
Seasonality
Cycles with calendar-related periods of one year or less are of special interest and are
referred to as seasonality.
Seasonality
Seasonality is present when there is a periodic movement in a time series that is related
to a calendar-related period – for example a year, a month or a week.
Seasonal movements tend to be more predictable than other time series features, and occur
because of variations in the weather, such as ice-cream sales increasing in the summer, or
institutional factors, like the increase in the number of unemployed people at the end of the
school year.
Example 4 Identifying seasonality
The plot below shows the total percentage of hotel rooms occupied in Australia by
quarter, over the years 2012–2016. Comment on the seasonality shown in the plot.

75.0
74.0
Room occupancy rate (%) 73.0
72.0
71.0
70.0
69.0
68.0
67.0
Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun
2012 2013 2014 2015 2016
Quarter
Solution
The regular peaks and troughs in the plot that occur at the same time each year signal the
presence of seasonality. In this case, the demand for accommodation is at its lowest in
the June quarter and highest in the December quarter.
This time series plot reveals both seasonality and trend in the demand for hotel rooms.
The upward sloping trend line signals the presence of a general increasing trend. This
tells us that, even though demand for accommodation has fluctuated from month to
month, demand for hotel accommodation has increased over time.
Structural change
A structural change in a time series is a sudden change in the pattern of the time series at a
point in time.
Structural change
Structural change is present when there is a sudden change in the established pattern of a
time series plot.
Example 5 Identifying structural change
The time series plot below shows the power bill for a rental house (in kWh) for the
12 months of a year. Comment on any structural change in the plot.

350
Electricity use (kWh)

300
250
200
150
100
50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Solution
The plot reveals an abrupt change in power usage in June to July. During this period,
monthly power use suddenly decreases from around 300 kWh per month from January to
June to around 175 kWh for the rest of the year. This is an example of structural change
that can probably be explained by a change in circumstances, for example, from a family
with children to a person living alone.
Structural change is also displayed in the birth rate time series plot we saw earlier. This
revealed three quite distinct trends during the period 1931–1990. These reflect significant
external events (like a war) or changes in social and economic circumstances.
Outliers
Outliers
Outliers are individual values that stand out from the general body of data.
Example 6 Identifying outliers
The time series plot below shows the daily power bill for a house (in kWh) for a fortnight.
Comment on any outliers in the plot.
12
Electricity use (kWh)
10
8
6
4
2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Day

Solution
For this household, daily electricity use follows a regular pattern that, although
fluctuating, averages about 10 kWh per day. In terms of daily power use, day 4 is a clear
outlier, with less than 2 kWh of electricity used. A follow-up investigation found that, on
this day, the house was without power for 18 hours due to a storm, so much less power
was used than normal.
Irregular (random) fluctuations
Irregular (random) fluctuations

Irregular (random) fluctuations include all the variations in a time series that we
cannot reasonably attribute to systematic changes like trend, cycles, seasonality and
structural change or an outlier. There will always be irregular, random variation present
in any real world time series data.
There can be many sources of irregular fluctuations, mostly unknown. A general

characteristic of these fluctuations is that they are unpredictable.
One of the aims of time series analysis is to develop techniques to identify regular patterns in
time series plots that are often obscured by irregular fluctuations. One of these techniques is
smoothing, which you will meet in the next section.
Identifying patterns in time series plots

The features we look for in a time series are:
trend cycles seasonality
structural change possible outliers irregular (random) fluctuations.
Trend is present when there is a long-term upward or downward movement in a time

series.
Cycles are present when there is a periodic movement in a time series. The period is the
time it takes for one complete up and down movement in the time series plot. In practice,
this term is reserved for periods greater than 1 year.
Seasonality is present when there is a periodic movement in a time series that has a
calendar related period – for example a year, a month or a week.
Structural change is present when there is a sudden change in the established pattern of
a time series plot.
Outliers are present when there are individual values that stand out from the data.
Irregular (random) fluctuations are always present in any real-world time series plot.
They include all the unexplained variations in a time series.

5A 5A Time series data 263
Exercise 5A
Constructing a time series plot

Note: A CAS calculator may be used to construct the time series plots. You may assign numerical values to
the values of the time variable where convenient to do so.
Example 1 1 Construct a time series plot to display the following data.
Year 2015 2016 2017 2018 2019 2020 2021 2022

Sales 2 23 35 31 45 23 67 70
2 Researchers recorded the number of penguins present on a remote island each month
for 12 months. Construct a time series plot of the data.
Month Number of penguins Month Number of penguins

January 449 July 180
February 214 August 241
March 170 September 311
April 265 October 499
May 434 November 598
June 102 December 674
3 The following table shows the maximum temperature in Melbourne during one week in
March. Construct a time series plot of the data.
Day Mon Tues Wed Thur Fri Sat Sun

Temperature (◦ C) 24.0 24.2 17.4 17.7 18.3 19.5 17.4
Identifying key features in a time series plot

Example 2 4 Complete the table below by indicating which of the listed features are present in each
of the time series plots shown below.
40
Plot
35 plot A
Feature A B C 30
Irr fluctuations 25
plot B
Increasing trend 20
15
Decreasing trend
10
5
plot C
0
2011 2012 2013 2014 2015
Year

264 Chapter 5 Investigating and modelling time series 5A
Example 4 of the time series plots shown below.
40
Plot plot A
35
Feature A B C 30
Irr fluctuations 25 plot B
Increasing trend 20
15
Decreasing trend
10
Cycles 5
plot C
Seasonality 0
2011 2012 2013 2014 2015
Year
40
Plot
35
Feature A B C
30 plot A
Irr fluctuations
25
Struc change
Increasing trend 20 plot B
Decreasing trend 15
Seasonality 10
5 plot C
0
2011 2012 2013 2014 2015
Year
70
Plot
60
Feature A B C plot A
Irr fluctuations 50
Struc change 40
Increasing trend 30
Decreasing trend
20
Outliers plot B
10
plot C
0
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Describing time series plots
Mobile phones per 100 people

8 The time series plot for the number 120
of mobile phones per 100 people
from 2000–2019 is shown to the 100
right. Describe the features of the 80

time series plot.
60
40
2000 2005 2010 2015 2020
Year
9 The data below shows the population (in millions) in Australia over the period
2012–2021.
Year 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Population 22.73 23.13 23.48 23.82 24.19 24.60 24.98 25.37 25.69 25.97
a Construct a time series plot of the data.

b Describe the features of the plot.
10 The table below shows the motor vehicle theft rate per 100 000 cars in Australia from
2003 to 2018.
Year 2003 2004 2005 2006 2007 2008 2009 2010
Theft rate 500.9 442.4 398.3 367.2 337.6 320.0 274.2 214.8
Year 2011 2012 2013 2014 2015 2016 2017 2018
Theft rate 220.0 228.4 204.2 191.0 194.5 231.0 213.3 214.1

b Describe the features of the plot.
11 The time series plot below 5000

shows the number of measles
4000
cases reported in Australia
measles cases
from 1988 to 2019. Describe 3000

the features of the plot.
2000
1000
0
1985 1990 1995 2000 2005 2010 2015 2020
Year

266 Chapter 5 Investigating and modelling time series 5A
12 The time series plot below shows the number of overseas arrivals (millions of people
per month) in Australia from November 2011 until December 2021. Describe the
features of the plot.
2.50
2.00
1.50
1.00
0.50
0.00
Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct
2015 2016 2017 2018 2019 2020 2021
13 a The time series plot shown shows the 80

smoking rates (%) of Australian males 70
Smokers (percentage)
and females over the period 1945–92. 60
50
i Describe any trends in the time series males
40
plot. females
30
ii Did the difference in smoking rates
20
increase or decrease over the period
10
1945–92? 0
1945 1955 1965 1975 1985 1995
Year
b The table below shows the smoking rates for females and males aged 15 years at
several time points from 2000–2018 (smoking rate data is not collected every year).
Year 2000 2005 2007 2010 2011 2012 2013 2018 2014 2015 2018
Female 22.4 18.9 19.9 17.9 15.4 16.6 14.4 15.6 13.5 14.5 13.6
Male 26.7 22.9 24.9 22.9 19.1 21.9 18.0 20.7 17.0 19.7 18.7
i Use a CAS calculator to construct time series plots of the male and female data.
ii Describe any trends in the time series plot.
iii Did the difference in smoking rates change over the period 2000–2018?


14 The time series plot below shows the quarterly room occupancy rate for a chain of
hotels over the years 2013 to 2019.
82.0
80.0
Occupancy rate (%)

78.0
76.0
74.0
72.0
70.0
68.0
2013 2014 2015 2016 2017 2018 2019
Year
The time series plot is best described as having
A seasonality only B irregular fluctuations only
C seasonality with irregular fluctuations
D an increasing trend with irregular fluctuations
E an increasing trend with seasonality and irregular fluctuations
15 The time series plot below shows the annual sales (in $ millions) for a car sales
company.
30
20
Sales
10
0
2012 2014 2016 2018 2020 2022
Year

A seasonality only
B irregular fluctuations only
D irregular fluctuations with an outlier
E seasonality with an outlier

5B Smoothing a time series using moving means

Learning intentions
I To be able to smooth a time series plot using moving means.
I To be able to know when and how to use centring when smoothing.
A time series plot can incorporate many of the sources of variation previously mentioned:
trend, cycles, seasonality, structural change, outliers and irregular fluctuations. One effect
of the irregular fluctuations and seasonality can be to obscure an underlying trend. The
technique of smoothing can sometimes be used to overcome this problem.
In this section we consider moving mean smoothing, which involves replacing individual
data points in the time series with the mean of the data point and some adjacent points. The
simplest method is to smooth over a small odd number of data points – for example, three or
five, but any number of points can be used.
The three-moving mean
To use three-moving mean smoothing, replace each data value with the mean of that
value and the one on each side. That is, if y1 , y2 and y3 are sequential data values, then:
y1 + y2 + y3
smoothed y2 =
3
The first and last points in the time series do not have values on each side, so they are
omitted.
The five-moving mean

To use five-moving mean smoothing, replace each data value with the mean of that
value and the two values on each side. That is, if y1 , y2 , y3 , y4 , y5 are sequential data
values, then:
y1 + y2 + y3 + y4 + y5
smoothed y3 =
5
The first two and last two points in the time series do not have two values on each side, so
they are omitted.
These definitions can be readily extended for moving means involving more points.
Example 7 Three- and five-moving mean smoothing
The table below gives the temperature (◦ C) recorded at a weather station at 9.00 a.m. each
day for a week.
Day Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Temperature 18.1 24.8 26.4 13.9 12.7 14.2 24.9
a Calculate the three-moving mean smoothed temperature for Tuesday.

b Calculate the five-moving mean smoothed temperature for Thursday.

5B Smoothing a time series using moving means 269
a 1 Write down the three 18.1, 24.8, 26.4
temperatures centred on (18.1 + 24.8 + 26.4)
Mean = = 23.1
Tuesday. 3
2 Find their mean and write down The three-moving mean smoothed
your answer. temperature for Tuesday is 23.1◦ C.
b 1 Write down the five temper- 24.8, 26.4, 13.9, 12.7, 14.2
atures centred on Thursday (24.8 + 26.4 + 13.9 + 12.7 + 14.2)
Mean =
2 Find their mean and write down 5
= 18.4
your answer.
The five-moving mean smoothed
temperature for Thursday is 18.4◦ C.
The next step is to extend these computations to smooth all terms in the time series.
Example 8 Three- and five-moving mean smoothing of a time series
The following table gives the number of births per month over a calendar year in a
country hospital. Use the three-moving mean and the five-moving mean methods,
rounded to one decimal place, to complete the table.
Solution
Complete the calculations as shown below.
Month Number of births 3-moving mean 5-moving mean
January 10
10 + 12 + 6
February 12 = 9.3
3
12 + 6 + 5 10 + 12 + 6 + 5 + 22
March 6 = 7.7 = 11.0
3 5
6 + 5 + 22 12 + 6 + 5 + 22 + 18
April 5 = 11.0 = 12.6
3 5
5 + 22 + 18 6 + 5 + 22 + 18 + 13
May 22 = 15.0 = 12.8
3 5
22 + 18 + 13 5 + 22 + 18 + 13 + 7
June 18 = 17.7 = 13.0
3 5
18 + 13 + 7 22 + 18 + 13 + 7 + 9
July 13 = 12.7 = 13.8
3 5
13 + 7 + 9 18 + 13 + 7 + 9 + 10
August 7 = 9.7 = 11.4
3 5
7 + 9 + 10 13 + 7 + 9 + 10 + 8
September 9 = 8.7 = 9.4
3 5
9 + 10 + 8 7 + 9 + 10 + 8 + 15
October 10 = 9.0 = 9.8
3 5
10 + 8 + 15
November 8 = 11.0
3
December 15

The result of this smoothing can be seen in the plot below, which shows the raw data,
the data smoothed with a three-moving means and the data smoothed with a five-moving
means.
25
20 raw data
Number of births
3-moving mean
15 5-moving mean
10
0
n
ar
pr
ct
ov
ec
ay
ug
p
Ju
Fe
Ja
Ju
Se
D
M
N
Note: In the process of smoothing, data points are lost at the beginning and end of the time series.
Two observations can be made from this plot:

1 Five-moving mean smoothing is more effective in reducing the irregular fluctuations than
three-mean smoothing.
2 The five-moving mean smoothed plot shows that there is no clear trend although the raw
data suggest that there might be an increasing trend.
Moving mean smoothing with centring

If we smooth over an even number of points, we run into a problem. The centre of the set of
points is not at a time point belonging to the original series. Usually, we solve this problem
by using a process called centring.
Centring
Smoothing with centring involves taking a two-moving mean of the already smoothed
values so that they line up with the original data values. Smoothing with centring is only
required when smoothing using an even number of data values, for example 2-moving
mean smoothing, or 4-moving mean smoothing.
We will illustrate the process by finding the two- Day Temperature

moving mean, centred on Tuesday, for the daily
Monday 18.1
temperature data opposite.
Tuesday 24.8
Wednesday 26.4
It is straightforward to calculate a series of two-moving means for this data by calculating

the mean for Monday and Tuesday, followed by the mean for Tuesday and Wednesday.
However, as we can see in the diagram below, these means do not align with a particular day,
but lie between days. To solve this problem find the average of these two means, as shown in
the following diagram.
Two-moving mean with

Day Temperature Two-moving means centring
Monday 18.1
(18.1 + 24.8) = 21.45
2
(21.45 + 25.6)
Tuesday 24.8 = 23.525
2
(24.8 + 26.4) = 25.60
2
Wednesday 26.4
In practice, we do not have to draw such a diagram to perform these calculations. The
purpose of doing so is to show how the centring process works. Calculating a two-moving
mean with centring is illustrated in the following example.
Example 9 Two-moving mean smoothing with centring
The temperatures (◦ C) recorded at a weather station at 9 a.m. each day for a week are
displayed in the table.

Temperature 18.1 24.8 26.4 13.9 12.7 14.2 24.9
Calculate the two-moving mean smoothed temperature with centring for Tuesday.
1 For two-mean smoothing with centring, 18.1 24.8 26.4
write down the three data values
centred on Tuesday (highlighted in
red).
(18.1 + 24.8)
2 Calculate the mean of the first two mean 1 = = 21.45
2
values (mean 1). Calculate the mean of
(24.8 + 26.4)
the second two values (mean 2). mean 2 = = 25.60
2
(mean 1 + mean 2)
3 The centred mean is then the average of Centred mean =
2
mean 1 and mean 2. (21.45 + 25.60)
=
2
= 23.525
4 Write down your answer, rounded to The two-moving mean smoothed
one decimal place. temperature for Tuesday is 23.5◦ C.
The process of smoothing with centring across more data values is the same as two-mean
smoothing except that the means are determined in larger groups. This process is illustrated
in the following example with groups of four and six.

Example 10 Four- and six-moving mean smoothing with centring
The table below gives the temperature (◦ C) recorded at a weather station at 9.00 a.m. each
day for a week.

Temperature 18.1 24.8 26.4 13.9 12.7 14.2 24.9
a Calculate the four-moving mean smoothed temperature with centring for Thursday.
b Calculate the six-moving mean smoothed temperature with centring for Thursday.
a 1 For four-mean smoothing 24.8 26.4 13.9 12.7 14.2
with centring, write down the
five data values centred on
Thursday.
(24.8 + 26.4 + 13.9 + 12.7)
2 Calculate the mean of the mean 1 =
4
first four values (mean 1)
= 19.45
and the mean of the last four
(26.4 + 13.9 + 12.7 + 14.2)
values (mean 2). mean 2 =
4
= 16.8
(mean 1 + mean 2)
3 The centred mean is then the centred mean =
2
average of mean 1 and mean (19.45 + 16.8)
2. =
2
= 18.125
4 Write down your answer. The four-mean smoothed temperature centred on
Thursday is 18.1 ◦ C (to 1 d.p.).
b 1 For six-mean smoothing with 18.1 24.8 26.4 13.9 12.7 14.2 24.9
centring, write down the
seven data values centred on
Thursday.
(18.1 + 24.8 + 26.4 + 13.9 + 12.7 + 14.2)
2 Calculate the mean of the mean 1 =
6
first six values (mean 1)
= 18.35
and the mean of the last six
(24.8 + 26.4 + 13.9 + 12.7 + 14.2 + 24.9)
values (mean 2). mean 2 =
6
= 19.4833 . . .

(mean 1 + mean 2)
3 The centred mean is then the centred mean =
2
average of mean 1 and mean (18.35 + 19.483)
2. =
2
= 18.917
4 Write down your answer. The six-mean smoothed temperature centred on
Thursday is 18.9 ◦ C (to 1 d.p.).
The next step is to extend these computations to smooth all terms in the time series. This
process is illustrated using four-moving mean smoothing in the following example. Setting
up and using a table like the one shown in the example will help keep track of the process.
Example 11 Smoothing of a time series using four-mean smoothing with centring
The following table gives the number of births per month over a calendar year in a
country hospital. Use the four moving mean with centring method to complete the table.
Solution
Complete the calculations as shown below.
4-moving mean with
Month Number of births 4-moving mean
centring
January 10
10 + 12 + 6 + 5
February 12 = 8.25
4
8.25 + 11.25
March 6 = 9.75
2
12 + 6 + 5 + 22
= 11.25
4
11.25 + 12.75
April 5 = 12
2
6 + 5 + 22 + 18
= 12.75
4
12.75 + 14.5
May 22 = 13.625
2
5 + 22 + 18 + 13
= 14.5
4
14.5 + 15
June 18 = 14.75
2
22 + 18 + 13 + 7
= 15
4
15 + 11.75
July 13 = 13.375
2
18 + 13 + 7 + 9
= 11.75
4
11.75 + 9.75
August 7 = 10.75
2
13 + 7 + 9 + 10
= 9.75
4
9.75 + 8.5
September 9 = 9.125
2
7 + 9 + 10 + 8
= 8.5
4
8.5 + 10.5
October 10 = 9.5
2
9 + 10 + 8 + 15
= 10.5
4
November 8
December 15

274 Chapter 5 Investigating and modelling time series 5B
Skill-
sheet Exercise 5B
Note: A CAS calculator may be used to construct the time series plots.
Calculating the smoothed values of an odd number of individual points

Example 7 1
t 1 2 3 4 5 6 7 8 9
y 5 2 5 3 1 0 2 3 0
For the time series data in the table above, find:
a the three-mean smoothed y-value for
i t=4 ii t = 6 iii t = 2
b the five-mean smoothed y-value for
i t=3 ii t = 7 iii t = 4
c the seven-mean smoothed y-value for
i t=4 ii t = 6
d the nine-mean smoothed y-value for t = 5
2 The table below gives the temperature (◦ C) recorded at a weather station at 3.00 p.m.
each day for a week.
Day Mon Tue Wed Thu Fri Sat Sun

Temp (◦ C) 28.9 33.5 21.6 18.1 16.2 17.9 26.4
a Find the three-mean smoothed temperature for Wednesday.

b Find the five-mean smoothed temperature for Friday.
c Find the seven-mean smoothed temperature for Thursday.
Example 8 3 Complete the following table.
t 1 2 3 4 5 6 7 8 9
y 10 12 8 4 12 8 10 18 2
3-moving mean y – –
5-moving mean y – – – –
Smoothing and plotting a time series plot (odd number of points)

4 The maximum temperature of a city over a period of 10 days is given below.
Day 1 2 3 4 5 6 7 8 9 10
Temperature (◦ C) 24 27 28 40 22 23 22 21 25 26
3-moving mean 26.3 30.0 22.3 22.7 24.0
5-moving mean 28.2 27 23.4

5B 5B Smoothing a time series using moving means 275
a Construct a time series plot of the temperature data.

b Use the five-mean and seven-mean smoothing method to complete the table.
c Plot the smoothed temperature data and compare the plots.
5 The value of the Australian dollar in US dollars (exchange rate) over 10 days is given below.
Day 1 2 3 4 5 6 7 8 9 10
Exchange rate 0.743 0.754 0.737 0.751 0.724 0.724 0.712 0.735 0.716 0.711
3-moving mean 0.745 0.747 0.733 0.721 0.721
5-moving mean 0.738 0.730 0.722

b Use the three-mean and five-mean smoothing method to complete the table.
c Plot the smoothed exchange rate data and compare the plots.
Calculating the smoothed values of an even number of individual points

Example 9 6
t 1 2 3 4 5 6 7 8 9
y 5 2 5 3 1 0 2 3 0
For the time series data above, find the two-mean smoothed y-value centred at:
a t=3 b t=8
Example 10 7 Use the time series data in the table in Question 6 to find:
a the four-mean smoothed y-value centred at t = 3
b the four-mean smoothed y-value centred at t = 6
c the six-mean smoothed y-value centred at t = 3
d the six-mean smoothed y-value centred at t = 6
8 The table below gives the minimum daily temperature (◦ C) recorded at a weather
station over a 10 day period.
Day 1 2 3 4 5 6 7 8 9 10
Temperature (◦ C) 8.9 3.5 11.6 14.1 12.5 13.3 6.4 8.5 9.1 4.5
a Find the two-mean smoothed temperature with centring for Day 5.

b Find the four-mean smoothed temperature with centring for Day 5.
c Find the six-mean smoothed temperature with centring for Day 5.
Smoothing and plotting a time series plot (even number of points)

Example 11 9 The table below gives the number of complaints recorded at a customer service centre
over a 12 month period.

276 Chapter 5 Investigating and modelling time series 5B
Month Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec
Number of complaints 10 12 6 5 22 18 13 7 9 10 8 15
2-moving mean 10.0 7.3 9.5 16.8 17.8 12.8 9.0 9.3 10.3

b Show that the two-mean smoothed value with centring for September is equal to 8.8
(rounded to one decimal place).
c Plot the smoothed data and compare the plots.
10 The table below gives the amount of rain (in mm) recorded each month at a weather
station.
Month Apr May Jun July Aug Sep Oct Nov Dec
Rainfall (mm) 21.4 40.5 52.3 42.1 58.9 79.9 81.5 54.3 50.0
4-moving mean 43.8 53.4 67.1 67.5

b Show that the four-mean smoothed value with centring for August is equal to 62.0
(rounded to one decimal place).
c Plot the smoothed data and compare the plots.

11 Hay Lam records the number of emails (emails) he receives over a one-week period.

Emails 85 65 77 10
The numbers of emails he received on Thursday, Friday and Saturday are not shown.
The five-mean smoothed number of emails he received on Friday is 39.
The three-mean smoothed number of emails he received on Friday is:
A 36 B 39 C 40 D 42 E 45
12 The table shows the closing price (price) of a company’s shares on the stock market
over a 10 day period.
Day 1 2 3 4 5 6 7 8 9 10
Price($ ) 0.99 1.05 1.10 1.25 1.29 1.37 2.42 1.95 2.05 2.35
The six-mean smoothed with centring closing share price on Day 6 is closest to:
A $1.56 B $1.62 C $1.64 D $1.88 E $1.72

5B 5C Smoothing a time series plot using moving medians 277
The time series plot below shows the amount that Arnold saved each month (in dollars) over
a 12 month period.
250
200
Amount saved ($)
150
100
50
0
ch
y
ry
il
er
r
ay
us
be
be
be
n
ar
l
pr
Ju
ob
ua
ar
Ju
M
ug
em
em
m
nu
A
M
ct
br
e
A
Ja
pt
ov
ec
O
Fe
D
Se
N
13 If he saved a total of $831 over the period from May to September, the five-mean
smoothed amount that he saved in July is closest to:
A $277 B $190 C $182 D $152 E $166
14 If seven-mean smoothing is used to smooth this time series plot, the number of
smoothed data points would be:
A 3 B 4 C 5 D 6 E 7
5C Smoothing a time series plot using moving

medians
Learning intentions
I To be able to locate the median of a data set graphically.
I To be able to smooth a time series plot using moving medians.
Another simple and convenient method of smoothing a time series is to use moving median
smoothing. The advantage of this method is that it can be done directly on the graph without
needing to know the exact values of each data point.1 However, before smoothing a time
series plot graphically using moving medians we will first need to know how to locate
medians graphically.
1
Note that, in this course, median smoothing is restricted to smoothing over an odd number of points, so
centring is not required.

Locating medians graphically

The graph opposite shows three data points y
plotted on a set of coordinate axes. The task is 5
to locate the median of these three points. The 4
median will be a point somewhere on this set of 3

2
coordinate axes. To locate this point we proceed
1
as follows.
0 x
0 1 2 3 4 5
y
Step 1 5
middle
Identify the middle data point moving in the 4 x-value
x-direction. Draw a vertical line through this 3
2 middle
value as shown. y-value
1
Step 2 0 x
0 1 2 3 4 5
Identify the middle data point moving in the
y-direction. Draw a horizontal line through this
value as shown.
y
Step 3 5
The median value is where the two lines intersect 4

3 × (3, 3)
– in this case, at the point (3, 3).
2 median
Mark this point with a cross (×). point
1
0 x
0 1 2 3 4 5
Smoothing a time series using moving median smoothing

The process of graphically smoothing a time series plot requires no more than repeating
the above process for each group of three or five data points in the plot as required. The
following worked examples demonstrate the process.
Example 12 Three-moving median smoothing using a graphical approach
Construct a three-median 25
raw data
smoothed plot of the time series
Number of births
20
plot shown opposite. 15
10
5
0
ay
ar
ov
ug
pr
ec
b
p
ct
n
n
l
Ju
Fe
Se
Ju
Ja
O
M
D
A
M

5C Smoothing a time series plot using moving medians 279
1 Locate on the time series plot 25
raw data
the median of the first three first 3-median
Number of births
20
points (Jan, Feb, Mar). point
15 2
1 middle number of births
10
3
5
middle month
0
n
b
n
ec
p
ar
ug
ay
ct
pr
ov
l
Ju
Ju
Fe
Ja
Se
O
D
M
A
M
N
2 Continue this process by 25
raw data
moving onto the next three 20 3-median
Number of births
points to be smoothed (Feb, point
15
Mar, Apr).
10
Mark their medians on the
graph, and continue the 5
process until you run out of 0
n
b
n
ec
p
ar
ug
ay
ct
pr
ov
l
Ju
Ju
Fe
Ja
groups of three.
Se
O
D
M
A
M
N
3 Join the median points with a 25
line segment – see opposite. raw data
Number of births
20 3-median
point
15
10
5
0
n
b
n
ec
p
ar
ug
ay
ct
pr
ov
l
Ju
Ju
Fe
Ja
Se
O
D
M
A
M
Example 13 Five-moving median smoothing using a graphical approach
Construct a five-median 25
raw data
smoothed plot of the time series
Number of births
20
plot shown opposite.
15
Note: The starting point for a median
10
smoothing is a time series plot and
you smooth directly onto the plot. 5
Copies of the plots in this section can
0
be accessed through the skillsheet icon
ay
ar
ov
ug
ec
pr
b
p
ct
n
n
l
Ju
Fe
Se
Ju
Ja
O
M
D
A
M
in the Interactive Textbook.

N

280 Chapter 5 Investigating and modelling time series 5C
1 Locate on the time series plot 25 5 raw data
the median of the first five first 5-median
Number of births
20
middle month point
points (Jan, Feb, Mar, Apr, 15 2
May), as shown. 1 middle number of births
10
5 3 4
0
ay
ar
ov
ug
pr
ec
b
p
ct
n
n
l
Ju
Fe
Se
Ju
Ja
O
M
D
A
M
N
2 Then move onto the next five points to be smoothed (Feb, Mar, Apr, May, Jun). Repeat
the process until you run out of groups of five points. The five-median points are then
joined up with line segments to give the final smoothed plot, as shown.
25
raw data
Number of births
20 5-median
smoothed
15
10
5
0
ay
ar
ov
ug
pr
ec
b
p
ct
n
l
Ju
Fe
Ju
Se
Ja
O
M
D
A
M
N
Note: The five-median smoothed plot is much smoother than the three-median smoothed plot.
Skill-
sheet Exercise 5C
Note: Copies of all plots in this section can be accessed through the skillsheet icon in the Interactive
Textbook.
Locating the median of a set of data points graphically

1 Mark the location of the median point for each of the sets of data points below.
a 5 b 5
4 4
3 3
2 2
1 1
0 0
0 1 2 3 4 5 0 1 2 3 4 5

5C 5C Smoothing a time series plot using moving medians 281
c 5 d 5
4 4
3 3
2 2
1 1
0 0
0 1 2 3 4 5 0 1 2 3 4 5
Smoothing a time series graphically

Example 12 2 The time series plot below shows 45
40
the maximum daily temperatures 35
Temperature
(in ◦ C) in a city over a period of 10 30
25
consecutive days. 20
15
10
5
0
0 1 2 3 4 5 6 7 8 9 10
Day
Use three-median smoothing to determine the smoothed temperature for:
a day 4 b day 8
3 The time series plot below 30

28
shows the annual sales (in 26
Car sales ($millions)
$ millions) for a car sales 24

22
company. Use three-moving 20
18
median smoothing to 16
graphically smooth the 14
12
plot and comment on the 10
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
smoothed plot. Year
Example 13 4 Use the time series plot in Question 2 to find the five-median smoothed
temperature for:
a day 4 b day 8

282 Chapter 5 Investigating and modelling time series 5C
5 The time series plot opposite 0.76

shows the value of the 0.75
Exchange rate
0.74
Australian dollar in US dollars 0.73
(the exchange rate) over a 0.72
period of 10 consecutive days 0.71
0.70
in 2009.
0
Use five-moving median 0 1 2 3 4 5 6 7 8 9 10
smoothing to graphically Day
smooth the plot and comment
on the smoothed plot.
6 Use the graphical approach to smooth the time series plot below using:
a three-moving median smoothing b five-moving median smoothing.
70
Number of whales (’000s)
60
50
40
30
20
10
0
20 25 30 35 40 45 50 55 60 65 70 75 80 85
19 19 19 19 19 19 19 19 19 19 19 19 19 19
Year
7 The time series plot opposite shows

the percentage growth of GDP
(gross domestic product) over a
13-year period.
a Find the median value of the percentage growth in GDP over the 13 year period.
b Smooth the times series graph using:
i three-moving median smoothing ii five-moving median smoothing.
c What conclusions can be drawn about the variation in GDP growth from these
smoothed time series plots?

5C 5D Seasonal indices 283
Use the following information to answer Questions 8 to 10
The time series plot below shows the amount that Lulu saved each month (to the nearest $)
over a 12 month period.
260
Amount saved ($) 240
220
200
180
160
140
120
100
80
8 During the year shown in the time series plot the median monthly amount Lulu saved is
closest to:
A $180 B $155 C $130 D $190 E $200
9 The five-median smoothed amount saved by Lulu in July is closest to:

A $130 B $150 C $170 D $190 E $200
10 The nine-median smoothed amount saved by Lulu in August is closest to:

A $132 B $160 C $168 D $180 E $218
5D Seasonal indices
Learning intentions
I To be able to interpret the meaning of seasonal indices.
I To be able to seasonally adjust data using seasonal indices.
I To be able to calculate seasonal indices from time series data.
When the data is considered to have a seasonal component, it is often necessary to remove
this component so any underlying trend is clearer. The process of removing the seasonal
component is called deseasonalising the data. To do this we need to calculate seasonal
indices. Seasonal indices tells us how a particular season (generally a day, month or quarter)
compares to the average season.

The concept of a seasonal index

Consider the (hypothetical) monthly seasonal indices for unemployment given in the table.
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Total
1.1 1.2 1.1 1.0 0.95 0.95 0.9 0.9 0.85 0.85 1.1 1.1 12.0
Key fact 1
Seasonal indices are calculated so that their average is 1. This means that the sum of the
seasonal indices equals the number of seasons. Thus, if the seasons are months, the seasonal
indices add to 12. If the seasons are quarters, then the seasonal indices add to 4, and so on.
Key fact 2
Seasonal indices tell us how a particular season (generally a day, month or quarter) compares
to the average season.
For example:
seasonal index for unemployment for the month of February is 1.2 or 120%.
This tells us that February unemployment figures tend to be 20% higher than the monthly
average. Remember, the average seasonal index is 1 or 100%.
seasonal index for August is 0.90 or 90%.
This tells us that the August unemployment figures tend to be only 90% of the monthly
average. Alternatively, August unemployment figures are 10% lower than the monthly
average.
Example 14 Interpreting seasonal indices
Suppose that the seasonal indices (SI) for electricity usage in Esse’s home are as shown in
the table:
Summer Autumn Winter Spring

1.16 0.94 1.26 0.64
a What does the seasonal index for Winter tell us?

b What does the seasonal index for Spring tell us?
Solution
a The seasonal index for Winter is 1.26. This tells us that Esse’s electricity usage in
Winter is typically 26% higher than the average season.
b The seasonal index for Spring is 0.64. This tells us that Esse’s electricity usage in
Spring is typically 36% lower than the average season.

5D Seasonal indices 285
Using seasonal indices to seasonally adjust a time series

We can use seasonal indices to remove the seasonal component (deseasonalise) from a time
series, or to put it back in (reseasonalise).When we do this we are said to seasonally adjust
the data.
To calculate deseasonalised figures, each entry is divided by its seasonal index as follows.
Deseasonalising data
Time series data are deseasonalised using the relationship:
actual figure
deseasonalised figure =
seasonal index
The rule for determining deseasonalised data values can also be used to reseasonalise data –
that is, convert a deseasonalised value into an actual data value.
Reseasonalising data
Time series data are reseasonalised using the rule:
actual figure = deseasonalised figure × seasonal index
Example 15 Using seasonal indices
The seasonal indices (SI) for cold drink sales for Imogen’s kiosk are as shown in the
table:
1.75 0.66 0.46 1.13
a If the actual actual cold drink sales last summer totalled $21 653, what is the
deseasonalised sales figure for that time period?
b If the deseasonalised cold drink sales last spring totalled $10 870, what were the actual
sales for that time period?
21 653
a To deseasonalise we divide by the Deseasonalised sales =
1.75
seasonal index for Summer (1.75)
= $12 373.14
b To find the actual sales we multiply by the Actual sales = 10 870 × 1.13
seasonal index for Spring (1.13). = $12 283.10

Example 16 Using seasonal indices to determine percentage change required to

correct for seasonality
Consider the table below which gives the seasonal indices for heater sales at a discount
store:

0.65 1.25 1.35 0.0.75
a By what percentage should the sales in summer be increased or decreased in order

to deseasonalise the data? Give your answer as a percentage rounded to one decimal
place.
b By what percentage should the sales in winter be increased or decreased in order to
deseasonalise the data? Give your answer as a percentage rounded to one decimal
place.
a 1 Insert the seasonal index for summer In general for summer:
into the rule actual sales
deseasonalised sales =
actual sales 0.65
seasonal index 1
= × actual sales
0.65
= 1.538 × actual sales
2 Convert 1.538 into a percentage Multiplying the actual sales by 1.538 is the
increase or decrease. Write the equivalent of increasing the actual sales by
answer in a sentence. 53.8%.
To correct for seasonality, the actual sales
should be increased by 53.8%.
b 1 Insert the seasonal index for winter In general for winter:
into the rule actual sales
deseseasonalised sales =
actual sales 1.35
seasonal index 1
= × actual sales
1.35
= 0.741 × actual sales
2 Convert 0.741 into a percentage Multiplying the actual sales by 0.741 is the
increase or decrease. Write the
equivalent of decreasing the actual sales by
answer in a sentence.
(100%-74.1%) = 25.9%.
To correct for seasonality, the actual sales
should be increased by 25.9%.

Calculating seasonal indices

To complete this section, we will describe how to calculate a seasonal index. We will start
by using only one year’s data to illustrate the basic ideas and then move onto a more realistic
example where several years’ data are involved.
Example 17 Calculating seasonal indices (1 year’s data)
Mikki runs a shop and she wishes to

determine quarterly seasonal indices
920 1085 1241 446
for the number of customers to her
shop based on last year’s figures
which are shown in the table opposite.
value for season
1 The seasons are quarters. Write the Seasonal index =
seasonal average
formula in terms of quarters.
920 + 1085 + 1241 + 446
2 Find the quarterly average for the Quarterly average =
4
year.
= 923
3 The seasonal index (SI) for each
920
quarter is the ratio of that quarter’s SISummer = = 0.997
923
sales to the average quarter.
1085
SIAutumn = = 1.176
923
1241
SIWinter = = 1.345
923
446
SISpring = = 0.483
923
4 Check that the seasonal indices Check: 0.997 + 1.176 + 1.345 + 0.483 = 4.001
sum to 4 (the number of seasons).
The slight difference is due to
rounding.
5 Write out your answers as a table Seasonal indices
of the seasonal indices. Summer Autumn Winter Spring
0.997 1.176 1.345 0.483
The next example illustrates how seasonal indices are calculated with 3 years’ data. While
the process looks more complicated, we just repeat what we did in the previous example
three times and average the results for each year at the end.

Example 18 Calculating seasonal indices (several years’ data)
Suppose that Mikki has 3 years of data, as shown. Use the data to calculate seasonal
indices, rounded to two decimal places.
Year Summer Autumn Winter Spring

1 920 1085 1241 446
2 1035 1180 1356 541
3 1299 1324 1450 659
Solution
The strategy is as follows:
Calculate the seasonal indices for years 1, 2 and 3 separately. As we already have the
seasonal indices for year 1 in the previous example we will save ourselves some time
by simply quoting the result.
Average the three sets of seasonal indices to obtain a single set of seasonal indices.
1 Write down the result for Year 1 seasonal indices:
year 1. Summer Autumn Winter Spring
0.997 1.176 1.345 0.483
2 Now calculate the seasonal
indices for year 2.
value for quarter
a The seasons are quarters. Seasonal index =
quarterly average
Write the formula in terms
of quarters.
1035 + 1180 + 1356 + 541
b Find the quarterly average Quarterly average =
4
for the year.
= 1028
c Work out the seasonal
1035
index (SI) for each time SISummer = = 1.007
1028
period. 1180
SIAutumn = = 1.148
1028
1356
SIWinter = = 1.319
1028
541
SISpring = = 0.526
1028

d Check that the seasonal Check: 1.007 + 1.148 + 1.319 + 0.526 = 4.000
indices sum to 4.
e Write out your answers Year 2 seasonal indices:
as a table of the seasonal
indices.
1.007 1.148 1.319 0.526
3 Now calculate the seasonal
indices for year 3.
1299 + 1324 + 1450 + 659
a Find the quarterly average Quarterly average =
4
for the year.
= 1183
1299
b Work out the seasonal SISummer = = 1.098
1183
index (SI) for each time 1324
period. SIAutumn = = 1.119
1183
1450
SIWinter = = 1.226
1183
659
SISpring = = 0.557
1183
c Check that the seasonal Check: 1.098 + 1.119 + 1.226 + 0.557 = 4.000
indices sum to 4.
d Write out your answers Year 3 seasonal indices:
as a table of the seasonal
indices.
1.098 1.119 1.226 0.557
4 Find the 3-year averaged Final seasonal indices:
seasonal indices by averaging 0.997 + 1.007 + 1.098
SSummer = = 1.03
the seasonal indices for each 3
1.176 + 1.148 + 1.119
season. SAutumn = = 1.15
3
1.345 + 1.319 + 1.226
SWinter = = 1.30
3
0.483 + 0.526 + 0.557
SSpring = = 0.52
3
5 Check that the seasonal Check: 1.03 + 1.15 + 1.30 + 0.52 = 4.00
indices sum to 4.
6 Write out your answers as a
table of the seasonal indices.
1.03 1.15 1.30 0.52

Interpreting the seasonal indices

Having calculated these seasonal indices, what do they tell about the previous situation?
The seasonal index of:
1.03 tells us that in summer, customer numbers are typically 3% above average.
1.15 tells us that in autumn, customer numbers are typically 15% above average.
1.30 tells us that in winter, customer numbers are typically 30% above average.
0.52 tells us that in spring, customer numbers are typically 48% below average.
Using seasonal indices to deseasonalise a time series

Once we have determined the seasonal indices we can use the rule for deseasonalising the
time series introduced earlier on this section
actual figure
deseasonalised figure =
seasonal index
to deseasonalise the data.
Example 19 Deseasonalising a time series
The quarterly sales figures for Mikki’s shop over a 3-year period are given below.

1 920 1085 1241 446
2 1035 1180 1356 541
3 1299 1324 1450 659
Use the seasonal indices shown to
deseasonalise these sales figures. Write
1.03 1.15 1.30 0.52
answers rounded to the nearest whole
number.
1 To deseasonalise each sales figure in Deseasonalised Summer sales:
the table, divide by the appropriate 920
Year 1: = 893
seasonal index. 1.03
1035
For example, for summer, divide the Year 2: = 1005
1.03
figures in the ‘Summer’ column by 1.03. 1299
Round results to the nearest whole Year 3: = 1261
1.03
number.
2 Repeat for the other seasons. Deseasonalised sales figures
1 893 943 955 858
2 1005 1026 1043 1040
3 1261 1151 1115 1267

5D 5D Seasonal indices 291
Why deseasonalise?
The purpose of removing the seasonality component of a time series is generally so that
any trend in the time series is clearer. Consider again the actual customer data, and the
deseasonalised customer data from Example 18, both of which are shown in the following
time series plots.
1400
Deseasonalised customers
1600 1300
Actual customers
1400 1200
1200 1100
1000 1000
800 900
600 800
400 700
200 600
0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Quarter Quarter
It is hard to see from the first plot whether there has been any growth in Mikki’s business,
but the deseasonalised plot reveals revealed a clear underlying trend in the data.
It is common to deseasonalise time series data before you fit a trend line. We will consider
this further in the next section.
Skill-
sheet Exercise 5D
Basic skills and interpretation

Use the following information to answer Questions 1 to 4.
The table below shows the monthly sales figures (in $’000s) and seasonal indices (for
January to November) for a product produced by the U-beaut company.
Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Sales ($’000s) 9.6 10.5 8.6 7.1 6.0 5.4 6.4 7.2 8.3 7.4
Seasonal index 1.2 1.3 1.1 1.0 1.0 0.9 0.8 0.7 0.9 1.0 1.1
Example 14 1 a Find the seasonal index for December.

b Interpret the seasonal index for February.
c Interpret the seasonal index for September.
Example 15 2 a Find the deseasonalised sales figure (in $’000s) for March, giving your answer
b Find the deseasonalised sales figure (in $’000s) for June, giving your answer

292 Chapter 5 Investigating and modelling time series 5D
3 a The deseasonalised sales figure (in $’000s) for August is 5.6. Find the actual sales
(in $’000s), giving your answer rounded to one decimal place.
b The deseasonalised sales figure (in $’000s) for April is 6.9. Find the actual sales (in
$’000s), giving your answer rounded to one decimal place.
Example 16 4 a By what percentage should the sales in August be increased or decreased in order to
correct for seasonality? Give your answer as a percentage rounded to one decimal
place.
b By what percentage should the sales in February be increased or decreased in order
to correct for seasonality? Give your answer as a percentage rounded to one decimal
place.
5 The table below shows the quarterly newspaper sales (in $’000s) of a corner store.
Also shown are the seasonal indices for newspaper sales for the first, second and third
quarters.
Quarter 1 Quarter 2 Quarter 3 Quarter 4
Sales 1060 1868 1642
Seasonal index 0.8 0.7 1.3
a Find the seasonal index for quarter 4.

b Find the deseasonalised sales (in $’000s) for quarter 2.
c Find the deseasonalised sales (in $’000s) for quarter 3.
d The deseasonalised sales (in $’000s) for quarter 1 are 1256. Find the actual sales.
Deseasonalising a time series

6 The following table shows the number of students enrolled in a 3-month computer
systems training course along with some seasonal indices that have been calculated
from the previous year’s enrolment figures. Complete the table by calculating the
seasonal index for spring and the deseasonalised student numbers for each course.

Number of students 56 125 126 96
Deseasonalised numbers
7 The number of waiters employed by a restaurant chain in each quarter of 1 year, along
with some seasonal indices that have been calculated from the previous years’ data, are
given in the following table.

5D 5D Seasonal indices 293

Number of waiters 198 145 86 168
a What is the seasonal index for the second quarter?

b The seasonal index for quarter 1 is 1.30. Explain what this means in terms of the
average quarterly number of waiters.
c Deseasonalise the data.
Calculating seasonal indices

Example 17 8 The table below records quarterly sales (in $’000s) for a shop.

60 56 75 78
Use the data to determine the seasonal indices for the four quarters. Give your results
9 The table below records the monthly visitors (in ’000s) to a museum over one year.
12 13 14 17 18 15 9 10 8 11 15 20
Use the data to determine the seasonal indices for the 12 months. Give your results
Example 18 10 The table below records the monthly sales (in $’000s) for a shop over a two year
period.
22 19 25 23 20 18 20 15 14 11 23 30
21 20 23 25 22 17 19 17 16 11 25 31
Use the data to determine the seasonal indices for the 12 months. Give your results
Example 19 11 The daily number of cars carried on a car ferry service each day over a two-week
period, together with the daily seasonal indices, are shown in the table below:
Week Mon Tues Wed Thur Fri Sat Sun

1 124 110 45 67 230 134 330
2 120 108 57 74 215 150 345
Seasonal index 0.8 0.7 0.3 0.5 1.5 1.0 2.2

294 Chapter 5 Investigating and modelling time series 5D
a Use the seasonal indices to deseasonalise the data, rounding the answers to the
nearest whole number.
b Use your calculator to construct a time series plot the the deseasonalised data.
12 The numbers of retail job vacancies advertised on an online job board each quarter in
each of three consecutive years are shown in the following table.
Year Quarter 1 Quarter 2 Quarter 3 Quarter 4

1 212 194 196 227
2 220 197 196 239
3 231 205 203 245
b Use the data to calculate seasonal indices, rounded to two decimal places.
c Use the seasonal indices to construct a table of the deseasonalised data.
d Construct a time series plot of the deseasonalised data.

The table below shows the number of customers each month at a restaurant and the long
term seasonal indices for the number of customers at the restaurant each month of the year.
The number of customers for August is missing.
Seasonal index 1.36 1.19 1.05 1.01 0.93 0.82 0.75 0.68 0.87 0.9 1.18 1.26
Number of customers 934 836 736 716 649 554 541 598 626 826 873
13 To correct the number of customers in May for seasonality, the actual number of
customers should be:
A increased by 93.0% B decreased by 93.0% C decreased by 7.0%
D decreased 7.5% E increased by 7.5%
14 To correct the number of customers in November for seasonality, the actual number of
customers should be:
A increased by 18.0% B decreased by 84.7% C increased by 15.3%
D decreased 15.3% E decreased by 18.0%
15 If the deseasonalised number of customers for August is 700, the actual number of
customers in that month is closest to:
A 1029 B 768 C 570 D 607 E 476

5D 5E Fitting a trend line and forecasting 295
16 The table below records the monthly average electricity cost (in dollars) for a home.
223 190 253 236 201 189 203 153 143 111 235 307
The seasonal index for August is closest to:

A 1.00 B 0.75 C 1.25 D 0.80 E 0.87
17 The table below shows the room occupancy rate for a chain of hotels over the summer,
autumn, winter and spring quarters for the years 2020–2022.
Season 2020 2021 2022

summer quarter 72.0 71.5 72.0
autumn quarter 72.4 71.9 72.7
winter quarter 68.4 67.7 68.3
spring quarter 72.3 73.0 73.5
Quarterly average 71.3 71.0 71.6
The seasonal index for winter is closest to:

A 0.960 B 0.957 C 1.046 D 0.969 E 1.003
5E Fitting a trend line and forecasting

Learning intentions
I To be able to use the method of least squares to fit a trend line to a time series.
I To be able to use the trend line to make predictions.
I To be able to use seasonal indices to add seasonality to predicted values as appropri-
ate.
Fitting a trend line

If we identify a linear trend in the time series plot, we can use the least squares method to fit
a line to the data to model that trend.
The following example demonstrates fitting a trend line to times series data which shows no
seasonal component.
Example 20 Fitting a trend line
The table below shows the number of female students in Victoria enrolled in at least one
subject in the Mathematics learning area at year 12 over the period 2010–18. Fit a trend
line to the data, and interpret the slope.

Year 2010 2011 2012 2013 2014 2015 2016 2017 2018
Number 76260 78707 79797 79952 78237 80858 81587 83820 84069
1 Construct a time series plot 85000
84000
Number of female students

of the data to confirm that the 83000
trend is approximately linear. 82000
81000
80000
79000
78000
77000
76000
75000
2009 2010 2011 2013 2014 2015 2016 2017 2018 2019
Year
2 Fit a least squares line to the Number of female students = −1 633 580 +
data, giving the coefficients 851.017×year
rounded to six significant
figures.
3 Write down the slope rounded Slope = 851
to the nearest whole number Over the period 2010−2018 on average the number
and interpret. of female students in Victoria enrolled in at least
one subject in the Mathematics learning area at year
12 increased by 851 per year.
Forecasting
Using a trend line fitted to a time series plot to make predictions about future values is
known as trend line forecasting.
Example 21 Forecasting
How many female students in Victoria do we predict being enrolled in at least one
subject in the Mathematics learning area at year 12 in 2026 if the same increasing trend
continues? Give your answer rounded to the nearest whole number.
Substitute 2026 in the equation determined number of female students
using least squares regression, and round = −1 633 580 + 851.017 × year
to the nearest whole number. = −1 633 580 + 851.017 × 2026
= 90 580 to the nearest whole number.
Note: As with any prediction involving extrapolation, the results obtained when predicting well beyond
the range of the data should be treated with caution.

5E Fitting a trend line and forecasting 297
Forecasting taking seasonality into account

When time series data is seasonal, it is usual to deseasonalise the data before fitting the
trend line.
Example 22 Fitting a trend line (seasonality)
The deseasonalised quarterly sales data from Mikki’s shop are shown below.
Quarter 1 2 3 4 5 6 7 8 9 10 11 12
Sales 893 943 955 858 1005 1026 1043 1040 1261 1151 1115 1267
Fit a trend line and interpret the slope.

1 Using a calculator plot the time 1250
series.
2 Fit a least squares line (with Sales
quarter the EV and sales as the
1000
RV).
0
0 1 2 3 4 5 6 7 8 9 10 11 12
Quarter
3 Write down the equation of Sales = 838.0 + 32.07 × quarter
the least squares line with the
intercept and slope rounded to 4
4 Interpret the slope in terms of the Over the 3-year period, on average sales at
variables involved. Mikki’s shop increased by 32.07 sales per
quarter.
Making predictions with deseasonalised data

When using deseasonalised data to fit a trend line, you must remember that the result
of any prediction is a deseasonalised value. To be meaningful, this result must then be
reseasonalised by multiplying by the appropriate seasonal index.
Example 23 Forecasting (seasonality)
What sales do we predict for Mikki’s shop in the winter of year 4? (Because many items
have to be ordered well in advance, retailers often need to make such decisions.)

298 Chapter 5 Investigating and modelling time series 5E
1 Substitute the appropriate value for Deseasonalised sales = 838.0 + 32.07 × quarter
the time period in the equation for = 838.0 + 32.07 × 15
the trend line. Since summer year
= 1319.05
1 is quarter 1, then winter year 4 is
quarter 15.
2 To obtain the actual predicted sales Actual sales prediction for winter of year 4
figure reseasonalise the predicted = 1319.05 × 1.30
value by multiplying this value by = 1714.765
the seasonal index for winter, 1.30.
= 1715 (to the nearest whole number)
Exercise 5E
Fitting a least squares line to a time series plot (no seasonality)

Example 21 1 The data show the number of commencing university students (in thousands) in
Example 20 Australia for the period 2010–2019.
Year 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Number 488 490 510 538 569 569 595 619 632 645
The time series plot of the data is shown below.
a Comment on the 700
plot.
Commencing university students
650
b Fit a least squares
600
regression trend line
to the data, giving 550
the values of the 500

coefficients to 5 450
significant figures,
400
and interpret the 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
slope. Year
c Use this equation to predict the number of students expected to commence

university in Australia in 2030 to the nearest 1000 students.
2 The table below shows the percentage of total retail sales that were made in department
stores over an 11-year period:
Year 1 2 3 4 5 6 7 8 9 10 11
Sales (%) 12.3 12.0 11.7 11.5 11.0 10.5 10.6 10.7 10.4 10.0 9.4
a Using your CAS calculator, construct a time series plot.

b Comment on the time series plot in terms of trend.

5E 5E Fitting a trend line and forecasting 299
c Fit a trend line to the time series plot, find its equation and interpret the slope. Give
your answer rounded to 3 significant figures.
d Use the trend line to forecast the percentage of retail sales which will be made by
department stores in year 15.
3 The median ages of mothers in Australia over the years 2010–2020 are shown below.
Year 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Age 30.7 30.7 30.7 30.8 30.8 30.9 31.1 31.2 31.3 31.4 31.5
a Fit a least squares regression trend line to the data, and interpret the slope. (Give the
values of the coefficients rounded to 3 significant figures.)
b Use the trend line to forecast the average ages of mothers having their first child in
Australia in 2030. Explain why this prediction is not likely to be reliable.
4 The average weekly earnings (in dollars) in Australia during the period 2014–2021 are
given in the following table.
Year 2014 2015 2016 2017 2018 2019 2020 2021

Earnings 1454.10 1483.10 1516.00 1543.20 1585.30 1634.80 1713.90 1737.10
a Fit a least squares regression trend line to the data, and interpret the slope. Give the
values of the coefficients rounded to four significant figures.
b Use this trend relationship to forecast average weekly earnings in 2030. Explain
why this prediction is not likely to be reliable.
Fitting a least-squares line to a time series with seasonality

Example 23 5 The table below shows the deseasonalised quarterly washing-machine sales of a
company over 3 years.
Year 1 Year 2 Year 3

Quarter 1 2 3 4 5 6 7 8 9 10 11 12
Deseasonalised sales 53 51 54 55 64 64 61 63 67 69 68 66
a Use least squares regression to fit a trend line to the data.

Example 22 b Use this trend equation for washing-machine sales, with the seasonal indices below,
to forecast the sales of washing machines in the fourth quarter of year 4.
Quarter 1 2 3 4
Seasonal index 0.90 0.81 1.11 1.18
6 The quarterly seasonal indices for the sales of boogie boards in a surf shop are as
follows.
Seasonal index 1.13 0.47 0.62 1.77

300 Chapter 5 Investigating and modelling time series 5E
The actual sales of the boogie boards over a 2-year period are given in the table.
Year Quarter 1 Quarter 2 Quarter 3 Quarter 4

1 138 60 73 230
2 283 115 163 417
a Use the seasonal indices to calculate the deseasonalised sales figures for this period
to the nearest whole number.
b Use a CAS calculator to plot the actual sales figures and the deseasonalised sales
figures for this period and comment on the plot.
c Fit a trend line to the deseasonalised sales data. Write the slope and intercept
d Use the relationship calculated in c, together with the seasonal indices, to forecast
the sales for the first quarter of year 4 (you will need to reseasonalise here).

7 The number of visitors to an adventure park is seasonal. A least squares regression line
has been fitted to the data, and the equation is:
deseasonalised number of visitors= 38345 + 286.5 × quarter
where quarter number one is January – March 2022.
The quarterly seasonal indices for visitors to the adventure park are shown in the table
below.
Quarter Jan-Mar Apr-Jun Jul-Sept Oct-Dec
Seasonal index 1.17 0.91 0.78 1.14
The predicted number of actual visitors for the April-June quarter in 2025 is closest to:
A 42070 B 42356 C 38544 D 46545 E 37501
8 An electrical goods retailer knows that the sales of air conditioners are seasonal. A
least squares regression line has been fitted to the data collected by the retailer in 2021
and 2022, and the equation is:
deseasonalised number of air conditioners = 197 + 1.2 × month
where month number one is January 2021.
The monthy seasonal indices for air conditioner sales are shown in the table below.
Seasonal index 1.36 1.19 1.05 1.01 0.93 0.82 0.75 0.68 0.87 0.90 1.18 1.26
The predicted number of actual sales for November 2025 is closest to:
A 268 B 227 C 299 D 333 E 316

Review
Time series data Time series data are a special case of bivariate data, where the
Assign- explanatory variable is the time at which the values of the response
ment
variable were recorded.
Time series plot A time series plot is a bivariate plot where the values of the response
variable are plotted in time order. Points in a time series plot are joined
by line segments.
Features to look Trend Cycles

for in a time
Seasonality Structural change
series plot
Possible outliers Irregular (random) fluctuations
Trend A general increase or decrease over a significant period of time in a

times series plot is called a trend.
Cycles Cycles are present when there is a periodic movement in a time series.
The period is the time it takes for one complete up and down movement
in the time series plot. This term is generally reserved for periodic
movements with a period greater than one year.
Seasonality Seasonality is present when there is a periodic movement in a time

series that has a calendar related period – for example, a year, a month,
a week.
Structural Structural change is present when there is a sudden change in the

change established pattern of a time series plot.
Outliers Outliers are present when there are individual values that stand out
from the general body of data.
Irregular Irregular (random) fluctuations are always present in any real-world

(random) time series plot. They include all of the variations in a time series that
fluctuations
we cannot reasonably attribute to systematic changes like trend, cycles,
seasonality, structural change or the presence of outliers.
Smoothing Smoothing is a technique used to eliminate some of the irregular

fluctuations in a time series plot so that features such as trend are more
easily seen.

Review 302 Chapter 5 Investigating and modelling time series
Moving mean In moving mean smoothing, each original data value is replaced by
smoothing the mean of itself and a number of data values on either side. When
smoothing over an even number of data points, centring is required to
ensure the smoothed mean is centred on the chosen point of time.
Moving median Moving median smoothing is a graphical technique for smoothing a

smoothing time series plot using moving medians rather than means.
Seasonal indices Seasonal indices are used to quantify the seasonal variation in a time
series.
Deseasonalise The process of accounting for the effects of seasonality in a time series
is called deseasonalisation.
Reseasonalise The process of a converting seasonal data back into its original form is
called reseasonalisation.
Trend line Trend line forecasting uses the equation of a trend line to make
forecasting predictions about the future.
Skills checklist
list
5A 1 I can construct a time series plot.

5A 2 I can identify trend in a time series plot.

5A 3 I can identify cycles in a time series plot.

5A 4 I can identify seasonality in a time series plot.

5A 5 I can identify structural change in a time series plot.

Review
5A 6 I can identify outliers in a time series plot.

5B 7 I can smooth a time series using moving mean smoothing.

5B 8 I can smooth a time series using moving mean smoothing with centring.

5C 9 I can smooth a time series using moving median smoothing.

5D 10 I can interpret seasonal indices.

5D 11 I can use seasonal indices to deseasonalise and reseasonalise data.

5D 12 I can use seasonal indices to determine percentage change required to
correct for seasonality.

5D 13 I can calculate seasonal indices from 1 year of data.

5D 14 I can calculate seasonal indices from several years of data.

5D 15 I can use seasonal indices to deseasonalise a time series.

5E 16 I can fit a trend line to a time series plot.

5E 17 I can use a trend line to forecast a future value (no seasonality).

5E 18 I can fit a trend line to a time series plot with seasonality.

5E 19 I can use a trend line to forecast a future value (with seasonality).

1 The time series plot below shows quarterly house sales for a real estate agency over a
three year period.
700
600
500
400
House sales
300
200
100
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Quarter
The time series plot is best described as showing

A seasonality only
B irregular fluctuations only
D an increasing trend seasonality and irregular fluctuations
E a decreasing trend with seasonality and irregular fluctuations
2 The time series plot below shows the annual profit (in $000) for a manufacturing
company.
750
Annual profit ($000)
700
650
600
550
1995 2000 2005 2010 2015 2020
Year

A increasing trend
B decreasing trend

Review
D increasing trend with an outlier
E increasing trend with a structural change
Use the following table to answer Questions 3 to 6.
Time period 1 2 3 4 5 6
Data value 2.3 3.4 4.4 2.7 5.1 3.7
3 The three-moving mean for time period 2 is closest to:

A 3.4 B 3.6 C 3.9 D 4.0 E 4.2
4 The five-moving mean for time period 3 is closest to:

A 3.4 B 3.6 C 3.9 D 4.1 E 4.2
5 The two-moving mean for time period 5 with centring is closest to:
A 2.7 B 3.6 C 3.9 D 4.0 E 4.2
6 The four-moving mean for time period 4 with centring is closest to:
A 2.7 B 3.6 C 3.9 D 4.1 E 4.2

The time series plot for hotel room occupancy rate (%) in a large city over a three year
period is shown below.
74.0
72.0
Room occupancy rate (%)
70.0
68.0
66.0
64.0
62.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Quarter
7 The three-median smoothed value for Quarter 2 is closest to:

A 62 B 63 C 64 D 65 E 67
8 The five-median smoothed value for Quarter 3 is closest to:

A 62 B 64 C 65 D 68 E 69

9 The seasonal indices for the number of customers at a restaurant are as follows.
1.0 p 1.1 0.9 1.0 1.0 1.2 1.1 1.1 1.1 1.0 0.7
The value of p is:
A 0.5 B 0.7 C 0.8 D 1.0 E 1.2
Day 1 2 3 4 5 6 7 8 9 10
Price($ ) 2.85 2.80 2.78 2.40 2.80 3.15 3.42 3.95 4.05 3.35
A $2.80 B $2.89 C $2.91 D $2.99 E $3.08
11 If five-mean smoothing was used to smooth this time series, the number of smoothed
values would be:
A 5 B 6 C 7 D 8 E 9
12 Suppose that Lyn spent a total of $427 on dining out over the period from January to
March, and then another $230 over the period April-May. The five-mean smoothed
amount that she spent in March is closest to:
A $115 B $129 C $131 D $142 E $329

The seasonal indices for the number of bathing suits sold at a surf shop are given in the
table.
Quarter Summer Autumn Winter Spring

Seasonal index 1.8 0.4 0.3 1.5
13 The number of bathing suits sold one summer is 432. The deseasonalised number is
closest to:
A 432 B 240 C 778 D 540 E 346
14 The deseasonalised number of bathing suits sold one winter was 380. The actual
number was closest to:
A 114 B 133 C 152 D 380 E 1267

Review
15 The seasonal index for spring tells us that, over time, the number of bathing suits sold
in spring tends to be:
A 50% less than the seasonal average
B 15% less than the seasonal average
C the same as the seasonal average
D 15% more than the seasonal average
E 50% more than the seasonal average
16 To correct for seasonality, the actual number of bathing suits sold in Autumn should be:
A reduced by 50% B reduced by 40% C increased by 40%
D increased by 150% E increased by 250%
17 The number of visitors to an information centre each quarter was recorded for one year.
The results are tabulated below.
Quarter Summer Autumn Winter Spring
Visitors 1048 677 593 998
Using this data, the seasonal index for autumn is estimated to be closest to:
A 0.25 B 1.0 C 1.22 D 0.82 E 0.21

A trend line is fitted to a time series plot displaying the percentage change in commencing
international student enrolments in Australia each year compared to the previous year
(enrolments) for the period 2012–2019.
The equation of this line is: % change in enrolments = −3480 + 1.73 × year
18 Using this trend line, the percentage change in enrolments from the previous year
forecast for 2026 is:
A 24.98 B -11.05 C 1.73 D 12.11 E 24.62
19 From the slope of the trend line it can be said that:

A on average, the number of commencing international student enrolments in
Australia is increasing by by 1.73 each year.
B on average, the number of commencing international student enrolments in
Australia is increasing by 1.73% each year.
C on average, the number of commencing international student enrolments in
Australia is decreasing by 1.73% each month.
D on average, the number of commencing international student enrolments in
Australia is decreasing by 1730 each year.
E on average, the number of commencing international student enrolments in
Australia is decreasing by 1.73% each year.

20 Suppose that the seasonal indices for the wholesale price of petrol are:
Day Sunday Monday Tuesday Wednesday Thursday Friday Saturday

Index 1.2 1.0 0.9 0.8 0.7 1.2 1.2
The equation of the least squares regression line that could enable us to predict the
deseasonalised price per litre in cents from the day number is
deseasonalised price = 189.9 + 0.23 × day number
where day number 1 is Sunday March 20. The predicted actual price for Sunday April
3 is closest to:
A 161.1 cents B 193.1 cents C 193.4 cents D 231.7 cents E 232.0 cents

1 The table below shows the carbon dioxide emissions in Australia (in tonnes per capita)
for the period 2013 to 2020.
Year 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
CO2 18.2 17.6 17.3 17.0 16.4 15.8 15.8 15.9 15.7 15.5
a Use your calculator to construct a times series plot of the data.

b Briefly describe the general trend in the data.
c Fit a least squares line to the time series plot that will enable emissions to be
predicted from year. Write down the equation for the least squares line, rounding
the intercept and slope to four significant figures.
d Use the least squares equation to predict the carbon dioxide emissions in Australia
in 2026. Round to three significant figures.
e Explain why the prediction you made in part d may not be reliable.
2 The table below shows the annual inflation rates in Australia and China for the period
2010–20.
Year 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Inflation Australia (%) 2.9 3.3 1.7 2.5 2.5 1.5 1.3 2.0 1.9 1.6 0.9
Inflation China (%) 3.3 5.4 2.6 2.6 2.0 1.4 2.0 1.6 2.1 2.9 2.4
These data are plotted in the time series plot below.

Review
5
Inflation rate (%)

4
0
2009 2011 2013 2015 2017 2019
Year
Inflation Australia (%) Inflation China (%)
a i Find the equation of the least squares line which allows inflation to be predicted
from year for China.
ii Draw the least squares line on the time series plot.
b i Find the equation of the least squares line which allows inflation to be predicted
from year for Australia.
ii Draw the least squares line on the time series plot.
c Explain why the equations of the least squares lines predict that the inflation rate for
China will always remain higher than the inflation rate for Australia.
d Find the two-mean centred smoothed inflation rate for Australia for 2015.
3 The table below shows the number of dolphins spotted in a bay over each of the four
seasons for the years 2020-2021.

2020 97 112 480 678
2021 107 145 496 730
a Use the data in the table to find seasonal indices. Give your answers rounded to two
decimal places.
b The number of dolphins spotted in each of the four seasons in 2022 is shown in the
table below.

2022 78 86 350 540
Use the seasonal indices from part a to deseasonalise the data. Round your answers

6
Revision
Chapter
Revision: Data analysis

6A Exam 1 style questions: Univariate data 311
Revision
6A Exam 1 style questions: Univariate data
Use the following information to answer Questions 1–3.
The following table shows the data collected from a sample of five senior students at
secondary college. The variables in the table are:
campus – the campus they attend (C = city, R = regional))
time – the time in minutes each students took to get to school that day
transport – mode of transport (1 = walked or rode a bike, 2 = car, 3 = public transport)
number – number of siblings at the school
postcode – postcode of place of residence.
Campus Time Transport Number Postcode

C 12 1 0 7267
R 35 3 2 7268
C 15 2 1 7268
C 43 3 0 7250
R 27 2 3 7267
1 The variables campus and transport are:

A both nominal variables
B a nominal and an ordinal variable respectively
C an ordinal and a nominal variable respectively
D a nominal and a discrete numerical variable respectively
E a nominal and a continuous numerical variable respectively
2 The number of these variables that are discrete numerical is:

A 0 B 1 C 2 D 3 E 4
3 The number of regional students who used public transport to get to school is:
A 0 B 1 C 2 D 3 E 4
4 Consider this box-and-whisker plot. Which one of the following statements is true?
0 10 20 30 40 50 60 70
A The median is 45.

B Less than one-quarter of the observations are less than 30.
C Less than one-quarter of the observations are greater than 50.

Revision 312 Chapter 6 Revision: Data analysis
D All of the observations are less than 60.

E More than half of the observations are less than 30.
5 In a survey people were 30%

asked to respond to the
statement I enjoy my 25%
Percent
work, selecting from the
20%
alternatives ‘strongly
disagree’, ‘agree’, ‘neither 15%
agree nor disagree’,
‘disagree’, ‘strongly 10%
Strongly Agree Neither Disagree Strongly
disagree’. Their responses agree agree nor disagree
are summarised in the bar disagree
chart opposite. I enjoy my work
The percentage of people who chose the modal response to this question is closest to:
A 18% B 20% C 27% D 29% E 31%

A group of VCE mathematics students sat for a 11
10
test. There were 63 students in the group. Their 9
test scores are summarised opposite in the form 8
7
Frequency
of a histogram.
6
5
6 The distribution of test scores is best 4
described as: 3
2
A positively skewed 1
0
B negatively skewed 0 5 10 15 20 25 30 35 40 45 50 55 60
C symmetric Test score
D symmetrically skewed
E symmetric with a clear outlier
7 Displayed in the form of a boxplot, the distribution of test scores would look like:
A B
0 60 0 60
C D
0 60 0 60

Revision
E
0 60
8 Students who scored 50 or more on the test were awarded an A on the test. The
percentage of students who were awarded an A is closest to:
A 10% B 11% C 21% D 30% E 33%
9 The number of students who scored at least 30 but under 45 marks is:
A 4 B 7 C 19 D 11 E 20
10 The modal mark lies in the interval:

A 5−10 B 15−20 C 25−30 D 40−45 E 50−55
11 The median mark lies in the interval:

A 5−10 B 15−20 C 25−30 D 40−45 E 50−55
12 For the boxplot opposite, outliers are

defined as data values that are:
A less than 35 or greater than 75
50 55 60 65 70 75 80
B less than 48 or greater than 70
C less than 48 D greater than 70 E greater than 80

The following stem plot shows the distribution of the time it took (in minutes) for each of a
group of 20 people to travel to work.
Time (minutes) key: 4|0 represents 40

1 6 8
2 2 5 6 8 9
3 0 5 6
4 2 6 7 8
5 1 3 6
6 0 1
7 1
13 The median travel time is:

A 4 minutes B 35 minutes C 36 minutes D 39 minutes E 42 minutes
14 The interquartile quartile range (IQR) of the travel times is:

A 25 minutes B 27 minutes C 28 minutes D 30 minutes E 53 minutes

15 The travel time for the fastest 30% of people was:

A 28 minutes or less B more than 28 minutes C less than 25 minutes
D 51 minutes or less E more than 51 minutes
16 The following histogram shows 20

the annual expenditure on 18
health (in $US) for 72 different 16
countries, plotted on a log10 scale.
14
Based on this histogram, the
Frequency
12
percentage of countries spending
10
$US10,000,000 or more on health
is equal to: 8
6
4
2
0
6.0 6.5 7.0 7.5 8.0 8.5 9.0
log(health expenditure)
A 11.1 B 16.7 C 20.0 D 30.6 E 69.4

To test the temperature 40
control on an oven, the 36
control is set to 180◦ C and
32
the oven is heated for 15
Percentage frequency
28
minutes. The temperature of
the oven is then measured. 24
The temperatures of 300 20
ovens tested in this way 16
were recorded and the
12
data displayed using the
8
percentage frequency
histogram shown. 4
0
174 176 178 180 182 184 186 188
oven temperature
17 The median temperature (in ◦ C) could be

A 176.5 B 179.5 C 181.3 D 182.1 E 185.7
18 The maximum value for the interquartile range (IQR) in ◦ C is

A 2 B 4 C 5 D 6 E 7

Revision
Use the following information to answer Questions 19–24.
The lengths of sample of 1000 ants of a particular species are approximately normally
distributed with a mean of 4.8 mm and a standard deviation of 1.2 mm.
19 From this information it can be concluded that around 95% of the lengths of the ants
should lie between:
A 2.4 mm and 6.0 mm B 2.4 mm and 7.2 mm C 3.6 mm and 6.0 mm
D 3.6 mm and 7.2 mm E 4.8 mm and 7.2 mm
20 The standardised ant length of z = −1.2 corresponds to an actual ant length of:
A 2.40 mm B 3.36 mm C 4.2 mm D 5.0 mm E 6.24 mm
21 The percentage of ants with lengths less than 3.6 mm is closest to:
A 2.5% B 5% C 16% D 32% E 95%
22 The percentage of ants with lengths less than 6.0 mm is closest to:
A 5% B 16% C 32% D 68% E 84%
23 The percentage of ants with lengths greater than 3.6 mm and less than 7.2 mm is
closest to:
A 2.5% B 18.5% C 68% D 81.5% E 97.5%
24 In the sample of 1000 ants, the number with a length between 2.4 mm and 4.8 mm is
approximately:
A 3 B 50 C 475 D 975 E 997
25 The scores on an examination are known to be approximately normally distributed.

If 2.5% of students score more than 68% on the examination, and 0.15% of students
score less than 16% on the exmination, estimate the mean and standard deviation of the
examination scores. Give your answers to one decimal place.
A mean = 47.2, standard deviation = 10.4 B mean = 42.0, standard deviation = 16.0
C mean = 10.4, standard deviation = 47.2 D mean = 34.4, standard deviation = 16.8
E mean = 56.6, standard deviation = 10.4
26 A class of students sat for a biology test and a legal studies test. Each test had a
possible maximum score of 100 marks. The table below shows the mean and standard
deviation of the marks obtained in these tests.
Subject
Biology Legal Studies
Class mean 54 78
Class standard deviation 15 5

The class marks in each subject are approximately normally distributed.

Sashi obtained a mark of 81 in the biology test. The mark that Sashi would need to
obtain on the legal studies test to achieve the same standard score for both legal studies
and biology is:
A 81 B 82 C 83 D 87 E 95
6B Exam 1 style questions: Associations

1 Researchers believe that reaction time might be lower in cold temperatures. They
devise an experiment where reaction time in seconds is measured at three different
temperature levels (1 = less than 8◦ C, 2 = from 8◦ C to 18 ◦ C, 3 = more than 18◦ C).
The explanatory variable, and its classification are:
A reaction time, categorical B temperature, categorical
C reaction time, numerical D temperature, numerical
E temperature, ordinal
Use the following information to answer Questions 2-4
The data in the following table was collected to investigate the association between a
person’s age and their satisfaction with their career choice.
Age group
Satisfied with career choice? Under 35 35 or more Total
Yes 136 136 272
No 42 86 128
Total 178 222 400
2 The percentage of participants in the study aged 35 or more is closest to:

A 55.5% B 50.0% C 44.5% D 68.0% E 61.3%
3 Of those people aged under 35, the percentage who are satisfied with their career
choice is closest to:
A 23.6% B 34.0% C 50.0% D 61.3% E 76.4%
4 The data in the table supports the contention that there is an association between age
group and satisfaction with career choice because:
A 68.0% of people are satisfied with their career choice, compared to 32.0% who are
not.
B The number of people satisfied with their career choice aged under 35 is the same as
the number aged 35 or more who are satisfied with their career choice.

6B Exam 1 style questions: Associations 317
Revision
C 76.4% of people aged under 35 are satisfied with their career choice, which is more
than the 61.3% of those aged 35 or more who are satisfied with their career choice.
D 50.0% of people are satisfied with their career choice are aged under 35.
E 67.2% of people who are not satisfied with their career choice are aged 35 or more.
5 The box plots opposite display the distribution 25

of the average pay rate, in dollars per hour, of 20
workers in 35 countries for the years 1980, 1990
15
Pay rate
and 2000.
The aim is to investigate the association 10
between pay rate and year.
5
0
1980 1990 2000
Year
Which one of the following statements is not true?
A The IQR’s of pay rate in 1990 and 2000 are approximately the same.
B The median pay rate is lower in 1980 than the median pay rate in 1990 and 2000.
C The IQR of pay rate in 1980 is more than the IQR of pay rate in 1990 and 2000.
D The pay rate in 75% of the countries in 1980 was less than the median pay rate in
2000.
E All three distributions are approximately symmetric.
Questions 6 to 8 relate to the following information

In a study of the association between left-handedness and dyslexia, the dominant hand
(left or right) was recorded for two groups of students, a group of students who had been
diagnosed with dyslexia (yes), and a control group (no). The results are summarised in the
percentaged segmented bar chart below.
100%
90% Right handed
80% Left handed
70%
60%
50%
40%
30%
20%
10%
0%
Yes No
Dyslexia
6 The percentage of students diagnosed with dyslexia who are left-handed is closest to:
A 10% B 20% C 30% D 80% E 90%

7 The results could be summarised in a two-way frequency table. Which of the following
frequency tales could match the percentaged segmented bar chart?
Dyslexia Dyslexia
Dominant hand Yes No Dominant hand Yes No
A B
Left 80 90 Left 80 20
Right 20 10 Right 90 10
Dyslexia Dyslexia
Dominant hand Yes No Dominant hand Yes No
C D
Left 20 80 Left 20 10
Right 10 90 Right 30 40
Dyslexia
Dominant hand Yes No
E
Left 10 10
Right 40 90
8 The variables dominant hand and dyslexia appear to be associated because:

A very few students are left-handed
B 20% of students diagnosed with dyslexia are left-handed, compared to only 10% of
the control group
C only 30% of the students are left-handed
D 80% of students diagnosed with dyslexia are left-handed, compared to 90% of the
control group
E a higher percentage of the control group are left-handed compared to the students
diagnosed with dyslexia
9 For which one of the following pairs of variables would it be appropriate to construct a
scatterplot?
A eye colour (blue, green, brown, other) and country of birth
B weight in kg and blood pressure in mmHg
C number of cups of coffee drunk each day and stress level (high, medium, low)
D age in years and football team
E time spent watching TV each week in hours and educational level (primary,
secondary, tertiary)
10 The value of r for the scatterplot is closest to:

A 0.8 B 0.5 C 0
D −0.5 E −0.8

6B Exam 1 style questions: Associations 319
Revision
11 The association pictured in the scatterplot in the previous question is best described as:
A strong, positive, linear B strong, negative, linear C weak, negative, linear
D strong, negative, non-linear with an outlier E strong, negative, non-linear
12 When the correlation coefficient, r, was calculated for 25

the data displayed in the scatterplot, it was found to be
20
r = −0.64.
If the point (1, 5) was replaced with the point (6, 5) 15
and the correlation coefficient, r, recalculated, then the
10
value of r would be:
A unchanged B positive but closer to 1 5
C negative but closer to 0 D positive but closer to 0

E negative but closer to –1 O 1 2 3 4 5 6 7 8
The following information relates for Questions 13 to 15
For the association between between computer ownership (computers/1000 people) and car
ownership (cars/1000 people) the coefficient of determination is equal to 0.8464.
13 If the people who own more cars also tend to own more computers, then the value of
the correlation coefficient, r (rounded to two decimal places) is closest to.
A 0.64 B 0.72 C 0.85 D 0.92 E 0.96
14 The percentage of variation in computer ownership explained by the variation in the

car ownership is closest to:
A 71.6% B 84.6% C 92.0% D 8.0% E 15.4%
15 The percentage of variation in computer ownership NOT explained by the variation in

the car ownership is closest to:
A 71.6% B 84.6% C 92.0% D 8.0% E 15.4%
16 A back-to-back stem plot is a useful tool for displaying the association between:
A weight (kg) and handspan (cm)
B height (cm) and age (years)
C handspan (cm) and eye colour (brown, blue, green)
D height in centimetres and sex (female, male)
E meat consumption (kg/person) population and country of residence
17 To explore the association between owning an electric car (yes or no) and age group
(under 25 years, 25-44 years, 44 years or more), it would be best to use the data
collected to construct:

A an appropriately percentaged table B a back-to-back stem plot

C parallel box plots D a scatterplot
E a histogram
18 The association between the time taken to walk 5 km (in minutes) and fitness level
(below average, average, above average) is best displayed using:
A a histogram B a scatterplot C a time series plot
D parallel box plots E a back-to-back stem plot
6C Exam 1 style questions: Regression and data

transformation
1 A teacher collected the following statistical information about her students’ scores in
their mathematics examination in Year 11 and their scores in Year 12:
Year 11 Year 12
mean 75.1 64.8
The slope of the least squares regression line which would allow their score in Year 12
to be predicted from their score in Year 11 is closest to:
A 0.35 B 0.68 C 1.3 D 1.7 E 3.36
2 The statistical analysis of the set of bivariate data involving variables x and y resulted
in the information displayed in the table below:
x y
mean 123.5 38.7
least squares equation y = −140 + 0.475x
Using this information the value of the correlation coefficient r for this set of bivariate
data is closest to
A 0.73 B 0.34 C 0.46 D 0.49 E 0.97
The following data relate to Questions 3 and 4.
Number of hot dogs sold 190 168 146 155 150 170 185
Temperature (◦ C) 10 15 20 15 17 12 10

6C Exam 1 style questions: Regression and data transformation 321
Revision
3 The equation of the least squares regression line fitted to the data is closest to:
A number of hot dogs sold = 227 − 4.31 × temperature
B number of hot dogs sold = 48.4 − 0.206 × temperature
C number of hot dogs sold = 4.31 + 227 × temperature
D number of hot dogs sold = 0.206 − 48.4 × temperature
E number of hot dogs sold = 227 + 4.31 × temperature
4 The coefficient of determination will be closest to:

A −0.94 B −0.89 C 0.21 D 0.89 E 0.94

Eighteen students sat for a 15-question multiple- 10
choice test. In the scatterplot opposite, the 9
number of errors made by each student on the 8
Number of errors
test is plotted against the time they reported 7
studying for the test. 6
5
A least squares regression line has been 4
determined for the data and is also displayed 3
on the scatterplot. The equation for the least 2
squares regression line is: 1
0
number o f errors = 8.8 − 0.12 × study time 0 10 20 30 40 50 60 70
and the coefficient of determination is 0.8198. Study time (minutes)
5 The least squares regression line predicts that a student reporting a study time of 35
minutes would make:
A 4.3 errors B 4.6 errors C 4.8 errors D 5.0 errors E 13.0 errors
6 The student who reported a study time of 10 minutes made six errors. The predicted
score for this student would have a residual of:
A −7.6 B −1.6 C 0 D 1.6 E 7.6
7 Which of the following statements that relate to the regression line is not true?
A The slope of the regression line is –0.12.
B The equation predicts that a student who spends 40 minutes studying will make
around four errors.
C The least squares line does not pass through the origin.
D On average, a student who does not study for the test will make around 8.8 errors.
E The explanatory variable in the regression equation is number of errors.
8 This regression line predicts that, on average, the number of errors made:
A decreases by 0.82 for each extra minute spent studying
B decreases by 0.12 for each extra minute spent studying
C increases by 0.12 for each extra minute spent studying

D increases by 8.8 for each extra minute spent studying
E decreases by 8.8 for each extra minute spent studying
9 Given that the coefficient of determination is 0.8198, we can say that close to:
A 18% of the variation in the number of errors made can be explained by the variation
in the time spent studying
B 33% of the variation in the number of errors made can be explained by the variation
C 67% of the variation in the number of errors made can be explained by the variation
D 82% of the variation in the number of errors made can be explained by the variation
E 95% of the variation in the number of errors made can be explained by the variation
10 The average rainfall and temperature range 250
Average rainfall (cm)

at several locations in the South Pacific
200
region are displayed in the scatterplot
opposite. 150
100
50
0
0 2 4 6 8 10 12 14 16 18 20
Temperature range (°C)
A least squares regression line has been fitted to the data, as shown. The equation of
this line is closest to:
A average rain fall = 210 − 11 × temperature range
B average rain fall = 210 + 11 × temperature range
C average rain fall = 18 − 0.08 × temperature range
D average rain fall = 18 + 0.08 × temperature range
E average rain fall = 250 − 13 × temperature range
11 In a certain state, correlation coefficient between:

university entrance score and score on a Mathematics aptitude test is r = 0.462
university entrance score and score on an English aptitude test is r = 0.662
Given this information, which one of the following statements is true?
A Around 21.3% of the variation in score on the Mathematics aptitude test is
explained by the variation in score on the English aptitude test.

6C Exam 1 style questions: Regression and data transformation 323
Revision
B Around 43.8% of the variation in score on the English aptitude test is explained by
the variation in score on the Mathematics aptitude test.
C Together the scores on the Mathematics and English aptitude tests explain 100% of
the variation in university entrance score.
D The correlation between the scores on the Mathematics and English aptitude tests is
negative.
E The score on English aptitude tests is more strongly associated with the university
entrance score than is the score on the Mathematics aptitude test.
A student uses the data in the table below to construct the scatterplot shown:
140
x y
1 132 120
2 120 100
3 117 80
4 104 y
60
5 91
40
6 82
20
7 49
8 24 0
0 1 2 3 4 5 6 7 8 9
x
12 A squared transformation is applied to x to linearise the association. A least squares

line is fitted to the transformed data, with x2 as the explanatory variable.
A y = 132 − 1.64x
B y = 79.7 − 0.602x2
C y2 = 79.7 − 0.602x
D y2 = 132 − 1.64x
E y = 132 − 1.64x2
13 A y2 transformation could also be used to linearise this association. A least squares line
is fitted to the transformed data, with y2 as the response variable, and the equation of
the least squares line is
y2 = 20076 − 2397.2x
Using this equation, the predicted value of y when x = 2 is closest to:
A 102 B 124 C 120 D 10487 E 15282

14 The following data were collected for two related variables x and y.
x 0.4 0.5 1.1 1.1 1.2 1.6 1.7 2.3 2.4 3.4 3.5 4.3 4.7 5.3
y 5.8 4.7 3.3 5.5 4.2 3.4 2.3 2.8 1.8 1.3 1.9 1.2 1.6 0.9
A scatterplot indicates a non-linear relationship. The data is linearised using a 1/y
transformation. A least squares line is then fitted to the transformed data.
The equation of this line is closest to:
1 1 1
A = 0.08 + 0.16x B = 0.16 + 0.08x C = −0.08x + 5.23x
y y y
1 1
D = 5.23 − 0.08x E = 1.44 + 1.96x
y y
15 The equation of a least squares line that has been fitted to transformed data is:
population = 58 170 + 43.17 × year2
Using this equation, the predicted value of population when year = 10 is closest to:
A 9.2 B 9.9 C 10.6 D 62 417 E 62 487
16 The equation of a least squares line that has been fitted to transformed data is:
weight2 = 52 + 0.78 × area
Using this equation, the predicted value of weight when area = 8.8 is closest to:
A −7.7 B ±7.7 C 7.7 D ±58 E 58
17 The equation of a least squares regression line that has been fitted to transformed
data is: log(number) = 1.31 + 0.083 × month
Using this equation, the predicted value of number when month = 6 is closest to:
A 1.8 B 6.0 C 18 D 64 E 650
6D Exam 1 style questions: Time series

1 The pattern in the time series in the graph
shown is best described as:
A trend with random variation
B cyclical but not seasonal
C seasonal
D random
E trend with outliers

6D Exam 1 style questions: Time series 325
Revision
2 The time series 16
Percentage of the population aged 65

plot shows the
15.5
percentage of
the population 15
years or more
aged 65 years 14.5
or more in
14
Australia and
New Zealand, 13.5
over the years 13
from 2010 to 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Year
2018.
Australia New Zealand
From the plot, it can be concluded that over the interval 2010–2018, the difference in
the percentage of the population aged 65 years or more in the two countries has shown:
A a decreasing trend B an increasing trend
C seasonal variation D a 5-year cycle
E no trend
Use the information in the table below to answer Questions 3 to 6.
t 1 2 3 4 5 6 7 8 9 10
y 4 5 4 4 8 6 9 10 9 12
3 The three mean smoothed value for t = 2 is closest to:

A 4.3 B 6.2 C 6.4 D 6.5 E 7.25
4 The five mean smoothed value for t = 5 is closest to:
A 4.3 B 6.2 C 6.4 D 6.5 E 7.25
5 The centred two mean smoothed value for t = 6 is closest to:
A 4.3 B 4.75 C 6.25 D 6.5 E 7.25
6 The centred four mean smoothed value for t = 3 is closest to:
A 4.3 B 4.75 C 6.25 D 7.25 E 9.75
7 To help work out her staffing roster, Fleur records the number of customers who come
into her cafe between 7.00 am and 8.00 am each morning for a week (customers).

Customers 42 25 84 100
The numbers of customers on Wednesday, Thursday and Friday are not shown. The
five-mean smoothed number of customers on Thursday is 38.
The three-mean smoothed number of customers on Thursday is:
A 27 B 29 C 30 D 38 E 55

Day 1 2 3 4 5 6 7 8 9
Price($) 1.30 1.15 1.10 1.25 1.29 1.37 2.42 1.95 2.55
A $1.43 B $1.50 C $1.56 D $1.68 E $1.81
The table below records the monthly electricity cost (in dollars) for an apartment over one
calendar year.
123 90 153 136 101 129 153 143 95 61 85 107
9 Based on this information, the seasonal index for September is closest to:
A 1.00 B 0.78 C 1.25 D 0.83 E 0.87
10 Using data collected over several years, the seasonal index for December was
determined to be 0.90. To correct the cost of electricity for seasonality in December,
the actual cost should be
A decreased by 11.1% B decreased by 9.0% C decreased by 10.0%
D increased by 10.0% E increased by 11.1%
Use the information below to answer Questions 11 and 12.

The quarterly sales figures for a soft drink company and the seasonal indices are as shown.
Quarter 1 2 3 4
Sales ($’000s) 1200 1000 800 1200
11 The deseasonalised figure (in $’000s) for quarter 3 is:

A 640 B 667 C 800 D 1000 E 1500
12 The seasonal index for quarter 4 is:

A 0.6 B 0.8 C 1.00 D 1.1 E 1.2
13 The deseasonalised sales (in dollars) for a company in June were $91 564. The
seasonal index for June is 1.45.

6D Exam 1 style questions: Time series 327
Revision
The actual sales for June were closest to:
A $41 204 B $61 043 C $63 148 D $91 564 E $132 768
14 Sales for a major department store are reported quarterly. The seasonal index for the
third quarter is 0.85. This means that sales for the third quarter are typically:
A 85% below the quarterly average for the year
B 15% below the quarterly average for the year
C 15% above the quarterly average for the year
D 18% above the quarterly average for the year
E 18% below the quarterly average for the year
Use the information below to answer Questions 15 and 16.

The time series plot opposite shows the 385
number of calls each month to a call-centre 380
375
Number of calls
over a 12-month period. 370
365
360
355
350
345
0 1 2 3 4 5 6 7 8 9 10 11 12
Month number
15 The three-median smoothed number of calls for month 9 is closest to:

A 358 B 362 C 371 D 375 E 377
16 The five-median smoothed number of calls for month 10 is closest to:

A 358 B 362 C 371 D 375 E 377
17 A time series for y is shown in the graph, where 25

t represents time. If a linear trend line is fitted
20
to this data, as shown, then the equation of the
line is closest to: 15
A y = 20 − 1.6t B y = −1.6t 10
C y = 20 + 1.6t D y = 20 − 0.6t
5
E y = 20 + 0.6t
0
0 1 2 3 4 5 6 7 8 9 10

6E Exam 2 style questions

1 In a large university students in some courses were asked if they would prefer to attend
on-campus lectures and tutorials, or study online and not attend the campus. Data was
collected for the following variables:
number student number
study mode 1 = on-campus, 2 = online
age age in years
course 1 = Business, 2 = Health, 3 = Social Science
gender F = female, M = male
distance the distance the student lives from the campus, to the nearest km
The data collected for 12 students are shown in the following table:
Number Study mode Age (years) Course Gender Distance (kms)

23455 1 18 1 M 2
13425 2 23 1 F 8
28445 1 18 2 M 4
19889 1 19 3 F 9
10340 2 25 2 F 13
23001 2 22 1 M 2
19968 1 19 1 F 7
20012 2 34 3 F 6
21980 1 18 3 M 12
17884 2 45 1 M 8
19456 2 22 2 F 9
21111 1 22 3 F 6
a Write down the names of the numerical variables in the table.

b Determine the mean and the standard deviation of the distance from their residences
to the campus for these 12 students. Round your answers to two decimal places.
c One student, Charlotte, lives 13 km from the campus. Calculate the standardised
score (z) for the distance Charlotte lives from campus. Round your answer to one
decimal place.
d Use the data in the table to complete the following two-way frequency table.
Gender
Study mode Female Male
On campus
Online
Total

6E Exam 2 style questions 329
Revision
e Data was collected from a 100%
total of 120 students. The 90%
80%
percentaged segmented bar 70%
chart shows the study mode 60%
preferences for students from 50%
40%
each of the three courses. 30%
20%
10%
0%
Business Health Social Science
Online On campus
i What percentage of Business students chose online?
ii Does the percentaged segmented bar chart support the contention that the choice
of study mode (on-campus or online) is associated with course? Justify your
answer by quoting appropriate percentages.
2 The histogram and the boxplot below show the distribution of the distances the 120
students surveyed in Question 1 live from the campus.
40
30
Frequency
20
10
0
0 2 4 6 8 10 12 14 16 18 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Distance (kms) Distance (kms)
a Describe the shape of the distribution of distance, including the values of outliers.
b Approximately how many of these students lived from 4 to 5 km from the campus?
c i Determine the values of the upper and lower fences for the boxplot.
ii Use the fences to explain why a distance of 1 km would not be shown as an
outlier.
d The boxplots compare the
distance students live from the online
campus for students who prefer

to study on-campus, and those on-campus
who prefer to study online.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Distance (kms)

Use the information in the boxplots to answer the following questions.

i The median distance for the students preferring to study online was km
higher than the median distance for students preferring to study on-campus.
ii The difference in the IQR for distance for the students preferring to study online
and the IQR for the students preferring to study on-campus is .
3 To address the question "Can we predict a person’s height from their armspan?" height
(in cm) and armspan (in cm) measurements were collected from 60 students in Year
11 and 12. Of the 60 students 30 were male and 30 female. The following scatterplot
shows the data collected, with least squares regression lines fitted for males and
females.
200
female
male
190
height (cm)
180
170
160
150
150 155 160 163 170 175 180 185 190 195 200
armspan (cm)
The equation of the least squares regression line for females is:
height= −4.199 + 1.028 × armspan
The equation of the least squares regression line for males is:
height= 31.705 + 0.815 × armspan
a Interpret the slope of the regression equation in terms of height and armspan for
males.
In determining this equation, the armspan height
summary statistics displayed in
mean (females) 164.0 164.5
the table were also calculated.
standard deviation (females) 6.319 8.083
mean (males) 178.1 177.0
standard deviation (males) 9.832 9.583
b i Determine the value of the coefficient of determination for females and interpret
in terms of height and armspan. Give your answer as a percentage rounded to
one decimal place.
ii Determine the value of the coefficient of determination for males and interpret in
terms of height and armspan. Give your answer as a percentage rounded to one
decimal place.
6E Exam 2 style questions 331
Revision
iii Explain why armspan is a better predictor of height for males than for females,
quoting appropriate statistics.
c i Use the least squares regression line to predict the difference in height between
males and females who both have armspans of 160 cm. Who is taller and by
how much?
ii Use the least squares regression line to predict the difference in height between
males and females who both have armspans of 190 cm. Who is taller and by
how much?
iii Are the prediction made in ci. and cii. reliable? Explain.
4 The average student PISA mathematics scores (score) for OECD countries, as well
as the expenditure per primary school child in those countries in $US per capita
(expenditure), are shown in the following scatterplot.
540
520
PISA mathematics score
500
480
460
440
420
400
380
0 5000 10000 15000 20000 25000
Expenditure primary ($US per capita)
a Describe the association between score and expenditure in terms of form and
strength.
b Which transformations could be used in order to linearise the association?
c A least squares regression line, with expenditure as the explanatory variable, was
fitted to the data, and the following residual plot constructed.
40
20
0
resiual
–20
–40
–60
0 5000 10000 15000 20000 25000

Expenditure primary ($US per capita)
i A residual plot can be used to test an assumption about the nature of the
association between two numerical variables. What is this assumption?
ii Does the residual plot support this assumption? Explain your answer.
d A log10 was applied to the variable expenditure to linearise the association. When
a least squares line was fitted to the transformed data, it was found to have an
intercept of 12.99, and a slope of 120.6.
i Write down the equation of this least squares line.
ii Use the equation from cii to predict the PISA mathematics score for a country
which has an expenditure of $US10 000 per capita. Round your answer to the
nearest whole number.
5 The value of one bitcoin in 20

19
Australian dollars at the 18
beginning of each month in 17
16
2020 is shown in the time 15
Bitcoin ($000’s)
series plot. 14
13
12
11
10
9
8
7
6
5
0 1 2 3 4 5 6 7 8 9 10 11 12
Month
a Determine:
i the 5-median smoothed value for Month 10, rounded to the nearest $000.
ii the 7-median smoothed value for Month 9, rounded to the nearest $000.
b The following table gives the value of bitcoin in Australian dollars at the beginning
of each month for the first six months of 2021.
Month Jan Feb Mar Apr May Jun
Bitcoin($) 29391.78 33543.77 58726.68 57836.01 36681.74 33524.98
Find the centred two-mean smoothed value of bitcoin for the month of March,
rounding your answer to the nearest cent.
c A least squares regression line fitted the monthly bitcoin data for 2021 (January
2021 is month 1), giving the following equation:
Bitcoin = 36382.73 + 1525.799 × month
i Write down the value of the slope to the nearest cent, and interpret in terms of
the variables in the question.
ii Use the equation to predict the value of bitcoin in January 2024. Give your
answer rounded to the nearest dollar.
iii How reliable is the prediction made in cii?

Chapter
7
Modelling growth and
decay using recursion
Chapter objectives
I What is a sequence?
I How do we generate a sequence of numbers from a starting value and a
rule?
I How do we identify particular terms in a sequence?
I What is recursion?
I What is linear growth and decay?
I How can recurrence relations be used to model simple interest, flat rate
depreciation and unit cost depreciation on assets?
I How can recurrence relations be used to model compound interest and
reducing-balance depreciation on assets?
I How can the CAS calculator be used to find the length of time or the
necessary interest rate required for an investment or loan to reach a
particular value?
I How can investments and loans be compared using effective interest rates?
In this chapter, the notion of a sequence and recurrence relation are introduced as well
as the concepts of linear growth and decay and geometric growth and decay.
Taken together, these ideas are applied to financial situations including investments,
loans and the depreciation of assets to investigate how much interest must be paid on a
loan, how much interest an investment earns or how much an asset depreciates, under
different assumptions.

334 Chapter 7 Modelling growth and decay using recursion
7A Sequences and recurrence relations

Learning intentions
I To be able to generate a sequence of terms recursively.
I To be able to generate a sequence of numbers from a worded description using a
calculator.
I To be able to generate a sequence from a recurrence relation.
I To be able to generate a sequence of numbers from a recurrence relation using a
calculator.
I To be able to number and name terms in a sequence.
A list of numbers, written down in succession, is called a sequence. Each of the numbers
in a sequence is called a term. We write the terms of a sequence as a list, separated by
commas. If a sequence continues indefinitely, or if there are too many terms in the sequence
to write them all, we use an ellipsis, ‘. . . ’.
Sequences may be either generated randomly or by recursion using a rule. For example, this
sequence
1, 3, 5, 7, 9, . . .
has a definite pattern.
The sequence of numbers has a starting value of 1. We add 2 to this number to generate the
next term, 3. Then, add 2 again to generate the next term, 5, and so on.
The rule is ‘add 2 to each term’.
+2 +2 +2 +2
1 3 5 7 9 ...
Example 1 Generating a sequence of terms recursively (1)
Write down the first five terms of the sequence with a starting value of 6 and the rule ‘add
4 to the previous term’.
1 Write down the starting value. 6
2 Apply the rule (add 4) to generate the 6 + 4 = 10
next term.
3 Calculate three more terms. 10 + 4 = 14
14 + 4 = 18
18 + 4 = 22
4 Write your answer. The first five terms are 6, 10, 14, 18, 22.

Example 2 Generating a sequence of terms recursively (2)
Write down the first five terms of the sequence with a starting value of 5 and the rule
‘double the number and then subtract 3’.
2 Apply the rule (double 5, then subtract 5×2−3=7
3) to generate the next term.
3 Calculate three more terms. 7 × 2 − 3 = 11
11 × 2 − 3 = 19
19 × 2 − 3 = 35
4 Write your answer. The first five terms are 5, 7, 11, 19, 35.
Using a calculator to generate a sequence of numbers

from a rule
All of the calculations to generate sequences from a rule are repetitive. The same
calculations are performed over and over again – this is called recursion. A calculator can
perform recursive calculations very easily because it automatically stores the answer to the
last calculation it performed, as well as the method of calculation.
Example 3 Generating a sequence of numbers with a calculator
Use a calculator to generate the first five terms of the sequence with a starting value of 5
and the rule ‘double and then subtract 3’.
Steps TI-Nspire
1 Start with a blank computation screen.
2 Type 5 and press · or .
3 Next type ×2 − 3 and press · or
to generate the next term in the sequence.
The computation generating this value is
shown as ‘5·2–3’ on the TI-Nspire and
‘ans × 2 − 3 on the ClassPad (here ‘ans’
represents the answer to the previous
calculation).

4 Pressing · or repeatedly applies ClassPad

the rule ×2 − 3 to the last calculated value.
5 State your answer. The first five terms are 5, 7, 11, 19, 35
Recurrence relations
A recurrence relation is a mathematical rule that we can use to generate a sequence. It has
two parts:
1 a starting value: the value of the first term in the sequence
2 a rule: that can be used to generate the next term from the current term.
For example, in words, a recurrence relation that can be used to generate the sequence:
10, 15, 20, . . .
can be written as follows:
1 Start with 10.
2 To obtain the next term, add 5 to the current term.
A more compact way of communicating this information is to translate this rule into
symbolic form. We do this by defining a subscripted variable. Here we will use the variable
Vn , but the V can be replaced by any letter of the alphabet.
Let Vn be the term in the sequence after n applications of the rule, called iterations.
In words In symbols
Starting value = 10 V0 = 10
Next term = current term +5 Vn+1 = Vn + 5
Using this definition, we can write a formal recurrence relation where the starting value is
defined, followed by the rule for generating the next term.
V0 = 10, Vn+1 = Vn + 5
Note: Because of the way we defined Vn , the starting value of n is 0. At the start there have been no
applications of the rule.

Example 4 Generating a sequence from a recurrence relation
Write down the first five terms of the sequence defined by the recurrence relation
V0 = 29, Vn+1 = Vn − 4
1 Write down the starting value. V0 = 29
2 Use the rule to find the next V1 = V0 − 4
term, V1 . = 29 − 4
= 25
3 Use the rule to determine V2 = V1 − 4 V3 = V2 − 4 V4 = V3 − 4

three more terms. = 25 − 4 = 21 − 4 = 17 − 4
= 21 = 17 = 13
4 Write your answer. The first five terms are 29, 25, 21, 17, 13
Example 5 Using a calculator to generate sequences from recurrence relations
A sequence is generated by the recurrence relation

V0 = 300, Vn+1 = 0.5Vn − 9
Use your calculator to generate this sequence and determine how many terms at the start
of the sequence are positive.
1 Start with a blank computation
300 300.
screen.
2 Type 300 and press · (or 300 · 0.5 − 9 141.
). 141 · 0.5 − 9 61.5
3 Next type ×0.5 − 9 and press 61.5 · 0.5 − 9 21.75
· (or ) to generate the 21.75 · 0.5 − 9 1.875
next term in the sequence. 1.875 · 0.5 − 9 −8.625
4 Continue to press · (or |
) until the first negative
term appears. The first five terms of the sequence are positive.
5 Write your answer.

338 Chapter 7 Modelling growth and decay using recursion 7A
Example 6 Naming terms in a sequence
Consider the recurrence relation

V0 = 3, Vn+1 = Vn + 6
State the values of:
a V1 b V4 c V5
1 Write the name for each term under its
3, 9, 15, 21, 27, 33
value in the sequence.
V0 V1 V2 V3 V4 V5
2 Read the value of each required term. V1 = 9, V4 = 27 V5 = 33
Exercise 7A
Generating a sequence recursively

Example 1 1 Use the following starting values and rules to generate the first five terms of the
following sequences recursively.
a Starting value: 2 b Starting value: 5
Rule: add 6 Rule: subtract 3
c Starting value: 1 d Starting value: 64
Rule: multiply by 4 Rule: divide by 2
following sequences recursively.
a Starting value: 6 b Starting value: 24
Rule: multiply by 2 then add 2 Rule: multiply by 0.5 then add 4
c Starting value: 1 d Starting value: 124
Rule: multiply by 3 then subtract 1 Rule: multiply by 0.5 then subtract 2
following sequences recursively using a calculator.
a Starting value: 4 b Starting value: 24 c Starting value: 2
Rule: add 2 Rule: subtract 4 Rule: multiply by 3
d Starting value: 50 e Starting value: 5 f Starting value: 18
Rule: divide by 5 Rule: multiply by 2 then Rule: multiply by 0.8
add 3 then add 2

7A 7A Sequences and recurrence relations 339
Generating sequences using recurrence relations

Example 4 4 Write down the first five terms of the sequences generated by each of the recurrence
relations below.
a W0 = 2, Wn+1 = Wn + 3 b D0 = 50, Dn+1 = Dn − 5
c M0 = 1, Mn+1 = 3Mn d L0 = 3, Ln+1 = −2Ln
e K0 = 5, Kn+1 = 2Kn − 1 f F0 = 2, Fn+1 = 2Fn + 3
g S 0 = −2, S n+1 = 3S n + 5 h V0 = −10, Vn+1 = −3Vn + 5
Example 5 5 Using your calculator, write down the first five terms of the sequence generated by each
of the recurrence relations below.
a A0 = 12, An+1 = 6An − 15 b Y0 = 20, Yn+1 = 3Yn + 25
c V0 = 2, Vn+1 = 4Vn + 3 d H0 = 64, Hn+1 = 0.25Hn − 1
e G0 = 48 000, Gn+1 = Gn − 3000 f C0 = 25 000, Cn+1 = 0.9Cn − 550
Example 6 6 Consider the following recurrence relations. Find the required term for each.
a A0 = 2, An+1 = An + 2. Find A2 .
b B0 = 11, Bn+1 = Bn − 3. Find B4
c C0 = 1, Cn+1 = 3Cn . Find C3
d D0 = 3, Dn+1 = 2Dn + 1. Find D5
7 Write a recurrence relation for each of the following worded descriptions.

a Starting value: 4 b Starting value: 24 c Starting value: 2
Rule: add 2 Rule: subtract 4 Rule: multiply by 3
8 State a recurrence relation that could be used to generate each of the following
sequences.
a 5, 10, 15, 20, 25, . . . b 13, 9, 5, 1, −3, . . .
c 1, 4, 16, 64, 256, . . . d 64, 32, 16, 8, 4, . . .
Exploring sequences with a calculator

9 How many terms of the sequence formed from the recurrence relation below are positive?
F0 = 150, Fn+1 = 0.6Fn − 5
10 How many terms of the sequence formed from the recurrence relation below are negative?
Y0 = 30, Yn+1 = 1.2Yn + 2

11 A sequence of numbers is generated by the recurrence relation shown below
A0 = 3, An+1 = 4An + 1
The value of A4 is
A 3 B 4 C 13 D 213 E 853

340 Chapter 7 Modelling growth and decay using recursion 7A
12 The following recurrence relation can generate a sequence of numbers

A0 = 15, An+1 = An + 4
The number 51 appears in this sequence as
A A1 B A7 C A8 D A9 E A10
13 The first five terms of a sequence are

3, 7, 15, 31, 63
The recurrence relation that generates this sequence could be
A B0 = 3, Bn+1 = Bn + 4 B B0 = 3, Bn+1 = Bn + 8
C B0 = 3, Bn+1 = 3Bn − 1 D B0 = 3, Bn+1 = 4Bn − 5
E B0 = 3, Bn+1 = 2Bn + 1
7B Modelling linear growth and decay

Learning intentions
I To be able to graph the terms of a linear growth/decay sequence.
I To be able to model simple interest loans and investments using recurrence relations.
I To be able to use a recurrence relation to analyse a simple interest investment.
I To be able to model and analyse flat rate depreciation using a recurrence relation.
I To be able to model and analyse unit cost depreciation using a recurrence relation.
Linear growth means a value is increasing by the same amount in each unit of time. For
example, if you have $300 in your bank account and you add $20 each week, then your
savings will have linear growth. Similarly, linear decay is characterised as decreasing by the
same amount in each unit of time. For example, the depreciation of a new car by a constant
amount each year.
A recurrence model for linear growth and decay

The recurrence relations
P0 = 20, Pn+1 = Pn + 2
Q0 = 20, Qn+1 = Qn − 2
both have rules that generate sequences with linear patterns, as can be seen from the table
below. The first generates a sequence whose successive terms have a linear pattern of
growth, and the second a linear pattern of decay.

7B Modelling linear growth and decay 341
Recurrence relation Rule Sequence Graph

P0 = 20, Pn+1 = Pn + 2 ‘add 2’ 20, 22, 24, . . . Pn
35
30
25
20
10
n
0 1 2 3 4 5
Q0 = 20, Qn+1 = Qn − 2 ‘subtract 2’ 20, 18, 16, . . . Qn
35
30
25
20
10
n
0 1 2 3 4 5
As a general rule, if D is a positive constant, a recurrence relation rule of the form:

Vn+1 = Vn + D can be used to model linear growth.
Vn+1 = Vn − D can be used to model linear decay.
We refer to D as the common difference and can graph the sequence to obtain a straight
line graph of dots (do not join the dots). An upward slope indicates growth and a downward
slope reveals decay.
Example 7 Graphing the terms of linear growth/decay sequence
For each of the following recurrence relations, list the first four terms and graph the
corresponding points.
a V0 = 2, Vn+1 = Vn + 5
b W0 = 20, Wn+1 = Wn − 3
a From the rule, the starting value is 2. The first four terms are 2, 7, 12, 17.
Vn
The rule is ‘add 5’.
The corresponding points can then be 20
graphed. 15
10
n
0 1 2 3

b From the rule, the starting value is 20. The first four terms are 20, 17, 14, 11.
Vn
The rule is ‘subtract 3’.
The corresponding points can then be 25
graphed. 20
15
10
n
0 1 2 3
Simple interest loans and investments

Simple interest is an example of linear growth in which the starting value is the amount
borrowed or invested. The amount borrowed or invested is called the principal. The amount
added at each step is the interest and is usually a percentage of this principal, found by
multiplying the annual interest rate r% by the principal for each year of the loan.
Recurrence model for simple interest

Let Vn be the value of the loan or investment after n years and r be the annual percentage
interest rate.
The recurrence relation for the value (or balance) of the loan or investment after n
years is
V0 = principal, Vn+1 = Vn + D
r
where D = × V0 .
100
Example 8 Modelling simple interest investments with a recurrence relation
Cheryl invests $5000 in an investment account that pays 4.8% per annum simple interest.
Model this simple investment using a recurrence relation of the form:
r
V0 = the principal, Vn+1 = Vn + D, where D = V0 .
100
Let Vn be the value of the investment after n years.
1 Write down the value of V0 . V0 = 5000
2 Write down the interest rate r and use it r = 4.8
r
to determine the value of D = V0 . 4.8
100 D= × 5000 = 240
100
3 Use the values of V0 and D to write V0 = 5000, Vn+1 = Vn + 240.
down the recurrence relation.

Once we have a recurrence relation, we can use it to determine the value of an investment
after a given number of years.
Example 9 Using a recurrence relation to analyse a simple interest investment
Cheryl’s simple interest investment is modelled by

V0 = 5000, Vn+1 = Vn + 240
where Vn is the value of the investment after n years.
a Use the recurrence relation to show that the value of Cheryl’s investment after 3 years
is $5720.
b When will Cheryl’s investment first exceed $6000, and what will its value be then?
a Calculate V0 , V1 , V2 and V3 .
V0 = 5000
V1 = 5000 + 240 = 5240
V2 = 5240 + 240 = 5480
V3 = 5480 + 240 = 5720
Thus, after three years, the value of
Cheryl’s investment is $5720.
b i On a blank calculation screen, type 5000
and press · (or ). 5000 5000.
ii Type +240 and press · (or ) until 5000. + 240 5240.
the value of the investment first exceeds
5240. + 240 5480.
$6000.
5480. + 240 5720.
iii Count the number of times that 240 was
5720. + 240 5960.
added. Write your answer.
5960. + 240 6200.
After 5 years; $6200.
Depreciation
For some large items, their value decreases over time. This is called depreciation.
Businesses take into account the impact of depreciation by tracking the likely value of an
asset at a point in time, called the future value. At some point in time or at a particular
value, called the scrap value, the item will be sold or disposed of as it is no longer useful to
the business.
There are a number of techniques for estimating the future value of an asset. Two of them,
flat rate depreciation and unit cost depreciation, can be modelled using linear decay
recurrence relations.

Flat rate depreciation

Flat rate depreciation is an example of linear decay where a constant amount is subtracted
from the value of the asset each time period. This constant amount is called the depreciation
amount and is often given as a percentage of the initial purchase price of the asset. The
scrap value is the value at which the item is no longer of use to the business.
Recurrence model for flat-rate depreciation

Let Vn be the value of the asset after n years and r be the percentage depreciation rate.
The recurrence relation for the value of the asset after n years is
V0 = initial value of the asset, Vn+1 = Vn − D
r
where D = × V0 .
100
Example 10 Modelling flat rate depreciation with a recurrence relation
A new car was purchased for $24 000 in 2014. The car depreciates by 20% of its
purchase price each year. Model the depreciating value of this car using a recurrence
relation of the form:
r
V0 = initial value, Vn+1 = Vn − D, where D = V0
100
Let Vn be the value of the car after n years depreciation.
1 Write down the value of V0 . Here, V0 is V0 = 24 000
the value of the car when new.
2 Write down the annual rate of r = 20
depreciation, r, and use it to determine 20
r D= × 24 000 = 4800
the value of D = V0 . 100
100
3 Use the values of V0 and D to write V0 = 24 000, Vn+1 = Vn − 4800
Once we have a recurrence relation, we can use it to determine things such as the value of an
asset after a given number of years of flat rate depreciation.
Example 11 Using a recurrence relation to analyse flat rate depreciation
The flat rate depreciation of a car is modelled by

V0 = 24 000, Vn+1 = Vn − 4800
where Vn is the value of the car after n years.
a Use the model to determine the value of the car after 2 years.
b If the car was purchased in 2023, in what year will the car’s value depreciate to zero?
c What was the percentage depreciation rate?

a i Write down the recurrence relation. V0 = 24 000, Vn+1 = Vn − 4800
ii On a blank calculation screen, type
24000 24000.
24 000 and press · (or ).
24000. − 4800 19200.
iii Type –4800 and press · (or ) 19200. − 4800 14400.
twice to obtain the value of the car 14400. − 4800 9600.
after 2 years’ depreciation. 9600. − 4800 4800.
Write your answer. 4800. − 4800 0.
b i Continue pressing · (or )
until the car has no value. a $14 400
ii Write your answer.
b In 2028
c Use the amount of depreciation and 4800
c × 100% = 20%
initial value. 24000
The percentage depreciation
rate is 20%
Unit cost depreciation

Some items lose value because of how often they are used. A photocopier that is 2 years
old but has never been used could still be considered to be in ‘brand new’ condition and
therefore worth the same as it was 2 years ago. But if that photocopier was 2 years old and
had printed many thousands of pages over those 2 years, it would be worth much less than
its original value.
When the future value of an item is based upon usage, we use a unit cost depreciation
method. Unit cost depreciation can be modelled using a linear decay recurrence relation.
Recurrence model for unit-cost depreciation

Let Vn be the value of the asset after n units of use and D be the cost per unit of use.
The recurrence relation for the value of the asset after n units of use is:
V0 = initial value of the asset, Vn+1 = Vn − D
Example 12 Modelling unit cost depreciation with a recurrence relation
A professional gardener purchased a lawn mower for $270. The mower depreciates in
value by $3.50 each time it is used.
a Model the depreciating value of this mower using a recurrence relation of the form:
V0 = initial value, Vn+1 = Vn − D

where D is the depreciation in value per use and Vn is the value of the mower after
being used to mow n lawns.

346 Chapter 7 Modelling growth and decay using recursion 7B
b Use the model to determine the value of the mower after it has been used three times.
c How many times can the mower be used until its depreciated value is first less
than $250?
a 1 Write down the value of V0 . Here, V0 = 270
V0 is the value of the mower when
new.
2 Write down the unit cost rate of D = 3.50
depreciation, D.
3 Write your answer. V0 = 270, Vn+1 = Vn − 3.50
b 1 Write down the recurrence relation. V0 = 270, Vn+1 = Vn − 3.50
2 On a blank calculation screen, type
270 and press · (or ). 270 270.
Type –3.50 and press · (or ) 270. − 3.5 266.5
three times to obtain the value of the
266.5. − 3.5 263.
mower after three mows.
263. − 3.5 259.5
259.5 − 3.5 256.
256. − 3.5 252.5
252.5 − 3.5 249.
3 Write your answer. $259.50

c 1 Continue pressing · (or )
until the value of the lawn mower is
first less than $250.
2 Write your answer. After six mows
Skill-
sheet Exercise 7B
Modelling linear growth and decay using recurrence relations

Example 7 1 For each of the following recurrence relations, write down the first four terms and
graph the corresponding points.
a V0 = 3, Vn+1 = Vn + 2
b V0 = 38, Vn+1 = Vn − 5
Modelling and analysing simple interest with recurrence relations

Example 8 2 Ashwin invests $8000 in an account that pays 4% per annum simple interest.
a Let Vn be the value of Ashwin’s investment after n years. State the starting value,
V0 , given by the principal.
7B 7B Modelling linear growth and decay 347
r
b Calculate the value of D using the interest rate and the rule D = V0 to find the
100
amount of interest paid each year.
c Model this simple investment using a recurrence relation of the form
V0 = starting value, Vn+1 = Vn + D
3 Huang invests $41 000 in an account that pays 6.2% per annum simple interest.
a Let Hn be the value of Huang’s investment after n years. State the value of H0 .
b Find the amount, in dollars, that Huang will receive each year from the investment.
c Complete the recurrence relation, in terms of H0 , Hn+1 and Hn , that would model the
investment over time. Write your answers in the boxes below.
H0 = , Hn+1 = Hn +
Example 9 4 The following recurrence relation can be used to model a simple interest investment of
$2000, paying interest at the rate of 3.8% per annum.
V0 = 2000, Vn+1 = Vn + 76
In the recurrence relation, Vn is the value of the investment after n years.
a Use the recurrence relation to show that the value of the investment after 3 years is
$2228.
b Use your calculator to determine how many years it takes for the value of the
investment to first be worth more than $3000.
5 The following recurrence relation can be used to model a simple interest loan of $7000
with interest charged at the rate of 7.4% per annum.
V0 = 7000, Vn+1 = Vn + 518
In the recurrence relation, Vn is the value of the loan after n years.
a Use the recurrence relation to find the value of the loan after 1, 2 and 3 years.
b Use your calculator to determine how many years it takes for the value of the loan to
first have a value of more than $10 000.
6 The following recurrence relation can be used to model a simple interest investment. In
the recurrence relation, Vn is the value of the investment after n years.
V0 = 15 000, Vn+1 = Vn + 525
a i What is the principal of this investment?
ii How much interest is earned each year?
iii Calculate 525 as a percentage of 15 000 to find the annual interest rate of this
investment.
b State how many years it takes for the value of the investment to first exceed $30 000.
Modelling flat-rate depreciation with recurrence relations

Example 10 7 Fernando purchased a cherry picker for $82 000 in 2022. The cherry picker depreciates
by 15% of its purchase price each year.
Let Cn be the value of the cherry picker n years after it was purchased.Cambridge University Press
© Peter Jones et al 2023
ISBN 978-1-009-11041-9
r
a Calculate the value of D using the interest rate and the rule D = V0 to find the
100
amount of depreciation each year.
b Model this simple investment using a recurrence relation of the form
C0 = starting value, Cn+1 = Cn − D
8 Wendy purchases a new chair for her dental surgery for $2800. The chair depreciates
by 8% of its purchase price each year.
a Show that the total amount, in dollars, that Wendy’s chair will depreciate by each
year is $244.
b Let Wn be the value of Wendy’s chair after n years. State the value of W0 .
c Complete the recurrence relation, in terms of W0 , Wn+1 and Wn , that would model
the investment over time by filling in the boxes below.
W0 = , Wn+1 = Wn +
Example 11 9 The following recurrence relation can be used to model the depreciation of a computer
with purchase price $2500 and annual depreciation of $400.
V0 = 2500, Vn+1 = Vn − 400
In the recurrence relation, Vn is the value of the computer after n years.
a Use the recurrence relation to find the value of the computer after 1, 2 and 3 years.
b Use your calculator recursively to determine how many years it takes for the value
of the computer to first be worth less than $1000.
10 The following recurrence relation can be used to model the depreciation of a car
purchased for $23 000 and depreciated at 3.5% of its original value each year.
V0 = 23 000, Vn+1 = Vn − 805
In the recurrence relation, Vn is the value of the car after n years.
a Use the recurrence relation to find the value of the car after 1, 2 and 3 years.
b Determine how many years it takes for the value of the car to first be worth less
than $10 000.
11 The following recurrence relation can be used to model the depreciation of a television.
In the recurrence relation, Vn is the value of the television after n years.
V0 = 1500, Vn+1 = Vn − 102
a i What is the purchase price of this television?
ii What is the depreciation of the television each year?
iii What is the annual percentage depreciation of the television?
b Use your calculator to determine the value of the television after 8 years.
c If the owner of the television decides to discard the television once it is first worth
less than $100, determine how long the owner will own the television before
discarding it.

7B 7B Modelling linear growth and decay 349
Modelling unit-cost depreciation with recurrence relations

Example 12 12 A minibus was purchased for $32 600 to take passengers to and from the airport. The
minibus depreciates by $10 on every round trip that it takes.
Let Mn be the value of the minibus after n round trips.
a State the starting value, M0 of the minibus.
b Model the value of the minibus using a recurrence relation of the form
M0 = starting value, Mn+1 = Mn − D
13 The following recurrence relation can be used to model the depreciation of a printer
with purchase price $450 and depreciation of 5 cents for every page printed.
V0 = 450, Vn+1 = Vn − 0.05
In the recurrence relation, Vn is the value of the printer after n pages are printed.
a Write the first five terms of the sequence.
b Use your calculator to find the value of the printer after 20 pages are printed.
14 The following recurrence relation can be used to model the depreciation of a delivery
van with purchase price $48 000 and depreciation by $200 for every 1000 kilometres
travelled.
V0 = 48 000, Vn+1 = Vn − 200
In the recurrence relation, Vn is the value of the delivery van after n lots of 1000
kilometres are travelled.
a Use the recurrence relation to find the value of the van after 1000, 2000 and 3000
kilometres.
b Use your calculator to determine the value of the van after 15 000 kilometres.
c Use your calculator to determine how many kilometres it takes for the value of the
van to reach $43 000.
15 Jasmine owns a cafe that sells juices. The commercial blender, purchased for $1440,
depreciates in value using the unit cost method.
The rate of depreciation is $0.02 per juice that is produced. The recurrence relation that
models the year-to-year value, in dollars, of the blender is
B0 = 1440, Bn+1 = Bn − 144
a Calculate the number of juices that the blender produces each year.
b Determine how many juices the blender can produce before its value becomes 0.
c Use your calculator to find the value of the blender after 36 000 juices have been
produced.
d The recurrence relation above could also represent the value of the blender
depreciating at a flat rate. What annual flat rate percentage of depreciation is
represented?


16 A coffee machine was purchased for $720.
After five years the coffee machine has a value of $586. On average, 670 coffees were
made each year during those five years.
The value of the coffee machine was depreciated using a unit cost method of
depreciation. The depreciation in the value of the coffee machine, per coffee made, is,
in cents, closest to
A 2 B 3 C 4 D 5 E 6
17 The value of a tandoori oven is depreciated using the flat rate method and can be
modelled using the following recurrence relation where T n is the value of the oven after
n years.
T 0 = 4500, T n+1 = T n − 405
The annual depreciation rate is closest to
A 8% B 8.5% C 9% D 9.5% E 10%
18 Jane purchased a motorbike for $5500. She will depreciate the value of her motorbike
by a flat rate of 10% of the purchase price per annum.
A recurrence relation that Jane can use to determine the value of the motorbike, Vn ,
after n years is
A V0 = 5500, Vn+1 = Vn + 550
B V0 = 5500, Vn+1 = Vn − 550
C V0 = 5500, Vn+1 = 0.9Vn
D V0 = 5500, Vn+1 = 1.1Vn
E V0 = 5500, Vn+1 = 0.2(Vn − 550)
7C Using an explicit rule for linear growth or decay

Learning intentions
I To be able to convert a recurrence relation to an explicit rule.
I To be able to model a simple interest investment using an explicit rule.
I To be able to use a rule to determine the value of a simple interest loan or investment.
I To be able to model flat rate depreciation of an asset using an explicit rule.
I To be able to use a rule for the flat rate depreciation of an asset.
I To be able to use an explicit rule for unit cost depreciation.
While we can generate as many terms of a sequence as we like through repeated addition
and subtraction, the process can be tedious and so instead a rule can be used.

7C Using an explicit rule for linear growth or decay 351
Consider the example of investing $2000 in a simple interest investment paying 5% per
annum. If we let Vn be the value of the investment after n years, we can use the following
recurrence relation to model this investment:
V0 = 2000, Vn+1 = Vn + 100
Using this recurrence relation we can write out the sequence of terms generated as follows:
V0 = 2000 = V0 + 0 × 100 (no interest paid yet)
V1 = V0 + 100 = V0 + 1 × 100 (after 1 year of interest paid)
V2 = V1 + 100 = (V0 + 100) + 100 = V0 + 2 × 100 (after 2 years of interest paid)
V3 = V2 + 100 = (V0 + 2 × 100) + 100 = V0 + 3 × 100 (after 3 years of interest paid)
V4 = V3 + 100 = (V0 + 3 × 100) + 100 = V0 + 4 × 100 (after 4 years of interest paid)
and so on.
Following this pattern, after n years of interest has been added, we can write:
Vn = 2000 + n × 100
This rule can be used to determine the value after n iterations in the sequence. For example,
using this rule, the value of the investment after 15 years would be:
V15 = 2000 + 15 × 100 = $3500
Explicit rule for linear growth

For a recurrence rule for linear growth of the form:
V0 = initial value, Vn+1 = Vn + D (D constant)
the value of the term Vn in the sequence generated by this recurrence relation is:
Vn = V0 + nD
Explicit rule for linear decay

In general, for a recurrence rule for linear decay of the form:
V0 = initial value, Vn+1 = Vn − D (D constant)
the value of the term Vn in the sequence generated by this recurrence relation is:
Vn = V0 − nD

Example 13 Converting a recurrence relation to an explicit rule
Write down a rule for Vn for each of the following recurrence relations. Calculate V10 for
each case.
a V0 = 8, Vn+1 = Vn + 3
b V0 = 400, Vn+1 = Vn − 12
c V0 = 30, Vn+1 = Vn − 7
a 1 Identify the starting value. V0 = 8
2 Identify the common difference, D. D=3
3 Write the rule for Vn , noting that this Vn = 8 + 3n
is an example of linear growth.
4 Calculate V10 . V10 = 8 + 3 × 10 = 38
b 1 Identify the starting value. V0 = 400
2 Identify the common difference, D. D = 12
3 Write the rule for Vn , noting that this Vn = 400 − 12n
is an example of linear decay.
4 Calculate V10 . V10 = 400 − 12 × 10 = 280
c 1 Identify the starting value. V0 = 30
2 Identify the common difference, D. D=7
3 Write the rule for Vn , noting that this Vn = 30 − 7n
is an example of linear decay.
4 Calculate V10 . V10 = 30 − 7 × 10 = −40
These general rules can be applied to simple interest investments and loans, flat rate
depreciation and unit cost depreciation.
Using a rule for simple interest loans or investments

Simple interest loans and investments are examples of linear growth so we use the rule
r
Vn = V0 + nD, where D = × V0 .
100

Example 14 Modelling simple interest investments
Amie invests $3000 in a simple interest investment with interest paid at the rate of 6.5%
per year.
Use a rule to find the value of the investment after 10 years.
1 Identify the starting value. V0 = 3000
6.5
2 Identify the common difference, D. D= × 3000 = 195
100
3 Write the rule for Vn , noting that this is Vn = 3000 + 195n
an example of linear growth.
4 Calculate V10 . V10 = 3000 + 195 × 10 = 4950
Example 15 Using a rule to determine the value of a simple interest investment
The following recurrence relation can be used to model a simple interest investment:
V0 = 3000, Vn+1 = Vn + 260
a What is the principal of the investment? How much interest is added each year?
b Write down the rule for the value of the investment after n years.
c Use a rule to find the value of the investment after 15 years.
d Use a rule to find when the value of the investment first exceeds $10 000.
a These values can be read directly from a Principal: $3000
the recurrence relation. Amount of interest = $260
b Start with the general rule: b Vn = 3000 + n × 260
Vn = V0 + nD and substitute V0 = 3000 = 3000 + 260n
and D = 260.
c Substitute n = 15 into the rule to c V15 = 3000 + 260 × 15
calculate V15 . = $6900
d Substitute Vn = 10 000 into the rule, d 10 000 = 3000 + 260n
and solve for n. Write your conclusion. so 7000 = 260n
Note: Because the interest is only paid into
or n = 7000/260
the account after a whole number of years, any
decimal answer will need to be rounded up to = 26.92 . . . years
the next whole number. The value of the investment will first
exceed $10 000 after 27 years.

Using a rule for flat rate depreciation of assets

Flat rate depreciation is an example of linear decay so we use the rule Vn = V0 − nD.
Example 16 Modelling flat rate depreciation of an asset using an explicit rule
A photocopier costs $6000 when new. Its value depreciates at the flat rate of 17.5% per
year. Write a rule and use this to find its value after 4 years.
1 Identify the starting value. V0 = 6000
17.5
2 Identify the common difference, D. D= × 6000 = 1050
100
3 Write the rule for Vn , noting that this is Vn = 6000 − 1050n
an example of linear decay.
4 Calculate V4 . V4 = 6000 − 1050 × 4 = 1800
The value after 4 years is $1800.
Example 17 Using a rule for the flat rate depreciation of an asset
The following recurrence relation can be used to model the flat rate of depreciation of a
set of office furniture:
V0 = 12 000, Vn+1 = Vn − 1200
where Vn is the value of the furniture after n years.
a What is the initial value of the furniture? How much does the furniture decrease by
each year?
b Write down the rule for the value of the investment after n years.
c Use a rule to find the value of the investment after 6 years.
d How long does it take for the furniture’s value to decrease to zero?
a These values can be read directly from the a Initial value: $12 000
recurrence relation. Depreciation = $1200 each year
b Start with the general rule Vn = V0 − nD and b Vn = 12 000 − n × 1200
substitute V0 = 12 000 and D = 1200. = 12 000 − 1200n
c Use the rule to calculate V6 . c V6 = 12 000 − 1200 × 6
= $4800
d Substitute Vn = 0, and solve for n. Write d 0 = 12 000 − n × 1200
your conclusion. so n = 10
The value of the furniture will
depreciate to zero after 10 years.

Using a rule for unit cost depreciation of assets

Example 18 Using an explicit rule for unit cost depreciation
A hairdryer in a salon was purchased for $850. The value of the hairdryer depreciates
by 25 cents for every hour it is in use.
Let Vn be the value of the hairdryer after n hours of use.
a Write down a rule to find the value of the hairdryer after n hours of use.
b What is the value of the hairdryer after 50 hours of use?
c On average, the salon will use the hairdryer for 17 hours each week. How many weeks
will it take for the value of the hairdryer to halve?
d The hairdryer has a scrap value of $100 before it is disposed of. Find the number of
hours of use before this occurs.
a 1 Identify the values of V0 and D. V0 = 850 and D = 0.25
2 Write down the rule for the value of Vn = 850 − 0.25n
the hairdryer after n hours of use.
b 1 Decide the value of n and substitute After 50 hours of use, n = 50.
into the rule. V50 = 850 − 0.25 × 50
V50 = 837.50
2 Write your answer. After 50 hours of use, the hairdryer has a
value of $837.50.
c 1 Halving the value of the hairdryer Solve Vn = 425
means it will have a value of $425.
2 Write down the rule, with the value 425 = 850 − 0.25n
of the hairdryer, Vn = 425.
3 Solve the equation for n. 0.25n = 850 − 425
0.25n = 425
n = 1700
4 Divide by 17 as the hairdryer is used Number of weeks = 100
for 17 hours each week.
5 Write your answer. After 100 weeks, the hairdryer is expected
to halve in value.
d Solve for Vn = 100. 100 = 850 − 0.25n
0.25n = 750
n = 3000
Write your answer. The hairdryer can be used for 3000 hours
before it reaches its scrap value.

356 Chapter 7 Modelling growth and decay using recursion 7C
Skill-
sheet Exercise 7C
Writing an explicit rule from a linear recurrence relation

Example 13 1 Write down a rule for An for each of the following recurrence relations. In each case
calculate A20 .
a A0 = 4, An+1 = An + 2 b A0 = 10, An+1 = An − 3
c A0 = 5, An+1 = An + 8 d A0 = 300, An+1 = An − 18
Using a rule for simple interest loans and investments

Example 14 2 Webster borrows $5000 from a bank at an annual simple interest rate of 5.4%.
a Let Vn be the value of the loan after n years. State the starting value, V0 .
b Determine how much interest is charged each year in dollars.
c Write down a rule for the value of the loan, Vn , after n years.
d Use your rule to find how much Webster will owe the bank after 9 years.
3 Anthony borrows $12 000 from a bank at an annual simple interest rate of 7.2%.
a Let Vn be the value of the loan after n years. State the starting value, V0 .
b Determine how much interest is charged each year in dollars.
c Write down a rule for the value of the loan, Vn , after n years.
d Use your rule to find how much Anthony will owe the bank after 9 years.
Example 15 4 The value of a simple interest loan after n years, Vn , can be calculated using the
rule Vn = 8000 + 512n.
a What is the principal of this loan?
b How much interest is charged every year in dollars?
c Use the rule to find:
i the value of the loan after 12 years
ii when the value of the loan first doubles in value.
5 The value of a simple interest investment after n years, Vn , can be calculated using the
rule Vn = 2000 + 70n.
a What is the principal of this investment?
b How much interest is earned every year in dollars?
c Use the rule to find:
i the value of the investment after 6 years
ii when the value of the initial investment will first double in value.

7C 7C Using an explicit rule for linear growth or decay 357
Using a rule for flat rate depreciation of assets

Example 16 6 A computer is purchased for $5600 and is depreciated at a flat rate of 22.5% per year.
a State the starting value.
b Determine the annual depreciation in dollars.
c Write down a rule for the value of the computer, Vn , after n years.
d Find the value of the computer after 3 years.
7 A machine costs $7000 new and depreciates at a flat rate of 17.5% per annum. The
machine will be written off when its value is $875.
a State the starting value.
b Determine the annual depreciation in dollars.
c Write down a rule for the value of the machine, Vn , after n years.
d Determine the number of full years that the machine will be used (that is, has a
value greater than zero).
Example 17 8 The value of a sewing machine after n years, Vn , can be calculated from the
rule Vn = 1700 − 212.5n.
a What is the purchase price of the sewing machine?
b By how much is the value of the sewing machine depreciated each year in dollars?
c Use the rule to find the value of the sewing machine after 4 years.
d Find its value after 7 years.
e Determine the number of years it takes for the sewing machine to be worth nothing.
9 The value of a harvester after n years, Vn , can be calculated from the

rule Vn = 65000 − 3250n.
a What is the purchase price of the harvester?
b By how much is the value of the harvester depreciated each year in dollars?
c What is the annual percentage depreciation for the harvester?
d Use the rule to find the value of the harvester after 7 years.
e How long does it take the harvester to reach a value of $29 250?
Using a rule for unit cost depreciation of assets

Example 18 10 The value of a taxi after n kilometres, Vn , can be calculated from the
rule Vn = 29000 − 0.25n.
a What is the purchase price of the taxi?
b By how much is the value of the taxi depreciated per kilometre of travel?
c What is the value of the taxi after 20 000 kilometres of travel?
d Find how many kilometres have been travelled if the taxi is valued at $5000.

358 Chapter 7 Modelling growth and decay using recursion 7C
11 A car is valued at $35 400 at the start of the year, and at $25 700 at the end of that year.
During that year, the car travelled 25 000 kilometres.
a Find the total depreciation of the car in that year in dollars.
b Find the depreciation per kilometre for this car.
c Using V0 = 35 400, write down a rule for the value of the car, Vn , after n kilometres.
d How many kilometres have been travelled if the car has a value of $6688?
12 A printing machine costing $110 000 has a scrap value of $2500 after it has printed
4 million pages.
a Find:
i the unit cost of using the machine
ii the value of the machine after printing 1.5 million pages
iii the annual depreciation of the machine if it prints 750 000 pages per year.
b Find the value of the machine after 5 years if it prints, on average, 750 000 pages
per year.
c How many pages has the machine printed by the time the value of the machine is
$70 053?

13 The value of a bicycle, purchased for $3800, is depreciated by 10% per annum using
the flat rate method.
Recursive calculations can determine the value of the bicycle after n years, Bn .
Which one of the following recursive calculations is not correct?
A V0 = 3800 B V1 = 0.9 × 3800
C V2 = 0.9 × 3420 D V3 = 0.9 × 3080
E V4 = 0.9 × 2770.20
14 An asset is purchased for $4280.

The value of the asset after n time periods, Vn , can be determined using the rule
Vn = 4280 + 25n
A recurrence relation that also models the value of this asset after n time periods is
A V0 = 4280, Vn+1 = Vn + 25n
B V0 = 4280, Vn+1 = Vn − 25n
C V0 = 4280, Vn+1 = Vn + 25
D V0 = 4280, Vn+1 = Vn − 25
E V0 = 4280, Vn+1 = 25Vn + 4280

7D Modelling geometric growth and decay 359
7D Modelling geometric growth and decay

Learning intentions
I To be able to graph the terms of a geometric sequence.
I To be able to model compound interest with a recurrence relation.
I To be able to model reducing balance depreciation with recurrence relations.
I To be able to use reducing balance depreciation with recurrence relations.
A recurrence model for geometric growth and decay

Geometric growth or decay occurs when quantities increase or decrease by the same
percentage at regular intervals. For example, a sequence that starts with 3 and doubles is
given by 3, 6, 12, 24, . . . and can be written as a recurrence relation V0 = 3, Vn+1 = 2Vn .
Example 19 Graphing the terms in a geometric sequence
For each recurrence relation, state the rule, find the first 6 terms and then plot each point
on a graph.
a V0 = 1, Vn+1 = 3Vn
b V0 = 8, Vn+1 = 0.5Vn
a 1 Convert to words. Starting value = 1
Next value = 3 × current value
2 Multiply each term by 3 to find the 1, 3, 9, 27, 81, 243
next term. Vn
250
200
150
100
50
n
3 Plot each of the points on the axis. 0 1 2 3 4 5
b 1 Convert to words. Starting value = 8
Next value = 0.5 × current value
2 Multiply each term by 0.5 to find the 8, 4, 2, 1, 0.5, 0.25
next term. Vn
10
9
8
7
6
5
4
3
2
1
n
0 1 2 3 4 5
3 Plot each of the points on the axis.

As can be seen from the previous example, the first recurrence relation generates a sequence
whose successive terms grow geometrically, while the second recurrence relation decays
geometrically.
Modelling geometric growth and decay

As a general rule, if R is a positive constant, a recurrence relation rule of the form:
Vn+1 = RVn for R > 1, can be used to model geometric growth.
Vn+1 = RVn for R < 1, can be used to model geometric decay.
Compound interest investments and loans

More common than simple interest is compound interest where any interest that is earned
after one time period is added to the principal and then contributes to the earning of interest
in the next time period. This means that the value of the investment grows in ever increasing
amounts, or grows geometrically, instead of by the same amount as in simple interest.
Consider an investment of $5000 that pays 8% interest per annum, compounding yearly.
This means that the investment’s value increases by 8% each year.
We can model the investment with a recurrence relation as follows:
Let Vn be the value of the investment after n years. Initially the investment is worth $5000 so
V0 = $5000.
To find the rule between terms:
next value = current value + 8% of current value
= 108% of current value
= 1.08 × current value
= 1.08 × Vn
We now have a recurrence relation that we can use to model and investigate the growth of an
investment over time. Compound interest loans and investments often accrue interest over
periods of less than a year which we will consider at the end of this chapter.
A recurrence model for compound interest investments

and loans that compound yearly
Let Vn be the value of the investment after n years.
Let r be the annual percentage interest rate.
The recurrence model for the value of the investment after n years is:
r
V0 = principal, Vn+1 = RVn , where R = 1 +
100

7D Modelling geometric growth and decay 361
Example 20 Modelling compound interest with a recurrence relation
The following recurrence relation can be used to model a compound interest investment
of $2000 paying interest at the rate of 7.5% per annum.
V0 = 2000, Vn+1 = 1.075 × Vn
$2484.59.
b Determine when the value of the investment will first exceed $2500.
a 1 Write down the principal, V0 . V0 = 2000
2 Use the recurrence relation to V1 = 1.075 × 2000 = 2150
calculate V1 , V2 and V3 and round to V2 = 1.075 × 2150 = 2311.25
the nearest cent. V3 = 1.075 × 2311.25 = 2484.59
b 1 Type ‘2000’. Press · (or ).
2000 2000.
2 Type ×1.075.
2000. · 1.075 2150.
3 Count how many times you press
· (or ) until the term value is 2150. · 1.075 2311.25
greater than 2500. 2311.25 · 1.075 2484.59375
2484.59375 · 1.075 2670.93828125
|
4 Write your answer. The investment will first exceed $2500
after 4 years.
Reducing balance depreciation

Reducing balance depreciation is another method of depreciation – one where the value of
an asset decays geometrically. Each year, the value will be reduced by a percentage, r%, of
the previous year’s value.
A recurrence model for reducing balance depreciation

Let Vn be the value of the asset after n years.
Let r be the annual percentage depreciation rate.
The recurrence model for the value of the asset after n years is:
r
V0 = initial value, Vn+1 = RVn , where R = 1 −
100

Example 21 Modelling reducing balance depreciation with recurrence relations
A sofa was purchased for $7500 and is depreciating at a reducing balance rate of
8.4% per annum. Write down a recurrence relation where Vn is the value of the sofa after
n years.
1 Identify the value of V0 . V0 = 7500
2 Calculate the value of R. The depreciation rate is 8.4% per annum.
8.4
R=1− so R = 0.916
100
3 Write your answer. V0 = 7500, Vn+1 = 0.916 × Vn
Example 22 Using reducing balance depreciation with recurrence relations
The following recurrence relation can be used to model the value of office furniture with a
purchase price of $9600, depreciating at a reducing-balance rate of 7% per annum.
V0 = 9600, Vn+1 = 0.93 × Vn
In the recurrence relation, Vn is the value of the office furniture after n years.
a Use the recurrence relation to find the value of the office furniture, correct to the
nearest cent, after 1, 2 and 3 years.
b If the office furniture was initially purchased in 2023, at the end of which year will the
value of the investment first be less than $7000?
a 1 Write down the purchase price of the V0 = 9600
furniture, V0 .
2 Use the recurrence relation to V1 = 0.93 × 9600 = 8928
calculate V1 , V2 and V3 . Use your V2 = 0.93 × 8928 = 8303.04
calculator if you wish. V3 = 0.93 × 8303.04 = 7721.83
b Steps
1 Type 9600 and press · or .

9600 9600.
2 Type × 0.93.
9600. · 0.93 8928.
3 Count how many times you press
· until the term value is less than 8928. · 0.93 8303.04
7000. 8303.04 · 0.93 7721.8272
7721.8272 · 0.93 7181.299296
7181.299296 · 0.93 6678.708345
4 Write your answer. The value of the furniture first drops below
$7000 after 5 years. Thus, it is first worth
less than $7000 at the end of 2028.

7D 7D Modelling geometric growth and decay 363
Skill-
sheet Exercise 7D
Example 19 1 Generate and graph the first five terms of the sequences defined by the recurrence
relations.
a V0 = 2, Vn+1 = 2Vn
b V0 = 3, Vn+1 = 3Vn
c V0 = 100, Vn+1 = 0.1Vn
Modelling compound interest with recurrence relations

Example 20 2 An investment of $6000 earns compounding interest at the rate of 4.2% per annum.
A recurrence relation that can be used to model the value of the investment after
n years is shown below.
V0 = 6000, Vn+1 = 1.042Vn
$6788.20.
b Determine how many years it takes for the value of the investment to first exceed
$8000.
3 A loan of $20 000 is charged compounding interest at the rate of 6.3% per annum.
A recurrence relation that can be used to model the value of the loan after n years is
shown below.
V0 = 20 000, Vn+1 = 1.063Vn
In the recurrence relation, Vn is the value of the loan after n years.
a Use the recurrence relation to show that the value of the loan after 3 years is
$24 023.14.
b Determine how many years it takes for the value of the loan to first exceed $30 000.
4 Sue invests $5000 at a compounding rate of 6.8% per annum.

Let Vn be the value of the investment after n years. This compound interest investment
can be modelled by a recurrence relation of the form
V0 = principal , Vn+1 = R × Vn
a State the value of V0 .
r
b Determine R using R = 1 + .
100
c Write down a recurrence relation for the investment.
d Find the value of the investment after 5 years.
e Find the total interest earned over 5 years.

364 Chapter 7 Modelling growth and decay using recursion 7D
5 Jay takes out a loan of $18 000 at a compounding interest rate of 9.4% per annum.
a State the principal (starting value).
r
b Determine R using R = 1 + .
100
c Let Vn be the value of the loan after n years. Write down a recurrence relation for
this loan.
d Use the recurrence relation to find the value of the loan after 4 years.
e When will the loan first be valued at more than $25 000.
Modelling reducing balance depreciation with recurrence relations

Example 21 6 A motorcycle, purchased new for $9800, will be depreciated using a reducing balance
depreciation method with an annual depreciation rate of 3.5%. Write a recurrence
relation to model the value of the motorcycle using Vn to represent the value of the
motorcycle after n years.
7 Let Mn be the value of a minibus after n years. Write down a recurrence relation for a
minibus that was initially valued at $28 600 and is depreciated at a reducing-balance
rate of 7.4% per annum.
Example 22 8 Office furniture was purchased new for $18 000. It will be depreciated using a reducing
balance depreciation method with an annual depreciation rate of 4.5%. Let Vn be the
value of the furniture after n years.
a Write a recurrence relation to model the value of the furniture, Vn .
b Use the recurrence relation to find the value of the furniture after each of the first 5
years. Write the values of the terms of the sequence correct to the nearest cent.
c What is the value of the furniture after 3 years?
d What is the total depreciation of the furniture after 5 years?
9 A wedding gown was purchased new for $4 000. The value of the wedding gown
depreciates using a reducing balance depreciation method with an annual depreciation
rate of 4.1%. Let Wn be the value of the wedding dress after n years.
a Write a recurrence relation to model the value of the wedding dress, Wn .
b Calculate the value of the wedding dress after three years.
c Determine the total amount of depreciation of the dress after five years.
10 A new computer server was purchased for $13 420. The value of the computer server
depreciates using a reducing-balance depreciation method with an annual depreciation
rate of 11.2%. Let S n be the value of the server after n years.
a Write a recurrence relation to model the value of the server, S n .
b Use the recurrence relation to find the value of the server after each of the first 5
years. Write the values of the terms of the sequence correct to the nearest cent.
c What is the value of the server after 5 years?
d What is the depreciation of the server in the third year?
7D 7D Modelling geometric growth and decay 365

11 The value of a compound interest investment, in dollars, after n years, Vn can be
modelled by the recurrence relation shown below.
V0 = 80 000, Vn+1 = 1.02 × Vn
The interest rate, per annum, for this investment is
A 0.02% B 0.2%
C 2% D 1.02%
E 102%
12 Raymond invests $6000 in an account that pays interest compounding annually.

The balance of his investment after n years, Rn , can be determined using the recurrence
relation.
R0 = 6000, Rn+1 = 1.0384 × Rn
The total interest earned by Raymond’s investment after the first three years is closest
to
A $230.40 B $248.43
C $469.64 D $469.65
E $718.08
13 Mana invests $28 000 at an interest rate of 6.2% per annum, compounding annually.
Her investment will first be more than double its original value after
A 1 year B 2 years
C 10 years D 11 years
E 12 years
14 Giles purchases a jet ski for $18 000.

The value of the jet ski is depreciated each year by 10% using the reducing balance
method.
In the third year, the jet ski will depreciate in value by
A $1312.20 B $13122
C $1458 D $11809.80
E $14580

7E Using an explicit rule for geometric growth or

decay
Learning intentions
I To be able to write explicit rules for geometric growth and decay.
I To be able to use an explicit rule to find the value of an investment after n years.
I To be able to calculate the value and total depreciation of an asset after a period of
reducing balance depreciation.
I To be able to use a calculator to solve geometric growth and decay problems.
As with linear growth and decay, we can derive a rule to calculate any term in a geometric
sequence directly.
Assume $2000 is invested in a compound interest investment paying 5% per annum,
compounding yearly. Let Vn be the value of the investment after n years, giving the
following recurrence relation to model this investment:
V0 = 2000, Vn+1 = 1.05Vn
Using this recurrence relation we can write out the sequence of terms generated as follows:
V0 = 2000
V1 = 1.05V0
V2 = 1.05V1 = 1.05(1.05V0 ) = 1.052 V0
V3 = 1.05V2 = 1.05(1.052 V0 ) = 1.053 V0
V4 = 1.05V3 = 1.05(1.053 V0 ) = 1.054 V0
and so on.
Following this pattern, after n years of interest are added, we have:
Vn = 1.05n V0
With this rule, we can now find the value of the investment for any specific year. For
example, using this rule, the value of the investment after 18 years would be:
V18 = 1.0518 × 2000 = $4813.24 (to the nearest cent)
A rule for individual terms of a geometric growth and decay sequence

For a geometric growth or decay recurrence relation
V0 = starting value, Vn+1 = RVn
the value after n iterations is given by the rule:
Vn = Rn × V0

7E Using an explicit rule for geometric growth or decay 367
Exactly the same rule will work for both growth and decay, noting that R > 1 is used for
growth and R < 1 for decay.
Example 23 Writing explicit rules for geometric growth and decay
Write down a rule for the value of Vn in terms of n for each of the following. Use the rule
to find the value of V6 .
a V0 = 5, Vn+1 = 4Vn
b V0 = 10, Vn+1 = 0.5Vn
a 1 V0 = 5, R = 4 Vn = 4n × 5
2 Substitute n = 6. V6 = 46 × 5 = 20480
b 1 V0 = 10, R = 0.5 Vn = 0.5n × 10
2 Substitute n = 6. V6 = 0.56 × 10 = 0.15625
Explicit rules for compound interest loans and

investments and reducing balance depreciation
Since compound interest loans and investments increase over time, the value of R is
greater than 1 and can be found using the interest rate of r% per annum using the formula
r
R=1+ .
100
Compound interest loans and investments

Let V0 be the amount borrowed or invested (principal).
Let r be the annual percentage interest rate, with interest compounding annually.
The value of a compound interest loan or investment after n years, Vn , is given by the rule
r n
Vn = 1 + × V0
100
Example 24 Using a rule to find the value of an investment after n years
The rule for the value of the investment after n years, Vn , is shown below.
Vn = 1.09n × 10 000
a State how much money was initially invested.
b Find the annual interest rate for this investment.
c Find the value of the investment after 4 years, correct to the nearest cent.
d Find the amount of interest earned over the first 4 years, correct to the nearest cent.
e Find the amount of interest earned in the fourth year, correct to the nearest cent.
f Determine if the investor has doubled their money within 10 years.

a Recall the formnof the direct rule $10 000
r
Vn = 1 + × V0 . Read off V0 .
100
r
b Since R = 1.09 = 1 + r=9
100
The annual interest rate is 9%.
c 1 Substitute n = 4 into the rule for the V4 = 1.094 × 10 000
value of the investment. V4 = 14 115.816 . . .
2 Write your answer, rounded to the After 4 years, the value of the investment is
nearest cent. $14 115.82, correct to the nearest cent.
d To find the total interest earned in Amount of interest
4 years, subtract the principal from the = $14 115.82 − $10 000
value of the investment after 4 years. = $4115.82
After 4 years, the amount of interest earned
is $4115.82.
e 1 Calculate V3 to the nearest cent. V3 = 1.093 × 10 000
V3 = 12 950.29 (nearest cent)
2 Calculate V4 − V3 . V4 − V3 = 14115.82 − 12950.29
= 1165.53
3 Write your answer. Interest of $1165.53 was earned in the
Note: An alternate method is to calculate fourth year.
9% of V3 .
f Calculate V10 and compare this to We require Vn = 2 × V0 = 20 000.
double the principal. Note V10 = 1.0910 × 10 000 = 23 673.64
Since 23 673.64 > 20 000, the investor has
doubled their money within 10 years.
With reducing balance depreciation, the value of an asset declines over time. The value of R
r
can be found using the formula R = 1 − where r% is the annual depreciation rate.
100
Reducing balance depreciation

Let V0 be the purchase price of the asset.
Let r be the annual percentage rate of depreciation.
The value of an asset after n years, Vn , is given by the rule
r n
Vn = 1 − × V0
100

7E Using an explicit rule for geometric growth or decay 369
Example 25 Calculating the value and total depreciation of an asset after a

period of reducing balance depreciation
A machine costs $9500 to buy, and decreases in value with reducing balance depreciation
of 20% each year. A recurrence relation that can be used to model the value of the
machine after n years, Vn , is shown below.
V0 = 9500, Vn+1 = 0.8 × Vn
a Write down the rule for the value of the machine after n years.
b Use the rule to find the value of the machine after 8 years. Write your answer, correct
to the nearest cent.
c Calculate the total depreciation of the machine after 8 years.
a 1 Write down the values of V0 and R. V0 = 9500
20
R=1− = 0.8
100
2 Write down the rule. Vn = Rn × V0
Vn = 0.8n × 9500
b 1 Substitute n = 8 into the rule. V8 = 0.88 × 9500
V8 = 1593.835 . . .
2 Write your answer, rounding as After 8 years, the value of the machine is
required. $1593.84, correct to the nearest cent.
c To find the total depreciation after 8 Depreciation = $9500 − $1593.84
years, subtract the value of the machine = $7906.16
after 8 years from the original value of After 8 years, the machine has depreciated
the machine. Write your answer. by $7906.16.
Using a CAS calculator

As well as finding the value of an investment or loan with compound interest or the value of
an asset with reducing balance depreciation, it is also possible to find how long it will take
for an investment to reach a particular value.
While this can be done using trial and error, it is also possible to solve these types of
problems using a CAS calculator. Similarly, it is also possible to find the annual rate of
interest or depreciation which will lead to a particular value in a given number of years.

Example 26 Using a calculator to solve geometric growth and decay problems to

find n
How many years will it take for an investment of $5000, paying compound interest at 6%
per annum, to grow above $8000? Write your answer correct to the nearest year.
6
1 Write down the values of V0 , Vn and R. V0 = 5000, R = 1 + = 1.06
100
Vn = 8000
2 Substitute into the rule for the Vn = Rn × V0
particular term of a sequence. 8000 = 1.06n × 5000
3 Solve this equation for n using a CAS
solve (8000 = (1.06)n · 5000, n)
calculator.
n = 8.06611354799
|
4 Write your answer, rounding up as The value of the investment will grow
interest is paid at the end of the year. above $8000 after 9 years.
After 8 years, the value is $7969.24.
Example 27 Using a calculator to solve geometric growth and decay problems to

find r
An industrial weaving company purchased a new loom at a cost of $56 000. It has
an estimated value of $15 000 after 10 years of operation. If the value of the loom is
depreciated using a reducing balance method, what is the annual rate of depreciation?
Write your answer correct to one decimal place.
1 Write down the values of V0 , Vn , R V0 = 56 000, Vn = 15 000, n = 10
r
and n. R=1−
100
2 Substitute into the rule for the nth term Vn = Rn × V0

r 10
of a sequence. V10 = 1 − × V0
100 r 10
15 000 = 1 − × 56 000
100
3 Solve this equation for r using a CAS !
r 10
calculator. solve 15000 = 1 − · 56000, r
100
r = 12.3422491484 or
r = 187.657750852
Note: there are two answers. Choose the

more appropriate of the two.
4 Write your answer. The annual rate of depreciation is 12.3%,
correct to one decimal place.
7E 7E Using an explicit rule for geometric growth or decay 371
Exercise 7E
Writing explicit rules for geometric recurrence relations

Example 23 1 Write down a rule for Vn in terms of n for each of the following recurrence relations.
Use each rule to find the value of V4 .
a V0 = 6, Vn+1 = 2Vn
b V0 = 10, Vn+1 = 3Vn
c V0 = 1, Vn+1 = 0.5Vn
d V0 = 80, Vn+1 = 0.25Vn
Using a rule for compound interest loans and investments

Example 24 2 The value of an investment earning compound interest every year is modelled using the
recurrence relation:
V0 = 3000, Vn+1 = 1.1Vn .
a i How much money was invested?
ii What is the annual interest rate for this investment?
b Write down a rule for the value of the investment after n years, Vn .
c Use the rule to find the value of the investment after 5 years. Round your answer to
the nearest cent.
3 The value of a loan that is charged compound interest every year is modelled using the
recurrence relation:
V0 = 2000, Vn+1 = 1.06Vn .
a i How much money was borrowed?
ii What is the annual interest rate for this loan?
b Write down a rule for the value of the loan after n years, Vn .
c Use the rule to find the value of the loan after 4 years. Round your answer to the
nearest cent.
d If the loan is fully repaid after 6 years, what is the total interest that is paid? Round
your answer to the nearest cent.
4 Pacey invests $8000 in an account earning 12.5% compound interest each year. Let Vn
be the value of the investment after n years.
a Write down a rule for the value of Pacey’s investment after n years.
b Use the rule to find the value of the investment after 3 years. Round your answer to
the nearest cent.
c How much interest has been earned after 3 years? Round your answer to the nearest
cent.
d How much interest was earned in the third year of the investment? Round your
answer to the nearest cent.

372 Chapter 7 Modelling growth and decay using recursion 7E
Using a rule for reducing balance depreciation

Example 25 5 The value of a stereo system depreciating annually using reducing balance
depreciation is modelled using the recurrence relation V0 = 1200, Vn+1 = 0.88Vn .
a i What is the purchase price of the stereo system?
ii At what percentage rate is the stereo system being depreciated?
b Write down a rule for the value of the stereo system after n years.
c Find the value of the stereo system after 7 years. Round your answer to the nearest
cent.
6 A car was purchased for $38 500 and depreciates at a rate of 9.5% per year, using a
reducing balance depreciation method. Let Vn be the value of the car after n years.
a Write down a rule for the value of the car after n years.
b Find the value of the car after 5 years. Round your answer to the nearest cent.
c What is the total depreciation of the car over 5 years? Round your answer to the
nearest cent.
Using a CAS calculator to solve geometric growth and decay problems

Example 26 7 Sarah invested $3500 at 6.75% per annum, compounding annually.
If the investment now has a value of $5179.35, for how many years was it invested?
8 After how many years would an investment of $200 invested at 4.75% per annum,
compounding annually, first exceed a value of $20 000?
Example 27 9 An investment of $1000 has grown to $1601.03 after 12 years invested at r% per
annum compound interest. Find the value of r to the nearest whole number.
10 What annual reducing balance depreciation rate would cause the value of a car to drop
from $8000 to $6645 in 3 years? Give your answer to the nearest percent.
11 How much money must you deposit in a compounding interest investment at a rate of
6.8% per annum if you require $12 000 in 4 years’ time? Round your answer to the
nearest cent.
12 A machine has a book value after 10 years of $13 770.

If it depreciated at a reducing balance rate of 8.2% per annum, what was the initial
value of the machine? Round your answer to the nearest cent.

13 A ute had an initial value of $68 000 and was depreciated using the reducing balance
method. After five years, it had a value of $37 971.60.
The annual rate of depreciation was closest to
A 9% B 10% C 11% D 55% E 56%
14 Amber invests $15 000 at an interest rate of 5.8% per annum, compounding annually.
After how many years will her investment first be more than double its original value?
A 1 B 2 C 10 D 12 E 13
7F Interest rates over different time periods and effective interest rates 373
7F Interest rates over different time periods and

effective interest rates
Learning intentions
I To be able to convert nominal (annual) interest rates to compounding period interest
rates.
I To be able to model loans with different compounding periods using recurrence
relations.
I To be able to model investments with different compounding periods using recurrence
relations.
I To be able to compare loans and investments using effective interest rates.
I To be able to calculate effective interest rates using a CAS calculator.
Compound interest rates are usually quoted as annual rates, or interest rates per annum. This
rate is called the nominal interest rate for the investment or loan. Despite this, interest can
be calculated and paid according to a different time period, such as monthly. The time period
which compound interest is calculated and paid upon is called the compounding period.
Interest rate conversions

The interest rate for the compounding period is calculated based on the following:
12 equal months in every year (even though some months have different numbers of days)
4 quarters in every year (a quarter is equal to 3 months)
26 fortnights in a year (even though there are slightly more than this)
52 weeks in a year (even though there are slightly more than this)
365 days in a year (ignore the existence of leap years).
A nominal interest rate is converted to a compounding period interest rate by dividing by

these numbers, which we will refer to with the letter p.
Example 28 Converting nominal interest rates to compounding period interest

rates
An investment account will pay interest at the rate of 4.68% per annum. Convert this
interest rate to each of the following rates:
a monthly b fortnightly c quarterly.
4.68
a Divide by p = 12. Monthly interest rate = = 0.39%
12
4.68
b Divide by p = 26. Fortnightly interest rate = = 0.18%
26
4.68
c Divide by p = 4. Quarterly interest rate = = 1.17%
4

Recurrence relations with different compounding periods

An annual interest rate can be converted to a compounding period interest rate and then used
in a recurrence relation to model a compounding investment or loan.
To do this, we update our definition of the growth multiplier, R, for compound interest loans
and investments as follows:
r
R=1+
100 × p
where r is the annual nominal interest rate and p is the number of compounding periods in
each year. If compounding is annual, we use p = 1.
Example 29 Recurrence relations with different compounding periods
Brian borrows $5000 from a bank. He will pay interest at the rate of 4.5% per annum.
Let Vn be the value of the loan after n compounding periods.
Write down a recurrence relation to model the value of Brian’s loan if interest is
compounded:
a yearly b quarterly c monthly.
a 1 Define the variable Vn . The Let Vn be the value of Brian’s loan after n
compounding period is yearly. years.
2 Determine the value of R where The interest rate is 4.5% per annum.
r = 4.5 and p = 1. 4.5
R=1+ = 1.045
100 × 1
3 Write the recurrence relation. V0 = 5000, Vn+1 = 1.045Vn
b 1 Define the variable Vn . The Let Vn be the value of Brian’s loan after n
compounding period is quarterly. quarters.
2 Determine the value of R, where The interest rate is 4.5% per annum.
r = 4.5 and p = 4. 4.5
R=1+ = 1.01125
100 × 4
c 1 Define the variable Vn . The Let Vn be the value of Brian’s loan after n
compounding period is monthly. months.
2 Determine the value of R where The interest rate is 4.5% per annum.
r = 4.5 and p = 12. 4.5
R=1+ = 1.00375
100 × 12

Example 30 Modelling an investment that compounds monthly using recurrence

relations
A principal value of $10 000 is invested in an account earning compound interest monthly
at the rate of 9% per annum.
Let Vn be the value of the investment after n months.
a Calculate the growth multiplier, R.
b Write down a recurrence relation for the value of the investment after n months.
c Write down a rule for the value of the investment after n months.
d Use this rule to find the value of the investment after 4 years.
9
a Since interest compounds monthly, R=1+ = 1.0075
100 × 12
p = 12.
b Substitute V0 and R to form the V0 = 10 000, Vn+1 = 1.0075Vn
recurrence relation.
c Substitute R = 1.0075 and V0 = 10 000 Vn = 1.0075n × 10 000
into the rule to find the rule for Vn .
d Substitute n = 48 (4 years = 48 V48 = 1.007548 × 10 000
months) into the rule. = $14314.05
Effective interest rates

When interest compounds with different compounding periods, the total amount of interest
earned in one year differs. To compare loans and investments, we can calculate the effective
interest rate which is the percentage the value increases in one year.
Total interest in one year
effective rate = × 100%
Principal
For example, if $5000 is invested paying a nominal rate of 4.8% per annum and interest
compounds quarterly, the value at the end of the year is $5244.35 so the interest is $244.35.
Using the formula,
244.35
effective rate = × 100%
5000
= 4.89%
In contrast, carrying out the same procedure when interest is compounding monthly gives an
effective interest rate of 4.91% due to a value of $5245.35 at the end of the year and hence
interest of $245.35. Thus, calculating effective interest rates with different compounding
periods provides us with a good way of comparing loans and investments.
In order to calculate the effective interest rates for different loans or investments, we can use
the following rule.

Effective interest rate

The effective interest rate of a loan or investment is the interest earned after one year
expressed as a percentage of the amount borrowed or invested.
Let:
r be the nominal interest rate per annum
reff be the effective annual interest rate
n be the number of times the interest compounds each year.
r n
The effective annual interest rate is given by: reff = 1+ − 1 × 100%
100 × n
Note: n is used here in line with the VCAA formula sheet and should not be confused with other usages
of n in this chapter.
Example 31 Comparing loans and investments with effective interest rates
Brooke would like to borrow $20 000 that she will repay entirely after one year. She is
deciding between two loan options:
option A: 5.95% per annum, compounding weekly
option B: 6% per annum, compounding quarterly.
a Calculate the effective interest rate for each investment.
b Which investment option is the best and why?
a 1 Decide on the values of r A: r = 5.95% and n = 52

and n for each option. B: r = 6% and n = 4
 !52 
5.95
A: reff =  1 + − 1 × 100% = 6.13%
 
2 Apply the effective interest
100 × 52
rate rule.  !4 
6
B: reff =  1 + − 1 × 100% = 6.14%
 
100 × 4
b Compare the effective interest Brooke is borrowing money, so the best option is the
rates. one with the lowest effective interest rate. She will
pay less interest with option A.
While the effective interest rate can be calculated manually, the CAS calculator can also be
used to quickly perform the calculation. To do this, the nominal interest rate and the number
of compounding periods in a year are required.

Example 32 Calculating effective interest rates using a CAS calculator
Marissa has $10 000 to invest. She chooses an account that will earn compounding
interest at the rate of 4.5% per annum, compounding monthly.
Use a CAS calculator to find the effective interest rate for this investment, correct to three
decimal places.
Steps
1 Press b and then select
8: Finance
5: Interest Conversion
2: Effective interest rate
to paste in the eff(. . . ) command.
The parameters of this function are
eff(nominal rate, number of times the
interest compounds each year).
2 Enter the nominal rate (4.5) and number
of times the interest compounds each
year (12) into the function, separated by
a comma. Press · to get the effective
interest rate.
Steps
1 Select Interactive, Financial, Interest
Conversion, ConvEff.
2 Enter the number of times the interest
compounds each year (12) and the
nominal rate (4.5) as shown .
3 Press EXE to get the effective interest

rate. Write your answer. The effective interest rate for this
investment is 4.594%.

378 Chapter 7 Modelling growth and decay using recursion 7F
Exercise 7F
Interest rate conversions
Example 28 1 Convert each of the annual interest rates below to an interest rate for the given time
period. Write your answers correct to two decimal places.
a 4.8% per annum to monthly b 8.3% per annum to quarterly
c 10.4% per annum to fortnightly d 7.4% per annum to weekly
e 12.7% per annum to daily
2 Convert each of the interest rates below to an annual interest rate.

a 0.54% monthly b 1.45% quarterly c 0.57% fortnightly
d 0.19% weekly e 0.022% daily
Recurrence relations with different compounding periods

Example 29 3 Verity borrows $8000 from a bank. She will pay interest at a rate of 4.8% per annum.
Let Vn be the value of the loan after n compounding periods. Write down a recurrence
relation to model the value of Verity’s loan if interest is compounded:
a yearly b quarterly c monthly
Example 30 4 A principal value of $20 000 is invested in an account earning compound interest of
6% per annum, compounding monthly.
a Write down a recurrence relation for the value of the investment, Vn , after n months.
b Write down a rule for Vn in terms of n.
c Use this rule to find the value of the investment after 5 years (60 months). Round
5 A principal value of $8 000 is invested in an account earning compound interest

quarterly at the rate of 4.8% per annum.
a Write down a recurrence relation for the value of the investment, Vn , after n
quarters.
c Use this rule to find the value of the investment after 3 years (12 quarters). Round
6 Wayne invests $7600 with a bank. He will be paid interest at the rate of 6% per annum,
compounding monthly. Let Vn be the value of the investment after n months.
a Write a recurrence relation to model Wayne’s investment.
c How much is Wayne’s investment worth after 5 months? Round your answer to the
nearest cent.
d After how many months will Wayne’s investment first double in value?

7F 7F Interest rates over different time periods and effective interest rates 379
7 Jessica borrows $3500 from a bank. She will be charged compound interest at the rate
of 8% per annum, compounding quarterly. Let Vn be the value of the loan after
n quarters.
a Write a recurrence relation to model the value of Jessica’s loan.
b If Jessica pays back everything she owes to the bank after 1 year, how much money
will she pay? Round your answer to the nearest cent.
Comparing loans and investments with effective interest rates

Example 31 8 Brenda invests $15 000 in an account earning (nominal) compound interest of
4.6% per annum.
a Calculate the effective interest rate for the current investment when interest
compounds quarterly, correct to two decimal places.
b Calculate the effective rate for this investment when interest compounds monthly,
c Should Brenda choose quarterly or monthly compounding?
9 Stella borrows $25 000 from a bank and pays nominal compound interest of
7.94% per annum.
a Calculate the effective rate for the current loan when interest compounds fortnightly,
b Calculate the effective rate for this loan when interest compounds monthly, correct
to two decimal places.
c Should Stella choose fortnightly or monthly compounding?
10 Luke is considering a loan of $35 000. His bank has two compound interest rate
options:
A: 8.3% per annum, compounding monthly
B: 7.8% per annum, compounding weekly.
a Calculate the effective interest rate for each of the loan options. Round your answers
to two decimal places.
b Calculate the amount of interest Luke would pay in the first year for each of the loan
options. Round your answers to the nearest cent.
c Which loan should Luke choose and why?
11 Sharon is considering investing $140 000. Her bank has two compound interest
investment options:
A: 5.3% per annum, compounding monthly
B: 5.5% per annum, compounding quarterly.
a Calculate the effective interest rate for each of the investment options. Round your
answers to two decimal places.

380 Chapter 7 Modelling growth and decay using recursion 7F
b Calculate the amount of interest Sharon would earn in the first year for each of the
investment options. Give your answer to the nearest dollar.
c Which investment option should Sharon choose and why?
Calculating effective interest rates using the CAS

Example 32 12 Use your calculator to determine the effective annual interest rate, correct to two
decimal places, for the following nominal rates and compounding periods.
a 6.2% per annum compounding monthly
b 8.4% per annum compounding daily
c 4.8% per annum compounding weekly
d 12.5% per annum compounding quarterly
13 An account increases by 7% in one year when interest compounds monthly. Find the
annual interest rate correct to 2 decimal places.

14 Chung invests $3300 with a bank. He will be paid compound interest at the rate of
4.8% per annum, compounding monthly. If Vn is the value of Chung’s investment after
n months, a recurrence model for Chung’s investment is
A V0 = 3300, Vn+1 = Vn + 4.8
B V0 = 3300, Vn+1 = 4.8Vn
C V0 = 3300, Vn+1 = 0.004Vn
D V0 = 3300, Vn+1 = Vn + 158.40
E V0 = 3300, Vn+1 = 1.004Vn
15 An amount of $4700 is invested, earning compound interest at the rate of 6.8% per
annum, compounding quarterly. The effective annual interest rate is closest to
A 6.80% B 6.97% C 6.98% D 7.02% E 7.03%
16 Isabella invests $5000 in an account that pays interest compounding monthly.

After one year, the balance of the account is $5214.09.
The effective interest rate for this investment, rounded to two decimal places, is
A 0.35% B 0.42% C 3.50% D 4.20% E 4.28%
17 Maya invested $25 000 in an account at her bank with interest compounding monthly.
After one year, the balance of Maya’s account was $26 253.
The difference between the rate of interest per annum used by her bank and the
effective annual rate of interest for Maya’s investment is closest to
A 0.112% B 0.2% C 4.89%
D 4.9% E 5.012%

Review
Sequence A sequence is a list of numbers or symbols written in succession, for

Assign- example: 5, 15, 25, . . .
ment
Term Each number or symbol that makes up a sequence is called a term.
Recurrence A relation that enables the value of the next term in a sequence to be
relation obtained by one or more current terms. Examples include ‘to find
the next term, add two to the current term’ and ‘to find the next term,
multiply the current term by three and subtract five’.
Modelling Modelling is the use of a mathematical rule or formula to represent or

model real-life situations. Recurrence relations can be used to model
situations involving the growth (increase) or decay (decrease) in values
of a quantity.
r
Percentage If a quantity grows by r% each year, then R = 1 + .
growth and decay 100
r
If a quantity decays by r% each year, then R = 1 − .
100
Principal The principal is the initial amount that is invested or borrowed.
Balance The balance is the value of a loan or investment at any time during the
loan or investment period.
Interest The fee that is added to a loan or the payment for investing money is
called the interest.
Simple interest Simple interest is a fixed amount of interest that is paid at regular time
intervals. Simple interest is an example of linear growth.
Depreciation Depreciation is the amount by which the value of an item decreases

after a period of time.
Scrap value Scrap value is the value of an item at which it is ‘written off’ or is
considered no longer useful or usable.
Flat rate Flat rate depreciation is a constant amount that is subtracted from
depreciation the value of an item at regular time intervals. It is an example of linear
decay.
Unit cost Unit cost depreciation is depreciation that is calculated based on units
depreciation of use rather than time. Unit-cost depreciation is an example of linear
decay.

Review 382 Chapter 7 Modelling growth and decay using recursion
Compounding Interest rates are usually quoted as annual rates (per annum). Interest
period is sometimes calculated more regularly than once a year, for example
each quarter, month, fortnight, week or day. The time period for the
calculation of interest is called the compounding period.
Compound When interest is added to a loan or investment and then contributes to

interest earning more interest, the interest is said to compound. Compound
interest is an example of geometric growth.
Reducing balance When the value of an item decreases as a percentage of its value after
depreciation each time period, it is said to be depreciating using a reducing balance
method. Reducing balance depreciation is an example of geometric
decay.
Nominal interest A nominal interest rate is an annual interest rate for a loan or
rate investment.
Effective interest The effective interest rate is the interest earned or charged by an
rate investment or loan, written as a percentage of the original amount
invested or borrowed. Effective interest rates allow loans or investments
with different compounding periods to be compared. Effective interest
r n
rates can be calculated using the rule reff = 1 + − 1 × 100%
100 × n
where r is the nominal annual interest rate and n is the number of
compounding periods in 1 year.
Skills checklist
list
7A 1 I can generate a sequence of terms recursively.
See Example 1 and 2, and Exercise 7A Question 1 and 2
7A 2 I can generate a sequence of numbers with a calculator.
7A 3 I can generate a sequence from a recurrence relation.
7A 4 I can use a calculator to generate sequences from recurrence relations.

Review
7A 5 I can name terms in a sequence.
7B 6 I can graph the terms of a linear growth/decay sequence.
7B 7 I can model simple interest loans and investments with a recurrence relation.
7B 8 I can use a recurrence relation to analyse a simple interest investment.
7B 9 I can model flat rate depreciation using a recurrence relation.
7B 10 I can use a recurrence relation to analyse flat rate depreciation.
7B 11 I can model unit cost depreciation with a recurrence relation.
7C 12 I can convert a recurrence relation to an explicit rule.
7C 13 I can model a simple interest investment using an explicit rule.
7C 14 I can use a rule to determine the value of a simple interest investment.
7C 15 I can model flat rate depreciation of an asset with an explicit rule.
7C 16 I can use a rule for the flat rate depreciation of an asset.
7C 17 I can use an explicit rule for unit cost depreciation.
7D 18 I can graph the terms in a geometric sequence.

7D 19 I can model compound interest with a recurrence relation.
7D 20 I can model reducing balance depreciation with recurrence relations.
7D 21 I can use reducing balance depreciation with recurrence relations.
7E 22 I can write explicit rules for geometric growth and decay.
7E 23 I can use a rule to find the value of an investment after n years.
7E 24 I can calculate the value and total depreciation of an asset after a period of
reducing balance depreciation.
7E 25 I can use a calculator to solve geometric growth and decay problems.
See Example 26 and 27, and Exercise 7E Question 7 and Question 9
7F 26 I can convert nominal (annual) interest rates to compounding period interest

rates.
7F 27 I can use recurrence relations to model loans with different compounding

periods.
7F 28 I can model an investment that compounds monthly using recurrence

relations.
7F 29 I can compare loans and investments using effective interest rates.
7F 30 I can calculate effective interest rates using a CAS calculator.

Review
1 Consider the following recurrence relation
V0 = 5, Vn+1 = Vn − 3
The sequence generated by this recurrence relation is
A 5, 15, 45, 135, 405, . . . B 5, 8, 11, 14, 17, . . .
C 5, 2, −1, −4, −7, . . . D 5, 15, 45, 135, 405, . . .
E 5, −15, 45, −135, 405, . . .

V0 = 2, Vn+1 = 2Vn + 8
The value of the term V4 in the sequence generated by this recurrence relation is
A 12 B 18 C 32 D 72 E 152

V0 = 5, Vn+1 = 3Vn − 6
The value of the term V3 in the sequence of numbers generated by this recurrence
relation is
A 5 B 9 C 21 D 57 E 165
4 Brian has two trees in his backyard. Every month, he will plant three more trees.
A recurrence relation for the number of trees, T n , in Brian’s backyard after n months is
A T 0 = 2, T n+1 = 3T n
B T 0 = 2, T n+1 = 3T n + 3
C T 0 = 2, T n+1 = T n + 3
D T 0 = 2, T n+1 = T n − 3
E T 0 = 2, T n+1 = 3T n − 3
5 A graph that shows the value of a simple interest investment of $1000, earning interest
of $5 per month is
A Mn B Mn
1050 1050
1000 1000
950 950
900 n 900 n
O 1 2 3 4 5 O 1 2 3 4 5

C Mn D Mn
1050 1100
1000 1050
950 1000
900 950 n
n O
O 1 2 3 4 5 1 2 3 4 5
E Mn
1100
1050
1000
950 n
O 1 2 3 4 5
6 A car is depreciated using a unit cost depreciation method. It was purchased for
$18 990 and, after travelling a total of 20 000 kilometres, it has an estimated value of
$15 990. The depreciation amount, per kilometre, is
A $0.15 B $0.80 C $0.95 D $6.67 E $3000
7 Arthur invests $2000 with a bank. He will be paid simple interest at the rate of 5.1%
per annum. If Vn is the value of Arthur’s investment after n years, a recurrence relation
for Arthur’s investment is
A V0 = 2000, Vn+1 = Vn + 5.1
B V0 = 2000, Vn+1 = 5.1Vn
C V0 = 2000, Vn+1 = 0.051Vn + 102
D V0 = 2000, Vn+1 = Vn + 102
E V0 = 2000, Vn+1 = 1.051Vn + 2000
8 An interest rate of 4.6% per annum is equivalent to an interest rate of

A 1.15% per quarter B 0.35% per month C 0.17% per week
D 0.17% per fortnight E $0.39% per month
9 A sequence is generated from the recurrence relation

V0 = 40, Vn+1 = Vn − 16
The rule for the value of the term Vn is
A Vn = 40n − 16 B Vn = 40 − 16n C Vn = 40n
D Vn = 40 + 16n E Vn = 40n − 16

Review
10 The recurrence relation that generates a sequence of numbers representing the value of
a car n years after it was purchased is
V0 = 18 000, Vn+1 = Vn − 1098
The car had a purchase price of $18 000 and is being depreciated using
A flat rate depreciation at 6.1% of its value per annum
B flat rate depreciation at $6.10 per kilometre travelled
C flat rate depreciation at $1098 per kilometre travelled
D unit cost depreciation at $6.10 per kilometre travelled
E unit cost depreciation at $1098 per kilometre travelled
11 A computer is depreciated using a flat rate depreciation method. It was purchased for
$2800 and depreciates at the rate of 8% per annum. The amount of depreciation after
4 years is
A $224 B $448 C $794 D $896 E $1904
12 Sandra invests $6000 in an account that pays interest at the rate of 4.57% per annum,
compounding annually. The number of years it takes for the investment to exceed
$8000 is
A 5 B 6 C 7 D 8 E 9
13 The value of a machine is depreciating by 8% every year. The initial value is 2700.
A recurrence relation model for the value of the machine after n years, Pn , is
A P0 = 2700, Pn+1 = 1.8 × Pn
B P0 = 2700, Pn+1 = 1.08 × Pn
C P0 = 2700, Pn+1 = 0.92 × Pn
D P0 = 2700, Pn+1 = 1 + 8 × Pn
E P0 = 2700, Pn+1 = 1.08 + Pn
14 An investment of $50 000 is compounding annually over a number of years. The graph
that best represents the value of the investment at the end of each year is
A B C
Amount
Amount
Amount
Year Year Year

D E
Amount
Amount
Year Year

15 An item is depreciated using a reducing balance depreciation method. The value of the
item after n years, Vn , is modelled by the recurrence relation
V0 = 4500, Vn+1 = 0.86Vn
The rule for the value of the item after n years is
A Vn = 0.86n × 4500
B Vn = 1.86n × 4500
C Vn = (1 + 0.86)n × 4500
D Vn = 0.86 × n × 4500
E Vn = (1 − 0.86)n × 4500
16 After 10 years, a compound interest investment of $8000 earned a total of $4000 in

interest when compounding annually. The annual interest rate of this investment was
closest to
A 2.5% B 4.14% C 5.03% D 7.2% E 50%
17 The interest rate on a compound interest loan is 12.6% per annum, compounding
monthly. The value of the loan after n months, Vn , is modelled by the recurrence
relation
V0 = 400, Vn+1 = R × Vn
The value of the growth multiplier, R, in this recurrence relation is
A 0.874 B 1.00 C 1.0105 D 1.126 E 2.05
18 An amount of $2000 is invested, earning compound interest at the rate of 5.4% per
annum, compounding quarterly. The effective annual interest rate is closest to
A 5.2% B 5.3% C 5.4% D 5.5% E 5.6%
19 A car was purchased for $74 500. It depreciates in value at a rate of 8.5% per year,
using a reducing balance depreciation method. The total depreciation of the car over
5 years is closest to
A $4439
B $26 718
C $37 522
D $47 782
E $112 022
20 Sam invested $6500 at 8.75% per annum with interest compounding monthly. If the
investment now amounts to $13 056, for how many years was it invested?
A 5 B 7 C 8 D 9 E 96

Review
1 Jack borrows $20 000 from a bank and is charged simple interest at the rate of 9.4% per
annum. Let Vn be the value of the loan after n years.
a Write down a recurrence relation for the value of Jack’s loan after n years.
b Use the recurrence relation to model how much Jack will need to pay the bank after
5 years.
The bank decides to change the loan to a compound interest loan on a yearly basis,
with an annual interest rate of 9.4%. Let Wn be the value of the loan after n years.
c Write a recurrence relation to model the value of Jack’s loan.
d Write a rule for Wn in terms of n.
e Use the rule to find the value of the loan after 5 years. Round your answer to the
nearest cent.
2 Ilana uses a personal loan to buy a dress costing $300. Interest is charged at 18% per
annum, compounding monthly.
If she repays the loan fully after 6 months, how much will she pay? Round your answer
3 Kelly bought her current car 5 years ago for $22 500.
Let Vn be the value of Kelly’s car after n years.
a If Kelly uses a flat rate depreciation of 12% per annum:
i write down a recurrence relation for the value of Kelly’s car after n years
ii use the recurrence relation to find the current value of Kelly’s car.
b If Kelly uses reducing value depreciation at 16% per annum:
i write down a recurrence relation for the value of Kelly’s car after n years
ii use the recurrence relation to find the current value of Kelly’s car using reducing
balance depreciation. Round your answer to the nearest cent.
c On the same axes, sketch a graph of the value of Kelly’s car against the number of
years for both flat rate and reducing balance depreciation.
4 A commercial cleaner bought a new vacuum cleaner for $650. The value of the
vacuum cleaner decreases by $10 for every 50 offices that it cleans.
a How much does the value of the vacuum cleaner depreciate when one office is
cleaned?
b Give a recurrence relation for the value of the vacuum cleaner, Vn , after n offices
have been cleaned.
c The cleaner has a contract to clean 10 offices, 5 nights a week for 40 weeks in a
year. What is the value of the vacuum cleaner after 1 year?

5 Meghan has $5000 to invest.

Company A offers her an account paying 6.3% per annum simple interest.
Company B offers her an account paying 6.1% per annum compound interest.
a How much will she have in the account offered by company A at the end of 5 years?
b How much will she have in the account offered by company B at the end of 5 years?
Round your answer to the nearest cent.
c Find the simple interest rate that company A should offer if the two investments are
to have equal values after 5 years. Round your answer to one decimal place.
6 A sum of $30 000 is borrowed at an interest rate of 9% per annum, compounding

monthly.
Let Vn be the value of the loan after n months.
a Write a recurrence relation to model the value of this loan.
b Use the recurrence relation to find the value of the loan at the end of the first five
months. Round your answer to the nearest cent.
c What is the value of the loan after 1 year? Round your answer to the nearest cent.
d If the loan is fully repaid after 18 months, how much money is paid? Round your
7 On the birth of his granddaughter, a man invests a sum of money at a rate of 11.65%
per annum, compounding twice per year.
On her 21st birthday he gives all of the money in the account to his granddaughter.
If she receives $2529.14, how much did her grandfather initially invest? Round your
8 Geoff invests $18 000 in an investment account. After 2 years the investment account
contains $19 300.
If the account pays r% interest per annum, compounding quarterly, find the value of r,
to one decimal place.

Chapter
8
Reducing balance loans,
annuities and
investments
Chapter objectives
I How can we combine both linear and geometric growth/decay?
I How do we model a compound interest investment where additional
payments are made?
I How can recurrence relations be used to model reducing balance loans?
I How can recurrence relations be used to model annuities?
I What are amortisation tables and how can they be used?
I How can a finance solver be used to analyse reducing balance loans,
annuities and investments with additional payments?
I What are interest-only loans?
I What are perpetuities?
Often loans and investments are more complex than described in the previous chapter.
In particular, loans are often paid off through regular payments, investments may have
additional contributions made throughout their life and interest rates may change. In
this chapter, we analyse investments with additional payments, reducing balance loans,
annuities, interest-only loans and perpetuities using our already developed tool of
recurrence relations as well as new tools such as amortisation tables and the Finance
Solver found on a CAS.

392 Chapter 8 Reducing balance loans, annuities and investments
8A Combining linear and geometric growth or

decay to model compound interest investments
with additions to the principal
Learning intentions
I To be able to generate a sequence from a recurrence relation that combines both
geometric and linear growth or decay.
I To be able to model compound interest investments with additions to the principal.
I To be able to use a recurrence relation to analyse compound interest investments with
additions to the principal.
I To be able to determine the annual interest rate from a recurrence relation.
In the previous chapter, recurrence relations were used to model financial situations
with linear and geometric growth/decay such as simple and compound interest and the
depreciation of assets. Recurrence relations can also be used to model situations that involve
elements of both linear and geometric growth/decay.
There are several examples in finance that involve both geometric and linear growth or
decay. For example, an investment with compound interest grows geometrically over time
but might also have linear growth if regular additions are made. Alternatively, a personal
loan may be paid off with regular payments rather than at the conclusion of the loan period.
In general, a recurrence relation of the form
V0 = starting value, Vn+1 = R × Vn ± D
can be used to model situations that involve both geometric and linear growth/decay.
Example 1 Generating a sequence from a recurrence relation of the form

V n+1 = R × V n ± D
Write down the first five terms of the sequence generated by the recurrence relation
V0 = 3, Vn+1 = 4Vn − 1
2 Apply the rule (multiply by 4, then 3 × 4 − 1 = 11
subtract 1) to generate four more terms. 11 × 4 − 1 = 43
43 × 4 − 1 = 171
171 × 4 − 1 = 683
3 Write your answer. The first five terms are 3, 11, 43, 171, 683

8A Combining linear and geometric growth 393
Compound interest investments with regular additions to

the principal
In the previous chapter, we considered compound interest investments. Here, we consider
the option of adding to the investment by making additional payments on a regular basis.
This is called an annuity investment.
Modelling compound interest investments with regular additions to the

principal
Let Vn be the value of the compound interest investment (annuity investment) after n
additional payments have been made. Then
V0 = the principal, Vn+1 = RVn + D
r
where D is the additional payment made, R = 1 + is the growth multiplier, r is
100 × p
the annual interest rate and p is the number of compounding periods per year.
To start, we recall compound interest investments without additional payments as in Chapter

7. Imagine Fred has $5000 to invest in an account paying compound interest of 4% per
annum, compounding annually. The starting value is V0 = 5000. To find R, we use the
r
formula R = 1 + where r = 4 and p = 1 to give R = 1.04. Thus, the recurrence
100 × p
relation is
V0 = 5000, Vn+1 = 1.04Vn
Now we can consider the possibility of regular additions to the principal.
Example 2 Modelling compound interest investments with additions to the

principal (1)
Fred has saved $5000 and invests this in a compound interest account paying 4% per
annum, compounding yearly. He also adds an extra $1000 each year.
Model this investment using a recurrence relation of the form
1 Write down the value of V0 and D V0 = 5000 and D = 1000
where D is the amount added each year.
4
2 Determine the value of R using R=1+ = 1.04
r 100 × 1
R=1+ where r = 4 and p = 1
100 × p
because interest compounds annually.
3 Use the values of V0 , R and D to write V0 = 5000, Vn+1 = 1.04Vn + 1000

When interest compounds at intervals other than a year, we need to find the interest rate for
the compounding period which is calculated based on the following:
12 equal months in every year (even though some months have different numbers of days)
4 quarters in every year (a quarter is equal to 3 months)
26 fortnights in a year (even though there are slightly more than this)
52 weeks in a year (even though there are slightly more than this)
365 days in a year (ignore the existence of leap years).
This gives us the value of p which we use in the formula

r
R=1+
100 × p
to find the growth multiplier, R.
Example 3 Modelling compound interest investments with additions to the

principal (2)
Nor invests $1200 and plans to add an extra $50 each month. The account pays interest at
a rate of 3% per annum, compounding monthly.
Model this investment using a recurrence relation of the form
where Vn is the value of the investment after n months.
1 Write down the value of V0 and D V0 = 1200 and D = 50
where D is the amount added each
month.
3
2 Determine the value of R using the R=1+ = 1.0025
r 100 × 12
formula R = 1 + where r = 3
100 × p
and p = 12 because interest compounds
monthly.
3 Use the values of V0 , R and D to write V0 = 1200, Vn+1 = 1.0025Vn + 50
Once we have a recurrence relation, we can use it to determine the value of the investment
after a given number of periods once interest has been paid and extra payments have been
added to the principal. This value can be plotted on a graph so that we can see the impact of
making additional payments over time.

8A Combining linear and geometric growth 395
Example 4 Using a recurrence relation to analyse compound interest

investments with additions to the principal
Albert has an investment that can be modelled by the recurrence relation

V0 = 400, Vn+1 = 1.005Vn + 30
where Vn is the value of the investment after n months.
a State the value of the initial investment.
b Determine the value of the investment after Albert has made three extra payments.
c What will be the value of his investment after 6 months? Round your answer to the
nearest cent.
d Plot the points for the value of the investment after 0, 1, 2 and 3 months on a graph.
a Note that V0 = 400. The initial investment was $400.
b Perform the calculations. V0 = $400
V1 = 1.005 × 400 + 30 = $432
V2 = 1.005 × 432 + 30 = $464.16
V3 = 1.005 × 464.16 + 30 = $496.48
The value of Albert’s investment is
$496.48.
c i Either continue performing the
400 400
calculations or use your CAS by:
ii Type 400 and press · (or ). 400 · 1.005 + 30 432
432 · 1.005 + 30 464.16
iii Type × 1.005 + 30 and press · (or
) six more times. 464.16 · 1.005 + 30 496.48
496.4808 · 1.005 + 30 528.96
528.9632 · 1.005 + 30 561.61
561.6180 · 1.005 + 30 594.42
ii Write your answer. The value of Albert’s investment is

$594.42.
d Plot each of the points on the graph.
Vn
500
400
300
200
100
n
0 1 2 3 4

396 Chapter 8 Reducing balance loans, annuities and investments 8A
From the example above, it is clear that making additional payments on a regular basis
causes the investment to increase more. In particular, making additional payments early on is
beneficial as compound interest is earned on the additional payment for longer. An example
of this type of investment might be saving for retirement.
Sometimes we are given a recurrence relation and asked to determine the annual interest
r
rate. To do this, we use the formula R = 1 + .
100 × p
Example 5 Determining the annual interest rate from a recurrence relation
Determine the annual interest rates for each of the following investments.
a Consider an investment given by the recurrence relation
A0 = 400, An+1 = 1.005Vn + 30

where An is the value of the investment after n months.
b Consider an investment given by the recurrence relation
W0 = 2000, Wn+1 = 1.012Vn + 500

where Wn is the value of the investment after n quarters.
r r
a Solve R = 1 + for r where Solve 1.005 = 1 + .
100 × p 100 × 12
R = 1.005 and p = 12 because interest r=6
is compounded monthly. Thus, the annual interest rate is 6%.
r r
b Solve R = 1 + for r where Solve 1.012 = 1 + .
100 × p 100 × 4
R = 1.012 and p = 4 because interest is r = 4.8
compounded quarterly. Thus, the annual interest rate is 4.8%.
Exercise 8A
Generating a sequence using a recurrence relation

Example 1 1 Write down the first five terms of the sequences generated by the following recurrence
relations.
a A0 = 2, An+1 = 2An + 1
b B0 = 50, Bn+1 = 2Bn − 10
c C0 = 128, Cn+1 = 0.5Cn + 32

8A 8A Combining linear and geometric growth 397
Modelling compound interest investments with additions to the principal

Example 2 2 Molly has saved $500 and plans to add an extra $100 per year to an investment account
immediately after the interest payment is calculated. The account pays 3% per annum,
compounding annually.
Let Vn is the value of the investment after n years.
a State the initial value of the investment, V0 .
b State the amount added each year, D.
c Determine the value of the growth multiplier, R.
d Model this investment using a recurrence relation of the form
3 Jane has already saved $300 000 and plans to add an extra $50 000 per year to an
investment account immediately after the interest payment is calculated. The account
pays interest of 5.2% per annum, compounding annually.
Let Vn is the value of the investment after n years.
a State the initial value of the investment, V0 .
b State the amount added each year, D.
c Determine the value of the growth multiplier, R.
d Model this investment using a recurrence relation of the form
Example 3 4 Henry invests $3500 and plans to add an extra $150 per month after the interest is
calculated. The account pays interest of 3.6% per annum, compounding monthly.
a Determine the value of the growth multiplier, R.
b Model this investment using a recurrence relation of the form

c What is the value of the investment after two months? Round your answer to the
nearest cent.
5 Lois invests $1700 and plans to add an extra $100 per quarter. The account pays
interest of 3.2% per annum, compounding quarterly.
Let Vn be the value of the investment after n quarters.
a Model this investment using a recurrence relation of the form

b What is the value of the investment after six quarters? Round your answer to the
nearest cent.

398 Chapter 8 Reducing balance loans, annuities and investments 8A
6 Sarah invests $1500 at 7.3% per annum, compounding daily. She plans to add an extra
$4 to her investment each day, immediately after the interest is calculated.
Let Vn be the value of the investment after n days.
Write down a recurrence relation to model Sarah’s investment.
7 Rachel invests $24 000 at 6% per annum, compounding monthly. She plans to add an
extra $500 to her investment each month.
Write down a recurrence relation to model Rachel’s investment and determine the value
of the investment after six months. Round your answer to the nearest cent.
Using a recurrence relation to model and analyse an investment with additions to

the principal
Example 4 8 A compound interest investment with regular yearly additions to the principal can be
modelled by the recurrence relation
V0 = 2000, Vn+1 = 1.08Vn + 1000
b How much is added to the principal each year?
c Use your calculator to determine the balance of the investment after 2 years. Round
d Plot the value of the investment after 0, 1 and 2 years on a graph.
9 A compound interest investment with regular quarterly additions to the principal can be
modelled by the recurrence relation
V0 = 20 000, Vn+1 = 1.025Vn + 2000
where Vn is the value of the investment after n quarters.
b How much is added to the principal each quarter?
c Use your calculator to determine the balance of the investment after three quarterly
payments have been made. Round your answer to the nearest cent.
d Plot the value of the investment after 0, 1, 2 and 3 quarters on a graph.
Example 5 10 Consider the compound interest investment with regular annual additions to the
principal given by the recurrence relation
V0 = 2000, Vn+1 = 1.08Vn + 1000
Determine the annual interest rate for the investment.

8A 8A Combining linear and geometric growth 399
11 Consider the compound interest investment with regular quarterly additions to the
principal given by the recurrence relation:
V0 = 20 000, Vn+1 = 1.025Vn + 2000
where Vn is the value of the investment after n quarters.
Determine the annual interest rate for the investment.

12 The value of an annuity investment, in dollars, after n years, Vn , can be modelled by the
recurrence relation shown below
V0 = 54 000, Vn+1 = 1.0055Vn + 1500
What is the value of the regular payment added to the principal of this annuity
investment?
A $55 B $297 C $1500 D $1797 E $5400
13 The value of an annuity investment, in dollars, after n quarters, Vn , can be modelled by

the recurrence relation shown below
V0 = 36 000, Vn+1 = 1.008Vn + 200
The increase in the value of this investment in the third quarter is closest to
A $200.00
B $495.84
C $499.81
D $1475.74
E $37 475.74
14 Consider the following five recurrence relations representing the value of an asset after
n years, Vn .
V0 = 10 000, Vn+1 = Vn + 1500
V0 = 10 000, Vn+1 = Vn − 1500
V0 = 10 000, Vn+1 = 1.15Vn − 1500
V0 = 10 000, Vn+1 = 1.125Vn − 1500
V0 = 10 000, Vn+1 = 1.25Vn − 1500
How many of these recurrence relations indicate that the value of an asset is
depreciating?
A 0 B 1 C 2 D 3 E 4

8B Using recurrence relations to analyse and

model reducing balance loans and annuities
Learning intentions
I To be able to model a reducing balance loan with a recurrence relation.
I To be able to use a recurrence relation to analyse a reducing balance loan.
I To be able to model an annuity with a recurrence relation.
I To be able to use a recurrence relation to analyse an annuity.
Reducing balance loans

When money is borrowed from a bank, the borrower usually makes regular payments to
reduce the amount owed, rather than waiting until the end of the loan to repay the balance.
This kind of loan is called a reducing balance loan.
Modelling reducing balance loans

Let Vn be the balance of the loan after n payments have been made. Then
V0 = the principal, Vn+1 = RVn − D
r
where D is the additional payment made, R = 1 + is the growth multiplier, r is
100 × p
the annual interest rate and p is the number of compounding periods per year.
Example 6 Modelling a reducing balance loan with a recurrence relation (1)
Flora borrows $8000 at an interest rate of 13% per annum, compounding annually. She
makes yearly payments of $2100.
Construct a recurrence relation to model this loan, in the form
where Vn is the balance of the loan after n years.
1 State V0 and D. V0 = 8000 and D = 2100
13
2 Determine the value of R using the formula R=1+ = 1.13
r 100 × 1
R=1+ , where r = 13 and p = 1.
100 × p
3 Use the values of V0 , R and D to write down the V0 = 8000, Vn+1 = 1.13Vn − 2100

8B Using recurrence relations to analyse 401
Example 7 Modelling a reducing balance loan with a recurrence relation (2)
Alyssa borrows $1000 at an interest rate of 15% per annum, compounding monthly. She
makes monthly payments of $250.
Construct a recurrence relation to model this loan, in the form
where Vn is the balance of the loan after n months.
1 State V0 and D. V0 = 1000 and D = 250
15
2 Determine the value of R using the formula R=1+ = 1.0125
r 100 × 12
R=1+ where r = 15 and p = 12.
100 × p
3 Use the values of V0 , R and D to write down the V0 = 1000, Vn+1 = 1.0125Vn −250
Once we have a recurrence relation, we can use it to determine things such as the balance of
a loan after a given number of payments.
Example 8 Using a recurrence relation to analyse a reducing balance loan
Alyssa’s loan can be modelled by the recurrence relation:

V0 = 1000, Vn+1 = 1.0125Vn − 257.85
a Use your calculator to find the balance of the loan after four payments.
b Find the balance of the loan after two payments have been made. Round your answer
a i Write down the recurrence relation. V0 = 1000, Vn+1 = 1.0125Vn − 257.85
ii Type ‘1000’ and press ‘·’ or
‘ ’. 1000 1000
iii Type ‘× 1.0125-257.85’ and press 1000 ·1.0125 − 257.85 754.65
‘·’ (or ) 4 times to obtain the 754.65·1.0125 − 257.85 506.23
screen opposite. 506.2331·1.0125 − 257.85 254.71
254.7110·1.0125 − 257.85 0.044927
|
b Read the third line of the calculator. Balance $0.04 (to the nearest cent).
$506.23 (to the nearest cent)

Annuities
An annuity is an investment where compound interest is earned and money is withdrawn
from the investment by the individual in the form of regular payments. The calculations used
to model the values of reducing balance loans and annuities are identical. The value of the
annuity represents how much money is left in the investment.
An annuity can be modelled with a recurrence relation. Once we have a recurrence relation,
we can use it to determine things such as the value of the annuity after a given number of
payments have been received.
Modelling an annuity
Let Vn be the value of the annuity after n payments have been made. Then
V0 = principal, Vn+1 = RVn − D
r
where D is the payment that has been made, R = 1 + is the growth multiplier, r
100 × p
is the annual interest rate and p is the number of compounding periods per year.
Example 9 Modelling an annuity with a recurrence relation
Reza invests $12 000 in an annuity that earns interest at the rate of 6% per annum,
compounding monthly, providing him with a monthly income of $2035.
a Model this annuity using a recurrence relation of the form

where Vn is the value of the annuity after n months.
b Use your calculator to find the value of the annuity after the first four months. Round
a i State the value of V0 and D. V0 = 12 000 and D = 2035
6
ii Determine the value of R using the R=1+ = 1.005
r 100 × 12
formula R = 1 + .
100 × p
iii Use the values of V0 , R and D to V0 = 12 000, Vn+1 = 1.005Vn − 2035
write down the recurrence relation.
12000.
12000. · 1.005 − 2035 10025.
b i Type 12000 and press · or . 10025. · 1.005 − 2035 8040.125.
ii Type × 1.005-2035 and press · or 8040.125. · 1.005 − 2035 6045.326.
four times to obtain the screen 6045.326. · 1.005 − 2035 4040.55.
opposite.

8B 8B Using recurrence relations to analyse 403
Skill-
sheet Exercise 8B
Modelling reducing balance loans with recurrence relations

Example 6 1 Brooke borrows $5000 at an interest rate of 5.4% per annum, compounding annually.
The loan will be repaid by making annual payments of $1400.
Construct a recurrence relation to model this loan, in the form:
where Vn be the balance of the loan after n years.
Example 7 2 Jackson borrows $2000 at an interest rate of 6% per annum, compounding monthly.
The loan will be repaid by making monthly payments of $339.
Let Vn be the balance of the loan after n months.
a State V0 and D.
r
b Determine the value of R, using the formula R = 1 + .
100 × p
c Use the values of V0 , R and D to write down the recurrence relation in the form:
3 Benjamin borrows $10 000 at an interest rate of 12% per annum, compounding
quarterly. The loan will be repaid with quarterly payments of $2600.
Let Bn be the balance of the loan after n quarters.
a Model this loan using a recurrence relation of the form:
B0 = the principal, Bn+1 = RBn − D

b Use the recurrence relation to determine the balance of the loan after two payments
have been made.
4 Write a recurrence relation to model a loan of $3500 borrowed at 4.8% per annum,
compounding monthly, with payments of $280 per month.
Let Vn be the balance of the loan after n months.
5 Write a recurrence relation to model a loan of $150 000 borrowed at 3.64% per annum,
compounding fortnightly, with payments of $650 per fortnight.
Let Vn be the balance of the loan after n fortnights.
6 Consider a loan of $235 000 borrowed at 3.65% per annum, compounding daily, with
payments of $150 per day.
Let Vn be the balance of the loan after n days.
a Write a recurrence relation to model this loan.
b Find the value of the loan after 3 days. Round your answer to the nearest cent.

404 Chapter 8 Reducing balance loans, annuities and investments 8B
Using a recurrence relation to analyse a reducing balance loan

Example 8 7 A reducing balance loan is modelled by the recurrence relation
V0 = 2500, Vn+1 = 1.08Vn − 626
where Vn is the balance of the loan after n years.
a State the initial balance of the reducing balance loan.
b State the payment that is made each year.
r
c Determine the annual interest rate, r, using 1.08 = 1 + .
100 × p
d Use your calculator to determine the balance of the loan after three years. Round
8 A reducing balance loan can be modelled by the recurrence relation:

V0 = 5000, Vn+1 = 1.01Vn − 865
where Vn is the balance of the loan after n months.
a State the initial balance of the reducing balance loan.
b State the payment that is made each month.
c Determine the annual interest rate, r.
d Find the balance of the loan after two payments have been made. Round your
Modelling and analysing an annuity with a recurrence relation

Example 9 9 Mark invests $20 000 in an annuity paying interest at the rate of 7.2% per annum,
compounding annually. He receives a payment of $3375 each year until the annuity is
exhausted.
Let Vn be the value of the annuity after n years.
a State the value of V0 and D.
b Determine the value of the growth multiplier, R.
c Use your values of V0 , D and R to model this annuity using a recurrence relation of
the form:
10 Sandra invests $750 000 in an annuity paying interest at the rate of 5.4% per annum,
compounding monthly. She receives a payment of $4100 per month until the annuity is
exhausted.
Let Vn be the value of the annuity after n payments have been received.
a State the value of V0 and D.
b Determine the value of the growth multiplier, R.
c Use your values of V0 , D and R to model this annuity using a recurrence relation of
the form:

8B 8B Using recurrence relations to analyse 405
11 Helen invests $40 000 in an annuity paying interest at the rate of 6% per annum,
compounding quarterly. She receives a payment of $10 380 each quarter.
Let Vn be the balance of the loan after n quarters.
a Model this loan using a recurrence relation of the form:

b Use the recurrence relation to determine the balance of the annuity after 3 quarters.
12 An annuity is modelled by the recurrence relation

V0 = 5000, Vn+1 = 1.01Vn − 1030
where Vn is the balance of the annuity after n monthly payments have been received.
a State the initial balance of the annuity.
b State the payment that is received each month.
r
c Determine the annual interest rate, r, using R = 1 + .
100 × p
d Use your calculator to determine the balance of the annuity after three payments
have been received. Round your answer to the nearest cent.
e How much will the annuity pay out in the first three months?
13 An annuity can be modelled by the recurrence relation

V0 = 6000, Vn+1 = 1.005Vn − 1500
where Vn is the balance of the annuity after n payments have been made.
a Use your calculator to determine the balance of the annuity after the two payments
have been received.
b Assuming that payments are made quarterly, how much will the annuity pay out in
the first year?
14 Jeff invests $1 000 000 in an annuity and receives a regular monthly payment.
The balance of the annuity, in dollars, after n months, An , can be modelled by a
recurrence relation of the form
A0 = 1 000 000, An+1 = 1.0024An − 4000
a State the initial balance of the annuity.
b State the payment that Jeff receives each month.
c Calculate the annual compound interest rate.
d Calculate the balance of this annuity after two months.

406 Chapter 8 Reducing balance loans, annuities and investments 8B
15 Esme invests $100 000 in an annuity and receives a regular monthly payment.
The balance of the annuity, in dollars, after n months, En , can be modelled by the
recurrence relation
E0 = 100 000, En+1 = 1.0055En − 18 400
a What monthly payment does Esme receive?
b Find the annual interest rate for this annuity.
c At some point in the future, the annuity will have a balance that is lower than the
monthly payment amount. What is the balance of the annuity when it first falls
below the monthly payment amount? Round your answer to the nearest cent.

16 Matthew would like to purchase a new home. He will establish a loan for $640 000
with interest charged at the rate of 4.2% per annum, compounding monthly.
Each month, Matthew will pay $3946.05.
Let Vn be the value of Matthew’s loan, in dollars, after n months.
A recurrence relation that models the value of Vn is
A V0 = 640 000, Vn+1 = 1.0035Vn
B V0 = 640 000, Vn+1 = 1.042Vn
C V0 = 640 000, Vn+1 = 1.042Vn − 3946.05
D V0 = 640 000, Vn+1 = 1.0035Vn − 3946.05
E V0 = 640 000, Vn+1 = 1.0035Vn + 3946.05
17 Tim invests $3800 in an annuity and receives a regular monthly payment of $480.
The balance of the annuity, in dollars, after n months, T n , can be modelled by a
T 0 = 3800, T n+1 = 1.002T n − 480
The balance of the annuity after three months is closest to
A $3327 B $3328 C $2854 D $2379 E $2380
18 Suzanne invests $9200 in an annuity at 4.2% per annum, compounding monthly.

Suzanne receives a regular monthly payment of $620.
The amount of interest earned in the second month is closest to
A $30 B $588 C $590 D $8612 E $8022

8C Amortisation tables 407
8C Amortisation tables
Learning intentions
I To be able to apply the amortisation process.
I To be able to construct an amortisation table.
I To be able to analyse an amortisation table for a reducing balance loan.
I To be able to read and interpret an amortisation table for an annuity to find the interest
rate.
I To be able to interpret and construct an amortisation table for a compound interest
investment with additions to the principal.
Amortisation tables provide additional information for each period, rather than just the
balance after each payment.
Amortisation tables for reducing balance loans

Loans that are repaid by making regular payments until the balance of the loan is zero are
called amortising loans. In an amortising loan, part of each of the regular payments goes
towards paying the interest owed on the unpaid balance of the loan with the remainder used
to reduce the principal of the loan (the amount borrowed).
For example, consider Alyssa’s loan from Example 7. Interest on the $1000 loan was
charged at the rate of 15% per year and the loan was to be repaid with monthly payments of
$250. The recurrence relation was given as
V0 = 1000, Vn+1 = 1.0125Vn − 250
To determine where the first payment goes:
15%
Calculate the interest charged (p = 12): = 1.25%. Thus, 1000 × 1.25% = $12.50
12
Calculate the principal reduction: $250 − $12.50 = $237.50
Calculate the new balance: $1000 − $237.50 = $762.50
This process is shown in the following diagram:
Make payment $250

Principal Add interest Pay Reduce New balance
(1.25%) interest principal
$1000 $12.50 $12.50 $237.50 $762.50
We can also represent this information in table format, showing the impact of a payment,
interest and the subsequent reduction of the principal to give a new balance.
Payment number Payment Interest Principal reduction Balance

1 250.00 12.50 237.50 762.50

When this analysis is repeated, the results can be summarised in an amortisation table. The
amortisation table shows all of the details that explain how the new balance was calculated.
Note that the first line shows the initial value of the loan as the balance when no payments
have been received.
The first three payments for Alyssa’s loan are shown below.

0 0.00 0.00 0.00 1000.00
1 250.00 12.50 237.50 762.50
2 250.00 9.53 240.47 522.03
3 250.00 6.53 243.47 278.56
Example 10 Applying the amortisation process
Flora borrows $20 000 at an interest rate of 8% per annum, compounding annually. She
makes annual payments of $2500.
a State the principal of the loan.
b Calculate the initial interest charged on the principal.
c Determine the impact of the first annual payment to find the principal reduction.
d Calculate the new balance.
e Complete the row in the table below with your calculations.

1 2500.00
a Read the principal from the The principal is $20 000.
question or recurrence
relation.
b Calculate the interest paid. Interest paid = 8% of $20 000 = $1600
c Principal reduction = payment − Principal reduction = 2500 − 1600
interest. = $900
d New balance = balance owing − New balance = 20 000 − 900 = $19 100
principal reduction
e Place each of the numbers from
Interest Principal reduction Balance
the calculations into the
1600.00 900.00 19 100.00
relevant boxes.

Constructing an amortisation table for a reducing balance loan

r
1 Interest paid = × previous balance, where r is the annual interest rate and p is
100 × p
the number of compounding periods each year.
2 Principal reduction = payment − interest
3 New balance = (previous) balance − reduction in balance
Example 11 Constructing an amortisation table for a reducing balance loan
Flora borrows $20 000 at an interest rate of 8% per annum, compounding annually. She
makes annual payments of $2500.
Construct an amortisation table for Flora’s reducing balance loan for the first three
payments.
Solution
Repeat the calculations from Example 10, rounding all numbers to the nearest cent.
Once the new balance has been calculated, repeat the process for the first three payments.

0 0.00 0.00 0.00 20 000.00
1 2500.00 1600.00 900.00 19 100.00
2 2500.00 1528.00 972.00 18 128.00
3 2500.00 1450.24 1049.76 17 078.24
Sometimes we are asked to fill in gaps of a given amortisation table.
Example 12 Analysing an amortisation table for a reducing balance loan
A business borrows $10 000 at a rate of 8% per annum, compounding quarterly. The loan
is to be repaid by making quarterly payments of $2700.00. The amortisation table for this
loan is shown below.
0 0.00 0.00 0.00 10 000.00
1 2700.00 2500.00 7500.00
2 2700.00 150.00 4950.00
3 2700.00 99.00 2601.00
a Calculate the interest paid on the initial balance.

b Calculate the principal reduction from the second payment.
c Calculate the balance of the loan after payment 3 has been made.

r 8
a Use , where r = 8 and p = 4 Interest paid = × 10 000 = $200
100 × p 100 × 4
since interest is calculated quarterly.
Interest paid = 2% × unpaid balance
Alternatively, note that $2700 is paid Or, Interest paid = $2700 − $2500 = $200
and the principal was reduced by $250.
b Principal reduction = payment − Principal reduction = 2700.00 − 150.00
interest = $2550.00
c New Balance = balance owing − Balance of the loan after 3 payments
principal reduction = 4950 − 2601 = $2349
Amortisation tables for annuities

An amortisation table for an annuity is very similar to one for a reducing balance loan. Each
row shows the payment number, the payment received, the interest earned, the principal
reduction and the balance of the annuity after each payment has been received.
Example 13 Analysing an amortisation table for an annuity to find the interest

rate
Consider the following amortisation table for an annuity after 3 monthly payments.

0 0.00 0.00 0.00 12 000.00
1 2200.00 60.00 2140.00 9860.00
2 2200.00 49.30 2150.70 7709.30
3 2200.00 A B 5547.85
a State the principal of the annuity and the amount of interest paid in the first month.
b Calculate the monthly interest rate.
c Find the value of A and B.
a Read off from the table. Principal: $12 000, Interest: $60
Interest 60
b Calculate: × 100 × 100 = 0.5% per month
Principal 12000
0.5
c A is the interest due on $7709.30 A: × 7709.30 = 38.55
100
B is the principal reduction after the B: 2200 − 38.55 = 2161.45
third payment.

Amortisation tables for compound interest investments

with additions to the principal
An amortisation table for a compound interest investment with additions to the principal is
similar to the previous examples but here, the payment increases the balance of the principal
further.
Example 14 Interpreting and constructing an amortisation table for a compound

interest investment with additional payments
Consider the following amortisation table for a compound interest investment with
monthly additions to the principal. Assume that interest compounds monthly.
Payment number Payment Interest Principal increase Balance

0 0.00 0.00 0.00 1200.00
1 50.00 3.00 53.00 1253.00
2 50.00 3.13 53.13 1306.13
3 50.00 3.27 53.27 1359.40
Complete two additional lines for the table corresponding to payment 4 and payment 5.
Solution
3
Begin by calculating the monthly interest rate × 100 = 0.25%.
1200
Now we can calculate the line associated with payment 4. The interest paid is calculated
on the balance $1359.40:
Interest = 0.25% × 1359.40 = $3.40
The principal increases by the interest and the additional payment:
Principal increase = interest + payment = 3.40 + 50 = $53.40
Thus, the new balance becomes:
New balance = previous balance + principal increase = 1359.40 + 53.40 = $1412.80
Repeating gives the following two lines of the table.

4 50.00 3.40 53.40 1412.80
5 50.00 3.53 53.53 1466.33

412 Chapter 8 Reducing balance loans, annuities and investments 8C
Exercise 8C
Applying the amortisation process and constructing an amortisation table for a

reducing balance loan
Example 10 1 Walter borrows $14 000 at an interest rate of 11% per annum, compounding annually.
Example 11 He makes annual repayments of $1800 per year.
b Calculate the interest charged on the principal in the first year.
c Determine the impact of the first annual payment to find the principal reduction.
d Calculate the new balance after the first year.
e Complete the row in an amortisation table corresponding to payment 1.

0 0.00 0.00 0.00 14 000.00
1 1800.00
f Complete the next two rows of the amortisation table corresponding to payment 2
and 3 for Walter.
2 Ellie borrows $12 000 at an interest rate of 6% per annum, compounding monthly. She
makes regular repayments of $300 per month.
r
b Calculate where r is the annual interest rate and p is the number of
100 × p
compounding periods each year.
c Calculate the interest charged on the principal in the first month.
d Determine the impact of the first monthly payment to find the principal reduction.
e Calculate the new balance after the first month.
f Complete the row in an amortisation table corresponding to payment 1.

0 0.00 0.00 0.00 12 000.00
1 300.00
g Complete the next two rows of the amortisation table corresponding to payment 2
and 3 for Ellie.

8C 8C Amortisation tables 413
3 Anna borrows $36 000 at an interest rate of 8% per annum, compounding quarterly.
She makes regular repayments of $1000 per quarter.
r
b Calculate where r is the annual interest rate and p is the number of
100 × p
compounding periods each year.
c Construct an amortisation table corresponding to the first three payments for the
loan.
Reading and interpreting an amortisation table for a reducing balance loan

Example 12 4 A student borrows $2000 at an interest rate of 12% per annum, compounding monthly.
The student makes monthly payments of $345.
The amortisation table for this loan after 5 payments is shown below.

0 0.00 0.00 0.00 2000.00
1 345.00 20.00 325.00 1675.00
2 345.00 A 328.25 1346.75
3 345.00 13.47 331.53 1015.22
4 345.00 10.15 B 680.37
5 345.00 6.80 338.20 C
a Calculate the monthly interest rate.

b Determine the values of A, B and C.
5 The amortisation table for a loan with quarterly payments is shown below.

0 0.00 0.00 0.00 4000.00
1 557.65 100.00 457.65 3542.35
2 557.65 88.65 469.09 3073.26
3 557.65 76.83 480.82 2592.44
4 557.65 A 492.84 2099.60
5 557.65 52.49 B 1594.44
6 557.65 39.86 517.79 C
7 557.65 D E 545.92
a State the principal and interest paid in the first quarter.

b Calculate the quarterly interest rate.
c Determine the values of A, B, C, D and E.

414 Chapter 8 Reducing balance loans, annuities and investments 8C
6 Ada has a reducing balance loan with an interest rate of 3.6% per annum, compounding
monthly. She makes monthly payments of $1800 as shown in the amortisation table
below.
0 0.00 0.00 0.00 460 000.00
1 1800.00 1380.00 420.00 459 580.00
2 1800.00 1378.74 A 459 158.74
3 1800.00 B
Calculate the value of A and B.
Analysing an amortisation table for an annuity to find the interest rate

Example 13 7 A student invested $6000 in an annuity paying an interest rate of 3% per annum,
compounding monthly.
She receives a monthly payment of $508 as per the amortisation table shown below.

0 0.00 0.00 0.00 6000.00
1 508.00 15.00 493.00 5507.00
2 508.00 13.77 494.23 5012.77
3 508.00 A B C
a Reading from the table, determine:

i the interest earned when payment 1 is received,
ii the monthly interest rate.
b Calculate the values of A, B and C.
Interpreting and constructing an amortisation table for a compound investment with

additional payments
Example 14 8 The amortisation table below charts the growth of an investment which compounds
monthly and has regular additions made to the balance each month.

0 0.00 0.00 0.00 5000.00
1 100.00 50.00 150.00 5150.00
2 100.00 51.50 151.50 5301.50
3 100.00 A B C
a Calculate the monthly interest rate of the investment.

b Determine the values of A, B and C.

8C 8C Amortisation tables 415

9 Edith invested $400 000 in an annuity that provides an annual payment of $51 801.82
Interest is calculated annually.
The first five lines of the amortisation table are shown below.

0 0.00 0.00 0.00 400 000.00
1 51 801.82 20 000.00 31 801.82 368 198.18
2 51 801.82 18 409.91 33 391.91 334 806.27
3 51 801.82 16 740.31 299 744.76
4 51 801.82 14 987.24 36 814.58 262 930.18
The principal reduction associated with payment number 3 is

A $31 801.82 B $33 391.82 C $35 061.50 D $35 061.51 E $36 814.58
10 Consider the following amortisation table for a reducing balance loan.

0 0.00 0.00 0.00 200 000.00
1 3000.00 1800.00 1200.00 198 800.00
2 3000.00 1789.20 1210.80 197 589.20
3 3000.00 11778.30 1221.70 196 367.50
The annual interest rate for this loan is 3.6%.

Interest is calculated immediately before each payment.
For this loan, the repayments are made
A weekly B fortnightly C monthly D quarterly E yearly
11 Four lines of an amortisation table for an annuity investment are shown below.
The interest rate for this investment remains constant, but the amount of the additional
payment may vary.

1 50.00 50.00 100.00 10 100.00
2 50.00 50.50 100.50 10 200.50
3 50.00 51.00 101.00 10 301.50
4 10 553.01
The balance of the investment after payment number 4 is $10 533.01.

The value of payment number 4 is
A $50 B $100 C $150 D $200 E $250

8D Analysing financial situations using amortisation

tables
Learning intentions
I To be able to find the final payment in a reducing balance loan and an annuity.
I To be able to find the total payment made/received and the total interest paid/earned.
I To be able to plot points using an amortisation table for a reducing balance loan, an
annuity and a compound interest investment with additional payments.
Finding the final payment in a reducing balance loan and

an annuity
A reducing balance loan must be paid off at some point in time and likewise, an annuity will
run out after a period of time. In both cases, it is possible that the final payment might need
to be different from the regular payment so that the final balance can reach zero.
Example 15 Finding the final payment for a reducing balance loan or annuity
Consider the following amortisation table for a reducing balance loan of $20 000 with an
interest rate of 8% per annum, compounding annually. Regular payments of $5009.12 are
made for the first four years as shown in the amortisation table.

0 0.00 0.00 0.00 20 000.00
1 5009.12 1600.00 3409.12 16 590.88
2 5009.12 1327.27 3681.85 12 909.03
3 5009.12 1032.72 3976.40 8932.63
4 5009.12 714.61 4294.51 4638.12
Calculate the final payment required in the fifth year to pay off the loan fully.
Solution
To find the final payment, we first need to calculate the interest on $4638.12:
8
Interest = × $4638.12 = $371.05
100
Thus, the final payment can be calculated by adding the balance that is still due and the
interest.
Final payment = Balance + interest
= 4638.12 + 371.05
= 5009.17
Thus, the final payment is $5009.17.

8D Analysing financial situations using amortisation tables 417
Finding the total payment made/received and total

interest paid/earned
Once we have a completed amortisation table, we can find the total interest paid/earned and
the total payment made/received for a reducing balance loan, an annuity or a compound
interest investment with additional payments.
Example 16 Finding the total payment made and total interest paid
Consider the following amortisation table for a reducing balance loan of $10 000 with
an interest rate of 8% per annum, compounding quarterly. Three quarterly payments of
$2626 are made.
0 0.00 0.00 0.00 10 000.00
1 2626.00 200.00 2426.00 7574.00
2 2626.00 151.48 2474.52 5099.48
3 2626.00 101.99 2524.01 2575.47
4 A B C 0.00
a Complete the amortisation table corresponding to payment four such that the final
payment ensures that the balance is 0.
b Calculate the total payment made for the loan.
c Calculate the total interest paid on the loan.
8
a Follow the previous example B: Interest = × $2575.47 = $51.51
100 × 4
by first finding the interest
C: Principal reduction = $2575.47
applied to the loan (B), the
principal reduction (C) and A: Final payment = 2575.47 + 51.51
the final adjusted payment = 2626.98
(A). Payment Interest Principal reduction
2626.98 51.51 2575.47
b Add up all of the payments Payments = 2626 × 3 + 2626.98 = $10 504.98
made over the four quarters.
c Subtract the principal from the Interest = 10 504.98 − 10 000 = $504.98
total payments. OR
Alternatively, we can add up Interest = 200 + 151.48 + 101.99 + 51.51 = $504.98
the interest column.

Plotting points from an amortisation table

The values in an amortisation table can be plotted to give a picture of the impact of regular
payments.
Example 17 Plotting from an amortisation table
Consider the following amortisation table for a reducing balance loan.

0 0.00 0.00 0.00 1000.00
1 257.85 12.50 245.35 754.65
2 257.85 9.43 248.42 506.23
3 257.85 6.33 251.52 254.71
4 257.89 3.18 254.71 0.00
Plot a graph of the interest and principal reduction on the same graph.
Solution
For this loan, we can plot a graph for each
300
payment period to show that the amount
of interest paid each payment (blue dots)
declines while the amount of principal paid
increases (red dots). 200
Amount ($)
100
O 1 2 3 4 5
Payment (n)
Note that for a loan, the graph above shows how the amount of interest paid for each
payment (blue dots) decreases with the payment number, while the amount of principal paid
off increases (red dots). This is because the balance is decreasing and so the interest is being
calculated on a lower balance each period.
For a compound interest investment, we would expect to see the interest earned each period
increase because the balance increases each time a payment is made.

8D 8D Analysing financial situations using amortisation tables 419
Exercise 8D
Finding the final payment in a reducing balance loan and an annuity

Example 15 1 Consider the following amortisation table for a reducing balance loan of $2000 with an
interest rate of 4% per annum, compounding annually. Regular payments of $550 are
made for the first three years as shown in the amortisation table.
0 0.00 0.00 0.00 2000.00
1 550.00 80.00 470.00 1530.00
2 550.00 61.20 488.80 1041.20
3 550.00 41.65 508.35 532.85
4 0.00
Calculate the value of the fourth payment to ensure the loan is repaid in full.
2 Consider the following amortisation table for a reducing balance loan of $5000 with an
interest rate of 4.8% per annum, compounding monthly. Regular payments of $1262.50
are made for the first three months as shown in the amortisation table.
0 0.00 0.00 0.00 5000.00
1 1262.50 20.00 1242.50 3757.50
2 1262.50 15.03 1247.47 2510.03
3 1262.50 10.04 1252.46 1257.57
4 0.00
Calculate the value of the fourth payment to ensure the loan is repaid in full.
3 Consider the following amortisation table for an annuity of $3000 with an interest rate
of 5% per annum, compounding annually. Regular payments of $693 are received for
the first four months as shown in the amortisation table.
0 0.00 0.00 0.00 3000.00
1 693.00 150.00 543.00 2457.00
2 693.00 122.85 570.15 1886.85
3 693.00 94.34 598.66 1288.19
4 693.00 64.41 628.59 659.60
Calculate the value of the fifth payment to ensure the annuity is completely exhausted.

420 Chapter 8 Reducing balance loans, annuities and investments 8D
4 Charlie invests $4500 into an annuity with an interest rate of 5.4% per annum,
compounding monthly. He receives monthly payments of $760 for five months.
Calculate the value of the sixth payment that Charlie receives to ensure the annuity is
completely exhausted.
Finding the total payment made/received and the total interest paid/earned
Example 16 5 Consider the following amortisation table for an investment of $11 000 with an interest
rate of 4.8% per annum, compounding quarterly. Regular quarterly payments of $1200
are added each quarter as shown below for the first four quarters.
0 0.00 0.00 0.00 11 000.00
1 1200.00 132.00 1332.00 12 332.00
2 1200.00 147.98 1347.98 13 679.98
3 1200.00 164.16 1364.16 15 044.14
4 1200.00 A B 16 424.67
a Find the value of A.
b Find the value of B.
c Find the total interest earned on the investment in the first four quarters.
6 Consider the following amortisation table for a reducing balance loan of $4000 with an
interest rate of 6% per annum, compounding monthly. A monthly payment of $344.14
is made for the first 11 months.
9 344.14 6.81 337.33 1023.71
10 344.14 5.12 339.02 684.69
11 344.14 3.42 340.72 343.97
12 A B C 0.00
a State the value of A, B and C in the amortisation table corresponding to payment

twelve such that the final payment ensures that the balance is 0.
b Calculate the total payment made for the loan.

8D 8D Analysing financial situations using amortisation tables 421
7 Consider the final two lines of the amortisation table for an annuity of $30 000 with
an interest rate of 3.6% per annum, compounding quarterly. A quarterly payment of
$3903.50 is made for the first 7 quarters.
7 3903.50 69.32 3834.18 3868.37
8 A B C 0.00
a State the value of A, B and C in the amortisation table corresponding to payment

eight such that the final payment ensures that the balance is 0.
b Calculate the total payments received from the annuity.
8 Tania invests $12 000 in an annuity with an interest rate of 6.6% per annum,
compounding monthly. She receives regular monthly payments of $3040 per month
for three months followed by a final payment in the fourth month. Calculate the total
payment and the interest for the annuity.
Plotting points from an amortisation table

Example 16 9 Consider the following amortisation table for a reducing balance loan.

0 0.00 0.00 0.00 2000.00
1 345.14 20.00 325.14 1674.86
2 345.14 16.75 328.39 1346.47
3 345.14 13.46 331.68 1014.79
4 345.14 10.15 334.99 679.80
5 345.14 6.80 338.34 341.46
6 344.87 3.41 341.46 0.00
Plot a graph of the interest and principal reduction on the same graph for the first six
payments.
10 The amortisation table below charts the growth of a compound interest investment with
regular additions made to the principal each month.

0 0.00 0.00 0.00 5000.00
1 100.00 50.00 150.00 5150.00
2 100.00 51.50 151.50 5301.50
3 100.00 53.02 153.02 5454.52
Plot a graph of the interest and principal increase on the same graph.

422 Chapter 8 Reducing balance loans, annuities and investments 8D

11 Francesca invests $6000 into an annuity with an interest rate of 5.8% per annum,
compounding quarterly.
She receives a quarterly payment of $1554.13 for the first three quarters and then a
final payment in the fourth quarter.
After three quarters, the value of the annuity is $1534.48.
The payment that Francesca receives in the fourth quarter is
A $0.04 B $2.60 C $1534.48 D $1554.13 E $1556.73
12 Ned borrows $15 000 at an interest rate of 6% per annum, compounding monthly. He
makes regular monthly payments of $3796.99 for three months followed by a final
payment in the fourth month.
The total interest that Ned pays on the loan is closest to
A $56 B $75 C $188 D $300 E $900
8E Using a finance solver to find the balance and

final payment
Learning intentions
I To be able to use the finance solver to find the value of a compounding interest
investment with additional payments.
I To be able to use the finance solver to find the balance and final payment of a reducing
balance loan.
I To be able to use the finance solver to find the balance of an annuity.
While the techniques used so far in this chapter are useful for performing a small number
of calculations, they are tedious over a long period. For example, a typical home loan may
involve monthly payments over 30 years. CAS calculators have a Finance Solver that allow
for larger calculations to be performed with ease.

8E Using a finance solver to find the balance and final payment 423
Using the Finance Solver on the TI-Nspire CAS
Steps
1 Press ctrl + N
2 Select Add Calculator.

Press b>Finance> Finance Solver.
3 To use Finance Solver you need to know the

meaning of each of its symbols.
N is the total number of payments.
I(%) is the annual interest rate.
PV is the present value of the loan/investment.
Pmt is the amount paid at each payment.
FV is the future value of the loan/investment.

PpY is the number of payments per year. Note: Use e or ¤ to move down
CpY is the number of times the interest is boxes, press £ to move up. For PpY
and CpY press e to move down to
compounded per year. (It is the same as PpY.)
the next entry box.
PmtAt is used to indicate whether the interest
is compounded at the end or at the beginning
of the time period. Ensure this is set at END.
4 When using Finance Solver to solve loan and
investment problems, there will be one unknown
quantity. To find its value, move the cursor to its
entry field and press · to solve.
In the example shown, pressing · will solve
for Pmt.

Using the Finance Solver on the Casio ClassPad
Steps
1 Tap Financial from the main
menu screen.
2 Select the compound interest
solver by tapping on Compound
Interest from the solver screen.
3 To use Finance Solver you need to know the meaning of

each of its symbols.
N is the total number of payments.
I% is the annual interest rate.
PV is the present value of the loan or investment.
PMT is the amount paid at each payment.
FV is the future value of the loan or investment.
P/Y is the number of payments per year.
C/Y is the number of times interest is
compounded per year. (It is the same as P/Y.)
4 Tap Format and confirm that the setting for ‘Odd Period’ is
set to ‘off’ and ‘Payment Date’ is set to ‘End of period’.
5 When using Finance Solver to solve loan problems, there
will be one unknown quantity. To find its value, tap its
entry field and tap Solve.
In the example shown, tapping Solve will solve for Pmt.

Using a financial solver to analyse a compound interest

investment with regular additions to the principal
A finance solver is a powerful computation tool. However, you have to be very careful in
the way you enter information because it needs to know which way the money is flowing. It
does this by following a sign convention.
In general terms:
if you receive money, or someone owes you money, we treat this as a positive (+ve)
if you pay out money or you owe someone money, we treat this as a negative (−ve).
Recall that a compound interest investment with regular additions to the principal is an
investment where the balance increases through both the interest earned and the additional
payments.
Using Finance solver for a compound interest investment with regular

additions to the principal
In finance solver:
PV: Negative: you make an investment by giving the bank some money.
PMT: Negative: you make regular payments to the bank.
FV: Positive: after the payment is made and the investment matures, the bank will
give you the money.
Example 18 Determining the value of an investment with regular additions made

to the principal using a financial solver
Lars invests $500 000 at 5.5% per annum, compounding monthly. He makes a regular
deposit of $500 per month into the account. What is the value of his investment after
5 years? Round your answer to the nearest cent.
1 Open Finance Solver and enter the information 60
below, as shown opposite. 5.
5
N: 60 (5 years) −500000
−500
I%: 5.5
692292.
297.
..
PV: –500 000 (you give this to the bank)
12
PMT: –500 (you give this to the bank) 12
FV: to be determined
Pp/Y: 12 payments per year
Cp/Y: 12 compounding periods per year
2 Solve for FV and write your answer, rounding to After 5 years, Lars’ investment
the nearest cent. Note that this is positive as the will be worth $692 292.30.
bank will give this money to you.

Using finance solver for a reducing balance loan

Recall that for a reducing balance loan, you receive money from the bank (in the form of
the loan) and then make payments until you no longer owe the bank any money. The sign
convention for a reducing balance loan is summarised below.
Finance solver for a reducing balance loan

In finance solver:
PV: Positive: the bank gives you money through a loan.
PMT: Negative: you repay the loan by making regular repayments to the bank.
FV: Negative, zero or positive: after the payment is made:
• you still owe the bank money (FV negative),
• the loan is fully repaid (FV zero), or
• you have overpaid your loan and the bank now owes you money (FV positive).
Example 19 Determining the balance and final payment of a reducing balance

loan after a given number of payments
Andrew borrows $20 000 at an interest rate of 7.25% per annum, compounding monthly.
This loan will be repaid over 4 years with regular payments of $481.25 each month for 47
months followed by a final payment to fully repay the loan.
a How much does Andrew owe after 3 years? Round your answer to the nearest cent.
b What is the final payment amount that Andrew must make to fully repay the loan
within 4 years (48 months)? Round your answer to the nearest cent.
a 1 Open Finance Solver and enter the following:
N: 36 (number of months in 3 years)

I%: 7.25 (annual interest rate)
PV: 20000 (positive to indicate that this is money
received by Andrew from the lender)
Pmt or PMT: −481.25 (negative as Andrew is giving
this back to the lender)
Pp/Y: 12 (monthly payments)
Cp/Y: 12 (interest compounds monthly)

2 Solve for the unknown future value (FV). On the:

TI-NspireCAS: Move the cursor to the FV entry box
and press to · solve.
ClassPad: Tap on the FV entry box and tap ‘Solve’.
The amount –5554.3626 . . . now appears in the FV
entry box.
Note: A negative FV indicates that Andrew will still owe the
lender money after the payment has been made.
3 Write your answer, correct to the nearest cent Andrew owes $5554.36.
b 1 Enter the information below.
N: 48 (number of months in 4 years)
PV: 20000
Pmt or PMT (the payment amount is negative):
−481.25
2 Solve for the unknown future value (FV). On the:
TI-NspireCAS: Move the cursor to the FV entry box
and press · to solve.
ClassPad: Tap on the FV entry box and tap Solve
The amount 0.1079. . . (11 cents) now appears in the
FV entry box.
Since FV is positive (+11 cents), the bank owes Andrew

11 cents so we subtract this from the regular payment. Final payment
3 Write your answer. = $481.25 − $0.11
= $481.14
Andrew’s final payment will
be $481.14.
Using a finance solver to analyse an annuity

Recall that an annuity is when an individual puts money into an account and then receives
payments from the bank.

Finance solver for an annuity

In finance solver:
PV: Negative: you buy an annuity by giving the bank some money.
PMT: Positive: you receive regular payments from the bank.
FV: Positive or zero: after the payment is made:
• the bank still owes you money (FV positive),
• the annuity is fully paid out (FV zero)
Note: An annuity should never have a negative FV as a bank would never overpay the individual.
Example 20 Determining the balance of an annuity using a finance solver
Charlie invests $300 000 into an annuity, paying 5% interest per annum, compounding
monthly. Over the next ten years, Charlie receives a payment of $3182 per month from
the annuity for each month except the final month.
a Find the value of the annuity after five years. Round your answer to the nearest cent.
b Find the final payment from the annuity. Round your answer to the nearest cent.
N: 60 (number of monthly payments in 5 years)
PV: −300000 (negative to indicate that this is
money paid by Charlie to the bank)
Pmt or PMT: 3182 (positive to indicate that the
bank is paying back to Charlie)
2 Solve for the unknown future value (FV). On the: 60
5.
00
TI-NspireCAS: Move the cursor to the FV entry
−300000
box and press to · solve. 3182
ClassPad: Tap on the FV entry box and tap 168612.
247951
‘Solve’. The amount 168612.24795 . . . now 12

12
appears in the FV entry box.
Note: A positive FV indicates that Charlie is still owed money The balance of the annuity is
from the annuity. $168 612.25
3 Write your answer, rounding to the nearest cent.

8E 8E Using a finance solver to find the balance and final payment 429
c 1 Find the value of the annuity after 120 payments.

b
Enter the information below, as shown opposite.
N: 120 (number of monthly payments in 10
years)
I%: 5.00
PV: −300000
Pmt or PMT: 3182
Pp/Y: 12
Cp/Y: 12
2 Solve for the unknown future value (FV). On the: 120
TI-NspireCAS: Move the cursor to the FV entry 5.
00
box and press · to solve. −300000

3182
ClassPad: Tap on the FV entry box and tap Solve
−5.
36388
The amount −5.36 (−$5.36) now appears in the
12
FV entry box. 12
Note: The FV is negative ($5.36). This means that Charlie
owes the annuity $5.36. To compensate, Charlie’s final
Final payment
payment will be decreased by $5.36. = $3182 − $5.36
= $3176.64
3 Write your answer. Charlie’s final payment will
be $3176.64.
Skill-
sheet Exercise 8E
Determining the value of an investment with regular additions made to the principal
using a financial solver
Example 18 1 Wanda invested $20 000 at 7.1% per annum, compounding annually. She makes a
regular deposit of $6000 per year into the account.
a State whether the PV is positive or negative.
b State whether the PMT is positive or negative.
c Find the value of the investment after 10 years. Round your answer to the nearest
cent.
d Find the value of the investment after 30 years. Round your answer to the nearest
cent.

430 Chapter 8 Reducing balance loans, annuities and investments 8E
2 Ingrid invested $20 000 at 4.9% per annum, compounding monthly. She makes a
regular deposit of $380 per month into the account.
a Find the value of the investment after 5 months. Round your answer to the nearest
cent.
b Find the value of the investment after 3 years. Round your answer to the nearest
cent.
Determining the balance of a reducing balance loan using a financial calculator

Example 19 3 Barry borrows $8000 at an interest rate of 4.5% per annum, compounding monthly.
This loan will be repaid with regular payments of $350 per month, followed by a final
payment.
b State whether the PMT is positive or negative.
c How much does Barry owe after six months? Round your answer to the nearest
cent.
4 Suzanne borrows $25 000 at an interest rate of 7.8% per annum, compounding
monthly. She repays the loan with regular payments of $1200 per month and then a
final payment to bring the balance to zero.
a How much does Suzanne owe after 3 months? Round your answer to the nearest
cent.
b How much does Suzanne owe after 1 year? Round your answer to the nearest cent.
5 Rachel borrows $240 000 at an interest rate of 8.3% per annum, compounding
quarterly. She makes 119 regular payments of $5442.90 each quarter followed by a
final payment. Rachel repays the loan over thirty years.
a How much does Rachel owe after 6 years? Round your answer to the nearest cent.
b What is the final payment that Rachel must make to fully repay the loan in 30 years?
6 David borrows $50 000 for a new car at an interest rate of 4.6%, compounding weekly.
He repays the loan over 3 years with regular payments of $343.27 per week except for
the final payment.
a How much does David owe after 1 year? Round your answer to the nearest cent.
b What is the final payment that David must make to fully repay the loan in three
years? Round your answer to the nearest cent.

8E 8E Using a finance solver to find the balance and final payment 431
Determining the balance of an annuity using a financial calculator

Example 20 7 Kazou invests $50 000 into an annuity, paying 6.1% per annum, compounding
annually. Kazou receives $6825.61 per year for nine years then a final payment so that
the annuity lasts exactly 10 years.
b Find the value of the annuity after five years. Round your answer to the nearest cent.
c What is the final payment made to Kazou so that the value of the annuity is zero
after 10 years? Round your answer to the nearest cent.
8 Eliza invests $20 000 into an annuity, paying 7.2% per annum, compounding monthly.
The annuity regularly pays $1732.37 per month, for eleven months followed by a final
payment to exhaust the annuity.
a Find the value of the annuity after three months. Round your answer to the nearest
cent.
b What is the final payment made to Eliza so that the value of the annuity is zero after
12 months? Round your answer to the nearest cent.
9 Ezra is going backpacking around Europe and has invested $15 000 into an annuity for
this trip. The annuity pays 6.8% per annum, compounding weekly for one year. The
annuity pays $298.57 per week for each week except for the final payment.
a Find the value of the annuity after twenty-six weeks. Round your answer to the
nearest cent.
b What is the final payment made to Ezra so that the value of the annuity is zero after
1 year? Round your answer to the nearest cent.

10 Josie invests $3000 in an account that pays interest at the rate of 2.8% per annum,
compounding monthly. She makes an additional payment of $200 each month.
The value of the investment, correct to the nearest cent, after 6 years is
A $4249.26 B $4249.27 C $12 112.32 D $19 208.55 E $19 208.56
11 Bronwyn borrows $450 000 to buy an apartment.

The interest rate for this loan was 4.24% per annum, compounding monthly for
20 years.
Bronwyn makes regular monthly payments of $2784 each month except for her final
payment. To pay out her loan fully in 20 years, her final payment is
A $58.57 B $2725.42 C $2725.43 D $2784.00 E $2842.57
12 Benjamin invests $75 000 in an annuity, paying 7.3% per annum, compounding
monthly.
Benjamin receives a payment of $2326 each month from the annuity.
The value of the annuity, correct to the nearest cent after 2 years is
A $26842.05 B $26842.06 C $69356.55 D $69356.56 E $85153.41
8F Using a finance solver to find interest rates, time

taken and regular payments
Learning intentions
I To be able to use the finance solver to find the interest rate or time taken for a
compounding interest investment with additional payments.
I To be able to use the finance solver to find the regular payment for a reducing balance
loan.
I To be able to use the finance solver to find the interest rate, time taken or regular
payment for an annuity.
As well as finding the future value or balance of a loan, annuity or investment, we can
also use the financial solver to find the interest rate, regular payment or length of the loan,
annuity or investment.
Using finance solver for investments with additional

payments to find interest rates and time taken
Recall that for an investment with compound interest and additional payments:
PV: Negative: you make an investment by giving the bank some money.
PMT: Negative: you make regular payments to the bank.
FV: Positive: when the investment matures, the bank gives you the money.
Example 21 Finding the interest rate for an investment with additional payments
Mingjia puts $20 000 into a compound interest investment where interest compounds
monthly. She adds $50 per month. She wants her investment to reach $40 000 in 10 years.
Find the annual interest rate required for this to occur. Round your answer to two decimal
places.
1 Open finance solver and enter the following: 120
N: 120 (10 years) 4.
807676
PV: −20000 −20000
PMT: −50 −50

40000
FV: 40000 (the annuity will be exhausted
12
after 10 years)
12
2 Solve for I.
3 Write your answer. Minjia would require an interest
rate of 4.81% per annum.

8F Using a finance solver to find interest rates, time taken and regular payments 433
When using a financial solver, rounding is very important so it is always a good idea to
check your answer in case you need to round up or down accordingly.
Example 22 Finding the regular monthly payment and time taken for an
investment with additions to the principal
Winston puts $20 000 into an investment, paying 5.1% interest per annum, compounding
monthly.
a If Winston wants his investment to be worth at least $40 000 in 5 years, what is the
minimum he will need to add each month?
b If Winston invests $1000 each month immediately after interest is calculated, what is
the minimum number of months required for his investment to at least triple in value?
a 1 Open finance solver and enter the following:
N: 60 (5 years)
PV: −20000
FV: 40000 (the annuity will be exhausted
after 10 years)
2 Solve for Pmt or PMT.
Note: The sign of Pmt or PMT is negative, because it
is money that Winston invests.
3 Write your answer, noting that $208.34 per
month is insufficient as it gives a balance of
$39999.89. . . Winston will add $208.35 each
month to the investment.
b 1 Change the payment Pmt or PMT to −1000
and the FV to 60 000 and solve for N.
2 Write your answer, noting that 34 months has The value of Winston’s
a FV of $59598.147. . . so we need to round investment will take 35 months to
up triple.

Analysing a reducing balance loan with financial solver

Recall that for a reducing balance loan:
PV: Positive - the bank gives you money through a loan
PMT: Negative - you repay the loan by making regular repayments
FV: Negative, zero or positive - balance after the payment is made.
Example 23 Determining the payment amount, total repayment and total amount
of interest paid for a reducing balance loan
Sipho borrows $10 000 to be repaid in 59 equal monthly payments followed by a 60th
payment of less than one dollar more than the regular payment. Interest is charged at the
rate of 8% per annum, compounding monthly.
a Find the regular monthly payment amount. Round your answer to the nearest cent.
b Find the final payment. Round your answer to the nearest cent.
c Find the total of the repayments on the loan. Round your answer to the nearest cent.
d Find the total amount of interest paid. Round your answer to the nearest cent.
N: 60 (number of monthly payments in 5
years, assuming 60 equal payments)
I%: 8 (annual interest rate)
PV: 10000
FV: 0 (the balance will be zero when the
loan is repaid)
2 Solve for the unknown future value (Pmt or
PMT). On the:
TI-Nspire: Move the cursor to the Pmt
entry box and press · to solve.
ClassPad: Tap on the PMT entry box and
tap Solve.
The amount –202.7639. . . now appears in the
Pmt or PMT entry box.
Note: The sign of the payment is negative to indicate
that this is money Sipho is giving back to the lender.
3 Write your answer. Sipho repays $202.76 as the
regular payment.

8F Using a finance solver to find interest rates, time taken and regular payments 435
b To find the final payment:

1 Find the final value after 60 payments of Final payment
$202.76. = 202.76 + 0.29
= 203.05
2 Since FV is −0.289, the final payment is 0.29 Final payment is $203.05
more than the regular payment.
c Total of repayments of the loan = 59 × regular Total of repayments = 59 ×
payment + final payment 202.76 + 203.05 = $12 165.89
d Total interest = total repayments − the principal Interest paid = 12 165.89 −
10 000 = $2165.89
Analysing an annuity with financial solver

Recall that for an annuity:
PV: Negative: you buy an annuity by giving the bank some money.
PMT: Positive: you receive regular payments from the bank.
FV: Positive or zero: balance after the payment is made.
Example 24 Finding the interest rate, time taken and regular payment for an
annuity
Joe invests $200 000 into an annuity, with interest compounding monthly.
a What interest rate would allow Joe to withdraw $2500 each month for 10 years?
Round your answer to one decimal place.
b Assume the interest rate is 5% per annum and that Joe receives a regular monthly
payment of $3000. For how many months will Joe receive a regular payment?
c Assume that the interest rate is 5% per annum and that Joe wishes to be paid monthly
payments for 10 years. How much will he regularly receive each month?
d If Joe receives the regular monthly payment found in part c for 119 months, what will
his final payment be? Round your answer to the nearest cent.
a 1 Open finance solver and enter the following:
N: 120 (10 years)
PV: −200000
PMT: 2500
FV: 0 (exhausted after 10 years)
Solve for I%.

2 Write your answer. Joe would require an interest

rate of 8.7% per annum to make
monthly withdrawals of $2500
for 10 years.
b 1 Change the payment Pmt or PMT to 3000
and solve for N.
2 Write your answer, rounding down as we are Joe will receive a regular
only counting regular payments. payment for 78 months.
c 1 Open the finance solver on your calculator
and enter the information below, as shown.
N: 120 (10 years)
PV: −200000
FV: 0 (the annuity will be exhausted after
10 years)
2 Solve for Pmt or PMT.
Note: The sign of Pmt or PMT is positive, because it
is money received.
3 Write your answer. Joe will receive $2121.31 each

month from the annuity.
d Find the final value after 120 months of 120
$2121.31. 5
−200000
2121.
31
0.
047
12
12
Since FV is 0.047. . . , the final payment is 0.05 Final payment is $2121.36

more than the regular payment.

8F 8F Using a finance solver to find interest rates, time taken and regular payments 437
Skill-
sheet Exercise 8F
Analysing an investment with additional payments using a financial calculator

Example 21 1 Armaan puts $15 000 into an investment that compounds annually.
Find the annual interest rate that would allow Armaan’s investment to reach $100 000
after 10 years if he invests an additional $4500 each year for 10 years. Round your
answer to two decimal places.
Example 22 2 Jemima puts $30 000 into an investment that compounds monthly.
a What annual interest rate would allow Jemima’s investment to double after 5 years if
she invests an additional $400 each month for 5 years? Round your answer to one
decimal place.
Assume the investment pays 3.2% per annum, compounding monthly.
b i If she wants her investment to be worth at least $40 000 in 1 year, what is the
minimum she will need to add to the investment each month? Round your
Note: You will need to round your answer up so that it reaches $40 000.
ii If she invests an additional monthly payment of at least $1000, what is the
minimum number of months that it will take for the investment to first reach
$100 000?
3 Kelven puts $7500 into an investment, paying 4.7% per annum, compounding monthly.
He makes regular additional contributions to the investment each month. After one
year, Kelven’s investment is worth $13 991.15 to the nearest cent.
a Find the amount of Kelven’s regular monthly payment. Round your answer correct
b i Find the amount that Kelven invested in the first year through monthly
payments.
ii Find the increase in the value of the investment in the first year.
iii Hence, find how much interest was earned in the first year. Round your answer
correct to the nearest cent.
c Given Kelven’s monthly payment found in a, how many months will it take for
Kelven’s investment to be worth at least $20 000?
Analysing a reducing balance loan using a financial calculator

Example 23 4 Dan arranges to make regular payments of $450 per month followed by a single
smaller final payment to repay a loan of $20 000. Interest is charged at 9.5% per
Find the number of monthly payments required to pay out the loan.

438 Chapter 8 Reducing balance loans, annuities and investments 8F
5 A building society offers $240 000 loans at an interest rate of 10.25% compounding
monthly for a 30 year period.
a If payments are $2200 per month, calculate the amount still owing on the loan after
12 years. Round your answer to the nearest cent.
b If the loan has a regular monthly payment of $2150.64 for the first 359 payments,
calculate:
i the final payment, rounding your answer to the nearest cent.
ii the total amount repaid, rounding your answer to the nearest cent.
iii the total amount of interest paid, rounding your answer to the nearest cent.
6 Rahul borrows $17 000 at an interest rate of 6.8% per annum, compounding monthly.
Rahul wishes to pay off the reducing balance loan in 30 months by making equal
payments for 29 months followed by a final payment that is as close to the regular
payment as possible.
a Find the regular monthly payment. Round your answer to the nearest cent.
b Find the final payment of the loan. Round your answer to the nearest cent.
c Find the total of the repayments of the loan.
d Find the total amount of interest that Rahul has paid.
7 Cale borrows $140 000 at an interest rate of 8.6% per annum, compounding quarterly.
Cale makes regular equal quarterly payments except for the final payment which is as
close to the regular payment as possible.
a If Cale pays off the loan in 10 years with 39 regular quarterly payments of
$5253.39:
i find the final payment. Round your answer to the nearest cent.
ii find the total amount that Cale repays.
b Rounding each of your answers to the nearest cent, if Cale pays off the loan in 15
years, find
i the regular quarterly payment.
ii the final payment.
iii the total cost of repaying the loan.
8 Lorenzo borrows $250 000 at 5.2% per annum, compounding fortnightly. Lorenzo
makes 649 equal fortnightly payments followed by a final payment which is as close to
the regular payment as possible. Find the total cost of repaying the loan. Round your
9 Joan takes out a loan of $50 000 with an interest rate of 4.9% per annum, compounding
monthly. She makes regular monthly payments for 23 months of $2191.33 followed by
a single final payment. How much interest does Joan pay in total over the duration of
the two year loan?

8F 8F Using a finance solver to find interest rates, time taken and regular payments 439
Analysing an annuity using a financial calculator

Example 24 10 Olek invests $100 000 into an annuity with interest compounding monthly.
a What is the smallest interest rate that would allow Olek to withdraw $2500 each
month for 4 years. Round your answer to two decimal places.
b Assume the interest rate is 6% and Olek wishes to be paid monthly payments for
4 years. How much will he receive each month as his regular payment? Assume that
the final payment is as close as possible to the regular payment.
c Assume the interest rate is 6% and Olek receives a regular payment of $2000. For
how many months will he receive his full payment?
11 Sophia invests $300 000 into an annuity, paying 4.3% interest per annum, compound-
ing quarterly. She wishes to receive a payment of at least $5000 every quarter.
For how many quarters will Sophia receive at least $5000?
12 Kai invests $500 000 in an annuity. The annuity earns interest at the rate of 4.7% per
annum, compounding monthly. The balance of Kai’s annuity at the end of the first year
of the investment is $474 965.28.
a What monthly payment did Kai receive? Round your answer to the nearest cent.
b How much interest would Kai’s annuity earn in the first year? Round your answer to
the nearest cent.

13 Simone invests $3000 in an account that pays interest at the rate of 3.1% per annum,
compounding monthly. She makes an additional payment of $250 each month.
The number of months that it will take the investment to reach a balance of at least
$30 000 is
A 40 B 41 C 91 D 92 E 93
14 Lachlan borrows $480 000 to buy an apartment using a reducing balance loan that
compounds monthly.
Lachlan makes regular monthly payments of $3075.72 followed by a final payment of
$3075.53. If the loan is paid out fully in 20 years, the annual interest rate is closest to
A 0.3875% B 0.0465% C 4.65% D 11.55% E 14.67%
15 Audrey invests $85 000 in an annuity, paying 6.3% per annum, compounding monthly.
Audrey receives a regular monthly payment from the annuity.
If the value of the annuity after one year is $71 983.41, the amount of interest earned in
the first year is closest to
A $1500 B $4983 C $5355 D $13 017 E $31 016

8G Solving harder financial problems

Learning intentions
I To be able to find the value of an investment when the regular payment changes.
I To be able to analyse the impact of a change in the interest rate on a reducing balance
loan, an annuity and an investment.
Sometimes the conditions of a reducing balance loan can change, requiring the regular
repayment to increase or decrease for the loan to be repaid in full. Similarly, a change in
the interest rate can also alter the payment received from an annuity or the balance of an
investment with compound interest rates. A financial solver on the CAS calculator can help
to solve for the regular payment or the new balance after a change has occurred.
Changing the regular payment to an investment

Sometimes an investor may want to change the regular additional payment to an investment.
When this happens, we need to consider the investment in two parts. First, we consider the
time before the change. Then, we consider the time after the change occurs.
Banks and other financial institutions do not round the value of an investment or loan until
the investment is withdrawn or the loan is fully repaid. Thus, for any intermediate step, we
will use unrounded values.
Example 25 Finding the value of an investment when the regular payment

changes
Derek invests $50 000 into a compound interest investment paying 6.1% per annum,
compounding annually. Derek invests an additional $8000 per year immediately after
interest is calculated.
After five years, Derek increases his additional investment to $10 000 per year.
Calculate the value of Derek’s investment after twelve years (in total).
1 Open finance solver and enter the following:
N: 5 (5 years before the change)
PV: -50000 (value of initial investment)
PMT: -8000 (additional amount added)
Pp/Y: 1 (annual payment)
Cp/Y: 1 (interest compounds annually)
Solve for FV.

8G Solving harder financial problems 441
2 Change the following in finance solver:

N: 7 (7 years after the change)
PV: -112414.364. . . (copied from FV above)
PMT: -10000 (additional amount added)
Solve for FV
3 Write your answer. The value of Derek’s

investment is
$254 343.80
Changing the interest rate

Example 26 Finding the final payment of a reducing balance loan when the
interest rate changes
Adrian borrows $150 000 for 25 years at an interest rate of 6.8% per annum, compound-
ing monthly.
For the first three years, Adrian repays $1041.11 each month.
After 3 years, the interest rate rises to 7.2% per annum. Adrian still wishes to pay off
the loan in 25 years so makes 263 monthly payments of $1076.18 followed by a final
payment.
Calculate the final payment to ensure the loan is fully repaid at the end of 25 years.
1 Open the finance solver on your calculator
and enter the information below, as shown
opposite.
N: 36 (number of monthly payments in
3 years)
PV: 150000 (initial value of loan)
Pmt: −1041.11 (monthly repayments)
Note: You can enter N as 3 × 12 (3 years of monthly
payments). The finance solver will calculate this as
36 for you.

442 Chapter 8 Reducing balance loans, annuities and investments 8G
2 Solve for FV.
3 If the loan is still to be repaid in 25 years,

there are still 22 years left.
Change:
N to 22 × 12 or 264 payments
I(%) to 7.2 (the new interest rate)
PV to 142391.83593707 (the balance after
3 years)
Pmt to −1076.18
4 Solve for FV.
5 Since FV is 0.1173. . . , the final payment Final payment =

decreases by $0.12. $1076.18 − $0.12 = $1076.06
Thus, the final payment will be
$1076.06
Note: It is important that you do not round prematurely or you will get the incorrect answer of
$1076.08
A similar analysis can be used for both annuities and investments with additions to the
principal.
Exercise 8G
Changing the regular payment

Example 25 1 Danielle invests $8000 into a compound interest investment paying 7.6% per annum,
She invests an additional $1000 per year immediately after interest is calculated.
After five years, Danielle increases her additional investment to $2000 per year.
Calculate the value of Danielle’s investment after twelve years (in total). Round your

8G 8G Solving harder financial problems 443
2 Peta invests $20 000 into a compound interest investment paying 4.8% per annum,
Peta invests an additional $200 per month immediately after interest is calculated.
After ten years, Peta increases her additional investment to $500 per month.
Caculate the value of Peta’s investment after twenty years (in total). Round your
3 Jarrod opens an account with an initial balance of $0 that pays interest at a rate of
6% per annum, compounding monthly.
He makes monthly deposits of $500 to the account for 10 years.
After 10 years of making deposits, Jarrod withdraws the balance and places it in an
annuity, also with an annual interest rate of 6%, compounding monthly. He withdraws
$500 each month from the account.
a How much does he invest in the annuity after the initial 10 years?
b How much will remain in the annuity after 10 years? Round your answer to the
nearest cent.
Reducing balance loans with changing conditions

Example 26 4 Julien borrows $35 000 for 20 years at an interest rate of 10.5% per annum,
For four years, he pays $349.43 each month.
After four years, the interest rate rises to 13.75% per annum. Julien still wishes to
pay off the loan in a total of 20 years so he makes 191 monthly payments of $418.66
followed by a final payment. For the loan to be fully repaid to the nearest cent, Julien’s
final repayment will be a smaller amount.
Calculate the final payment that Julien must make to repay the loan in 20 years. Round
5 A couple negotiates a 25-year mortgage of $500 000 at a fixed rate of 7.5% per annum
compounding monthly for the first seven years.
The monthly repayment amount of $3694.96 is paid each month for seven years.
After seven years, the interest rate rises to 8.5% per annum. The couple now pay
$3959.44 each month.
Calculate the value of the loan after a further seven years at the higher interest rate.
6 Zian borrows $750 000 for a new home at an interest rate of 8.5% per annum,
For the first five years, he only pays the interest so the value of the loan remains at
$750 000.
a Calculate Zian’s monthly repayments.

After five years the interest rate increases to 9.4%. Zian must now pay more each
month in order to pay the loan in full within the original 30 years. He does this by
making 299 regular monthly repayments followed by a final payment which is as close
to the regular payment as possible.
b Calculate the new regular monthly payment amount Zian must make.
c Find the final payment.
d Calculate the total amount that Zian pays over 30 years.
e How much interest will Zian pay over the lifetime of the loan?
Changed conditions with annuities

7 Helen has $80 000 to invest. She chooses an annuity that pays interest at the rate of
6.4% per annum, compounding monthly. Helen expects her investment to be fully
exhausted after 15 years.
She receives a monthly payment of $692.50 each month for two years.
After two years, the interest rate of Helen’s investment was reduced to 6.2% per
a If Helen continues to withdraw $692.50 each month, how many more months can
she withdraw this regular amount?
b The final payment received from this annuity will be less than the regular
repayments. Find the final payment that will exhaust the annuity.
8 Ethan invests $125 000 into an annuity from which he receives a regular monthly
payment of $850. The interest rate for the annuity is 5.4% per annum, compounding
monthly.
a Let Vn be the balance of the annuity after n monthly payments. Write a recurrence
relation written in terms of V0 , Vn+1 and Vn to model the value of this annuity from
month to month.
b After two years, the interest rate for this annuity will fall to 4.1%.
So that Ethan will continue to receive a monthly payment of $850 for the following
18 years, he will add an extra one-off amount to the annuity at this time.
Determine the minimum value of the one-off addition. Give your answer to the
nearest dollar.
9 Sameep deposits $150 000 into a savings account earning 6% per annum, com-
pounding monthly, for 10 years. He makes no withdrawals or deposits during that
time.
a Let S n be the balance of Sameep’s investment after n months. Write a recurrence
relation to model this investment.
b What is the value of this account after 10 years? Round your answer to the nearest
cent.

8G 8G Solving harder financial problems 445
After 10 years Sameep withdraws the money and invests the full amount into an
annuity. He will require this investment to provide monthly withdrawals of $2600.
c What is the minimum annual interest rate required if Sameep’s investment is to be
exhausted after 10 additional years? Round your answer to two decimal places.
Changed conditions on investments

10 Marcus’ grandparents place $2000 in an investment account that pays interest at a
rate of 4% per annum, compounding annually. For 18 years from Marcus’ birth, they
contribute an additional $1000 to the account each year.
a Find the balance of the investment after 18 years. Round your answer to the nearest
cent.
When Marcus turns 18, the interest rate increases to 5% and his grandparents stop
contributing.
b Give the balance of the account to the nearest dollar when Marcus turns 21.
11 When Jessica starts working, she sets up an investment account with an initial balance
of $1000. Each month she deposits $200 into the account.
The account has an annual interest rate of 4.9% compounding monthly.
a Find the balance of the investment after one year. Round your answer to the nearest
cent.
b How much interest has the investment earned in the first year? Round your answer
After three years, the interest rate increases to 6% per annum and Jessica will increase
her monthly deposit to $350 per month.
c Find the balance of Jessica’s investment account after two years at the higher
interest rate. Round your answer to the nearest cent.
d How many payments of $350 will Jessica need to make until her investment first
exceeds $35 000?

12 Cherry borrowed $500 000 to buy an apartment.
The interest rate for this loan was 4.31% per annum, compounding monthly.
Cherry paid $4200 per month for the first two years.
After these two years, the interest rate changed. Cherry was able to pay off the loan in a
further 8 years by paying $5361.49 each month.
The interest rate, per annum, for the final 8 years of the loan was closest to
A 3.90% B 4.00% C 4.10% D 4.30% E 4.31%

13 Thirty years ago, Irene invested a sum of money in an account earning interest at the
rate of 3.1% per annum, compounding monthly.
After 10 years, the interest rate changed.
For the next twenty years, the account earned interest at the rate of 2.7% per annum,
The balance of her account today is $876 485.10.
The sum of money that Irene originally invested is closest to
A $360 300 B $375 000 C $390 000 D $511 100 E $670 000
14 Calvin plans to retire from his work in 12 years’ time and hopes to have $800 000 in an
annuity investment at that time.
The present value of this annuity investment is $227 727.96, where the interest rate is
3.6% per annum, compounding monthly.
To make this investment grow faster, Calvin adds $2500 at the end of each month.
Two years from now, Calvin expects the interest rate to fall to 3.3% per annum,
compounding monthly, and to remain at this level until he retires.
When the interest rate changes, Calvin must change his monthly payment if he wishes
to make his retirement goal.
The value of his new monthly payment will be closest to
A $1950 B $2500 C $2560 D $2600 E $2630
8H Interest-only loans
Learning intentions
I To be able to find the regular payment amount for an interest-only loan with and
without a financial solver.
I To be able to find the amount borrowed for an interest-only loan.
I To be able to find the interest rate for an interest-only loan.
In an interest-only loan, the borrower repays only the interest that is charged. As a result,
the balance of the loan remains the same for the duration of the loan. To understand how
this happens, consider a loan of $1000 with an interest rate of 5% per annum, compounding
yearly. The interest that is charged after 1 year will be 5% of $1000, or $50. If the borrower
only repays $50, the value of the loan will still be $1000.
The recurrence relation V0 = 1000, Vn+1 = 1.05Vn − D can be used to model this loan.
The table below shows the balance of the loan over a 4-year period for three different
payment amounts: D = 40, D = 50 and D = 60.

8H Interest-only loans 447
D = 40 D = 50 D = 60
V0 = 1000, V0 = 1000, V0 = 1000,
Vn+1 = 1.05Vn − 40 Vn+1 = 1.05Vn − 50 Vn+1 = 1.05Vn − 60
V0 = 1000 V0 = 1000 V0 = 1000
V1 = 1010 V1 = 1000 V1 = 990
V2 = 1020.50 V2 = 1000 V2 = 979.50
V3 = 1031.525 V3 = 1000 V3 = 968.475
V4 = 1043.101 . . . V4 = 1000 V4 = 956.898 . . .
The amount owed keeps The amount owed stays The amount owed keeps
increasing. constant. decreasing.
We can plot these balances against the payment number.

If the periodic payments on this loan are
1050
smaller than $50 (e.g. D = 40), the amount
owed will increase over time. The balance of
the loan is shown as red dots.
Balance ($)
If the periodic payments on this loan are larger 1000

than $50 (e.g. D = 60), the amount owed will
decrease over time. The balance of the loan is
shown as blue dots.
950
If the periodic payments on this loan are
O 1 2 3 4 5
exactly $50, then the amount owed on the loan
Payment (n)
will always be $1000. The balance of the loan
is shown as green dots.
A loan where the balance stays constant is called an interest-only loan and is commonly
used for investment purposes.
Modelling interest-only loans

Let Vn be the value of the interest-only loan after n payments have been made. Then
r
where R = 1 + is the growth multiplier, r is the annual interest rate, p is the
100 × p
number of compounding periods per year and D is the regular payment per compounding
period which is equal to the interest charged, given by
r
D= × V0
100 × p

Example 27 Finding the regular payment for an interest-only loan
Jane borrows $50 000 to buy some shares. Jane negotiates an interest-only loan at an
interest rate of 9% per annum, compounding monthly. What is the monthly amount Jane
will be required to pay?
Calculation method
r
Use the rule D = × V0 . V0 = 50 000
100 × p
1 V0 is the amount borrowed = $50 000 r
D= × V0
100 × p
2 Calculate the interest payable where
9
r = 9 and p = 12. D= × 50 000
100 × 12
3 Evaluate the rule for these values and D = 375
write your answer. Jane will need to repay $375 every month
on this interest-only loan.
Finance solver method
Consider one compounding period

because all compounding periods will be
identical.
1 Open Finance Solver and enter the
following.
N: 1 (one compounding period)
PV: 50000
FV: −50000 (the amount owing will
be the same after one payment)
Cp/Y: 12 (interest compounds
monthly)
2 Solve for the unknown future value
(Pmt or PMT). On the:
TI-Nspire: Move the cursor to the
Pmt entry box and press · to
solve.
ClassPad: Tap on the PMT entry
box and tap Solve.
The amount –375 now appears in the
Pmt or PMT entry box.

8H 8H Interest-only loans 449
Example 28 Finding the amount borrowed for an interest-only loan
A loan at 6% per annum, compounding monthly, requires payments of $440 each month.
If the loan is an interest-only loan, what is the principal?
r
1 Use formula D = × V0 , where Solving for the principal:
100 × p 6
D = 440, r = 6, p = 12. 440 = × V0
100 × 12
V0 = 88 000
2 Write the answer.
3 The principal is $88 000.
Example 29 Finding the interest rate for an interest-only loan
An interest-only loan of $1 000 000 requires quarterly payments of $4000.

What is the annual interest rate on the loan?
r
1 Use formula D = × V0 , where Solving for r:
100 × p r
4000 = × 1 000 000
D = 4000, p = 4, V0 = 1 000 000. 100 × 4
r = 1.6
2 Write the answer.
3 The annual interest rate is 1.6%
Exercise 8H
Finding the regular payment for an interest-only loan

Example 27 1 Georgia borrows $100 000 to buy an investment property. If the interest on the loan is
7.2% per annum, compounding monthly, find her monthly payment on an interest-only
loan.
2 In order to invest in the stockmarket, Jamie takes out an interest-only loan of $50 000.
If the interest on the loan is 8.4% per annum compounding monthly, find his monthly
payment amount.
3 Robert takes out an interest-only loan for $220 000 at an interest rate of 5.46% per
annum, compounding fortnightly. Find the fortnightly payment.
4 Frannie borrows $180 000 at an interest rate of 4.95% per annum, compounding
quarterly. If Frannie only pays the interest, find the total payments made over a five
year period.

450 Chapter 8 Reducing balance loans, annuities and investments 8H
5 Jackson takes out an interest-only loan of $30 000 from the bank to buy a painting. He
hopes to resell it at a profit in 12 months’ time. The interest on the loan is 9.25% per
annum, compounding monthly. He makes monthly payments on the loan.
a Find the total amount that Jackson pays in 12 months.
b How much will he need to sell the painting for in order not to lose money?
6 Ric takes out an interest-only loan of $600 000 to buy an investment property. The
interest on the loan is 5.11% per year, compounding monthly.
a Calculate Ric’s monthly repayments if he only pays the interest.
b Ric sells the property after 10 years. Calculate the total interest paid on the loan.
c How much must Ric sell the property for if he wishes to make a profit of at least
$100 000?
7 Mindy borrows $35 000 for 20 years at 6.24% per annum, compounding monthly. For
the first five years, Mindy pays interest only.
a Calculate the monthly repayments that Mindy makes for the first two years.
b State the balance of the loan after five years.
c For the next 179 months, Mindy pays $300 per month followed by a smaller
payment to fully repay the loan. Find this final repayment. Round your answer to
the nearest cent.
d Find the total amount that Mindy paid for the duration of the 20 year loan.
Find the amount borrowed for an interest-free loan

Example 28 8 An interest-only loan with an interest rate of 5.3% per annum, compounding annually,
requires annual payments of $2120. What is the principal?
9 An interest-only loan with an interest rate of 6.6% per annum, compounding monthly,
requires a monthly payment of $88. What is the principal?
10 Yianni took out an interest-only loan with an interest rate of 4.2% per annum,
compounding monthly. Over a two year period, Yianni paid $2352 in total. Find the
principal of the loan.
Find the interest rate for an interest-free loan

Example 29 11 An interest-only loan of $4000 requires annual payments of $116. What is the annual
interest rate?
12 An interest-only loan of $12 000 compounds monthly and requires monthly payments
of $36. What is the annual interest rate?
13 Leo takes out an interest-only loan of $35 000 which compounds monthly and requires
monthly payments. Over a two year period, Leo pays a total of $3360. What is the
annual interest rate?

8H 8I Perpetuities 451
14 Svetlana borrows $320 000 on an interest-only loan at an interest rate of 4.92% per
annum, compounding monthly for the first 5 years. Following this, the interest rate
changes.
a Calculate the monthly repayments that Svetlana makes for the first five years.
b Calculate the total amount that Svetlana paid during the first five years.
For the next five years, Svetlana pays $86 400 on the interest-only loan.
c State the total interest that she paid during the second half of the loan.
d Calculate the monthly repayment that she made during the second half of the loan.
e Hence, find the annual interest rate during the second half of the loan.

15 Matthew would like to purchase a new home. He establishes a 20 year loan for
$310 000 with interest charged at the rate of 3.84% per annum, compounding monthly.
Each month, Matthew will only pay the interest charged for that month.
After three years, the amount that Matthew will owe is
A $274 288 B $277 222 C $310 000 D $345 712 E $418 350
16 Eve borrowed $780 000 to buy a house.

The interest rate for this loan was 4.82% per annum, compounding monthly.
A scheduled monthly repayment that allowed Eve to fully repay the loan in 20 years
was determined.
Eve decided to pay interest-only for the first two years.
After these two years, the interest rate changed. Eve was still able to pay off the loan in
the 20 years by paying the original monthly repayment amount each month.
The new interest rate of the loan was closest to
A 3.9% B 4.0% C 4.1% D 4.2% E 4.3%
17 Jason takes out an interest-only loan for five years.

The value of Jason’s interest-only loan, Vn , after n months, can be modelled by the
recurrence relation
V0 = 56 000, Vn+1 = 1.0034Vn − D
The total interest paid on the loan over the five years is
A $4.08 B $190.40 C $952 D $11 424 E $56 000
8I Perpetuities
Learning intentions
I To be able to calculate the regular payment from a perpetuity.
I To be able to calculate the investment required to establish a perpetuity.
I To be able to calculate the interest rate of a perpetuity.

Recall that an annuity involves money being deposited in an investment and then withdrawn
over time in the form of regular payments. In our earlier analysis, we considered the case
where the withdrawals were made to exhaust the annuity over a given time frame. That is,
the value of the annuity eventually reached zero.
If the regular payments are smaller than the interest received, the annuity will continue
to grow. If the payments received are exactly the same as the interest earned in each
compounding period, the annuity will maintain its value indefinitely. This type of annuity
is called a perpetuity and the payments that are equal to the interest earned can be
made forever (or in Perpetuity). Perpetuities have the same relationship to annuities as
interest-only loans have to reducing balance loans.
Modelling perpetuities
Let Vn be the value of the perpetuity after n payments have been made. Then
r
where R = 1 + is the growth multiplier, r is the annual interest rate, p is the
100 × p
number of compounding periods per year and D is the regular payment per compounding
period which is equal to the interest earned, given by
r
D= × V0
100 × p
Example 30 Calculating the payment from a perpetuity
Elizabeth invests her superannuation payout of $500 000 into a perpetuity that will
provide a monthly income.
If the interest rate for the perpetuity is 6% per annum, what monthly payment will
Elizabeth receive?
r
1 Find the monthly interest D= × V0
100 × p
earned.
6
= × 500000
100 × 12
= 2500
2 Write your answer, rounding Elizabeth will receive $2500 every month from her
as required. investment.
Example 31 Calculating the investment required to establish a perpetuity
Calculate how much money will need to be invested in a perpetuity account, earning
interest of 4.8% per annum compounding monthly, if $300 will be withdrawn every
month.

8I Perpetuities 453
r
1 Use the rule D = × V0 to write
100 × p
4.8
down an equation that can be solved 300 = × V0
100 × 12
for V0 .
300
V0 =
0.004
= 75000
2 Write your answer. $75 000 will need to be invested to

establish the perpetuity investment.
Problems involving perpetuities can also be solved using a financial calculator.
Example 32 Calculating the interest rate of a perpetuity
A university mathematics faculty has $30 000 to invest. It intends to award an annual
mathematics prize of $1500 with the interest earned from investing this money in a
perpetuity.
What is the minimum interest rate that will allow this prize to be awarded indefinitely?
We will consider just one compounding
period because all compounding periods
will be identical.
Calculation method
r r
1 Use the rule D = × V0 and 1500 = × 30 000
100 × p 100 × 1
solve the equation for r.
1500 = r × 300
1500
r= =5
300
2 Write your answer. The minimum annual interest rate to award
this prize indefinitely is 5%.
Financial solver
1 Open Finance Solver and enter the

following.
N: 1 (one payment)
PV: −30 000
Pmt or PMT: 1500 (prize is $1500
each year)

454 Chapter 8 Reducing balance loans, annuities and investments 8I
FV: 30 000 (the balance will be the

same after each payment)
Pp/Y: 1 (yearly payment)
Cp/Y: 1 (interest compounds yearly)
2 Solve for the unknown interest rate
(I%). On the:
TI-Nspire: Move the cursor to the
I% entry box and press · to
solve.
ClassPad: Tap on the I% entry box
and tap Solve The amount 5 now
appears in the I% entry box.
3 Write your answer, rounding as The minimum annual interest rate to award
required. this prize indefinitely is 5%.
Exercise 8I
Calculating the payment from a perpetuity

Example 30 1 Suzie invests her inheritance of $642 000 in a perpetuity that pays 6.1% per annum
a What quarterly payment does she receive?
b After five quarterly payments, how much money remains invested in the perpetuity?
c After 10 quarterly payments, how much money remains invested in the perpetuity?
2 Craig wins $1 000 000 in a lottery and decides to place it in a perpetuity that pays
5.76% per annum interest, compounding monthly.
a What monthly payment does he receive?
b How much interest does he earn in the first year?
3 Donna sold her cafe business for $720 000 and invested this amount in a perpetuity.
The perpetuity earns interest at a rate of 3.6% per annum. Interest is calculated and
paid monthly.
a What monthly payment will Donna receive from this investment?
b After three years, the interest rate for the perpetuity increases. Describe whether
Donna’s monthly payment will increase, decrease or stay the same.

8I 8I Perpetuities 455
Calculating the investment required to establish a perpetuity

Example 31 4 Geoff wishes to set up a fund so that every year $2500 is donated to the RSPCA in his
name.
If the interest on his initial investment averages 2.5% per annum, compounding
annually, how much should he invest?
5 Barbara wishes to start a scholarship that will reward the top mathematics student each
quarter with a $600 prize.
If the interest on the initial investment averages 4.8% per annum, compounding
quarterly, how much should be invested?
6 Omar inherits $920 000 and splits the money between a perpetuity and an annuity
investment.
The perpetuity pays $2340 each month based on an interest rate of 5.2% per annum
that compounds monthly.
The annuity investment has an interest rate of 4.8% per annum that compounds
monthly.
a Calculate how much Omar invested in the perpetuity.
b State how much is initially invested in the annuity investment.
c Omar realises that he only needs $2000 from his perpetuity each month and so he
adds $340 as an aditional payment into the annuity investment each month. Find the
value of the annuity investment after three years.
Calculating the interest rate of a perpetuity

Example 32 7 If Sandra has $80 000 to invest, what is the minimum interest rate she requires
to provide an annual donation of $2400 indefinitely into the future if interest is
compounding annually?
8 Benjamin has $12 000 to invest in a perpetuity to provide a prize of $750 each
year. What is the minimum interest rate that he requires in order to pay the prize in
perpetuity if interest compounds annually?
9 On retiring from work, Tyson received a superannuation payout of $694 400.

If Tyson invests the money in a perpetuity, he would then receive $3645.60 each
month for the rest of his life. At what annual percentage rate is interest earned by this
perpetuity?
Analysis of perpetuities
10 Marco invests $350 000 in a perpetuity from which he will receive a regular monthly
payment of $1487.50.
The perpetuity earns interest at the rate of 5.1% per annum.
a Determine the total amount, in dollars, that Marco will receive after one year of
monthly payments.
456 Chapter 8 Reducing balance loans, annuities and investments 8I
b Write down the value of the perpetuity after Marco has received one year of
monthly payments.
c Let Mn be the value of Marco’s perpetuity after n months. Write down a recurrence
relation in terms of M0 , Mn+1 and Mn , that would model the value of this perpetuity
over time.
11 Zihan invests $200 000 in a perpetuity from which he will receive a regular payment.
There are two options available:
Option A: Earn interest at a rate of 3.6% per annum, compounding monthly with a
regular monthly payment.
Option B: Earn interest at a rate of 3.8% per annum, compounding annually with a
regular annual payment.
a Calculate the monthly payment from Option A.
b Calculate the annual payment from Option B.
c Determine which option pays the most over one year.
d Let Zn be the value of the perpetuity after n payments that pays the most over the
course of a year. Write down a recurrence relation in terms of Z0 , Zn+1 and Zn , that
would model the value of this perpetuity over time.

12 Which of the following recurrence relations could be used to model the value of a
perpetuity investment, Pn , after n months?
A P0 = 100 000, Pn+1 = 1.005Pn + 500
B P0 = 100 000, Pn+1 = 1.005Pn − 500
C P0 = 100 000, Pn+1 = 0.005Pn − 500
D P0 = 200 000, Pn+1 = 1.003Pn + 600
E P0 = 200 000, Pn+1 = 1.103Pn − 600
13 Aaliyah invests $120 000 in a perpetuity from which she will receive a regular
monthly payment. The perpetuity has a compound interest rate of 5.2% per annum and
compounds monthly. The amount that Aaliyah will receive from the perpetuity in the
first two years is closest to
A $520 B $624 C $6240 D $12 480 E $149 760

Review
Reducing balance A reducing balance loan is a loan that attracts compound interest but is
Assign-
loan reduced in value by making regular payments.
ment
Each payment partly pays the interest that has been added and partly
reduces the value of the loan.
Annuity An annuity is an investment that earns compound interest and from

which regular payments are made.
Amortisation An amortising loan is one that is paid back with periodic payments. An
amortising investment is one that is exhausted by regular withdrawals.
Amortisation of reducing balance loans tracks the distribution of each
periodic payment, in terms of the interest paid and the reduction in the
value of the loan.
Amortisation of an annuity tracks the source of each withdrawal,
in terms of the interest earned and the reduction in the value of the
investment.
Amortisation An amortisation table shows the amortisation (payment) of all or part

table of a reducing balance loan or annuity. It has columns for the payment
number, the payment amount, the interest paid or earned, the principal
reduction or increase and the balance after the payment has been made.
Finance Solver Finance Solver is a function on a CAS calculator that performs

financial calculations. It can be used to determine any of the principal,
interest rate, periodic payment, future value or number of payments
given all of the other values.
Interest-only An interest-only loan is a loan where the regular payments made are
loan equal in value to the interest charged. Interest-only loans have the same
value after each payment is made.
Perpetuity A perpetuity is an annuity where the regular payments or withdrawals

are the same as the interest earned. The value of a perpetuity remains
constant.

Review 458 Chapter 8 Reducing balance loans, annuities and investments
Skills checklist
list
8A 1 I can generate a sequence from a recurrence relation that combines both

geometric and linear growth or decay.
8A 2 I can model compound interest investments with additions to the principal.
See Example 2 and 3, and Exercise 8A Question 2 and 4
8A 3 I can use a recurrence relation to analyse compound interest investments with

additions to the principal.
8A 4 I can determine the annual interest rate from a recurrence relation.
8B 5 I can model a reducing balance loan with a recurrence relation.
See Example 6 and 7, and Exercise 8B Question 1 and 2
8B 6 I can use a recurrence relation to analyse a reducing balance loan.
8B 7 I can model an annuity with a recurrence relation.
8C 8 I can apply the amortisation process.
8C 9 I can construct an amortisation table for a reducing balance loan.
8C 10 I can analyse an amortisation table for a reducing balance loan.
8C 11 I can analyse an amortisation table for an annuity to find the interest rate.

Review
8C 12 I can interpret and construct an amortisation table for a compound interest
investment with additional payments.
8D 13 I can find the final payment for a reducing balance loan or annuity.
8D 14 I can find the total payment made and the total interest paid.
8D 15 I can plot from an amortisation table.
8E 16 I can determine the value of an investment with regular additions made to the
principal using a financial solver.
8E 17 I can determine the balance and final payment of a reducing balance loan after
a given number of payments.
8E 18 I can determine the balance of an annuity using a finance solver.
8F 19 I can find the interest rate for an investment with additional payments.
8F 20 I can find the regular monthly payment and the time taken for an investment
with additions to the principal.
8F 21 I can determine the payment amount, total cost and total amount of interest
paid for a reducing balance loan.
8F 22 I can find the interest rate, time taken and regular payment for an annuity.
8G 23 I can find the value of an investment when the regular payment changes.
8G 24 I can analyse the impact of a change in the interest rate on a reducing balance
loan, an annuity and an investment.
See Example 26, and Exercise 8G Question 4, 8 and 10

8H 25 I can find the repayment amount for an interest-only loan with and without
finance solver.
8H 26 I can find the amount borrowed for an interest-only loan.
8H 27 I can find the interest rate for an interest-only loan.
8I 28 I can calculate the payment from a perpetuity.
8I 29 I can calculate the investment required to establish a perpetuity.
8I 30 I can calculate the interest rate of a perpetuity.
1 An investment of $18 000, earning compound interest at the rate of 6.8% per annum,
compounding yearly, and with regular additions of $2500 every year can be modelled
with a recurrence relation. If Vn is the value of the investment after n years, the
recurrence relation is
A V0 = 18000, Vn+1 = 1.006Vn − 2500 B V0 = 2500, Vn+1 = 1.068Vn − 18 000
C V0 = 18 000, Vn+1 = 1.068Vn + 2500 D V0 = 18000, Vn+1 = 1.068Vn − 2500
E V0 = 2500, Vn+1 = 1.006Vn − 18000
2 Let Vn be the value of an investment after n months. The investment is modelled by the
recurrence relation V0 = 25 000, Vn+1 = 1.007Vn − 400. The annual interest rate for
this investment is
A 0.084% B 0.7% C 2.8% D 8.4% E 36.4%
3 The value of an annuity investment, in dollars, after n months, Vn , can be modelled by

the recurrence relation shown below
V0 = 20 000, Vn+1 = 1.045Vn + 500
The increase in the value of this investment in the second month is closest to
A $500
B $900
C $1100
D $1500
ISBN 978-1-009-11041-9
E $2300 © Peter Jones et al 2023 Cambridge University Press
Review
Questions 4 and 5 relate to the following information.
A loan of $28 000 is charged interest at the rate of 6.4% per annum, compounding monthly.
It is repaid with regular monthly payments of $1200.
4 Correct to the nearest cent, the value of the loan after 5 months is
A $21 611.35 B $22 690.33 C $23 763.59 D $24 831.16 E $31 363.91
5 Following 24 regular payments of $1200, a final payment is made to fully repay the
loan. The final payment on the loan, correct to the nearest cent, will be
A $1125.41 B $1131.41 C $1175.20 D $1181.47 E $1200
6 A loan of $6000 is to be repaid in full by 11 quarterly payments followed by a final

payment that is as close to the regular payment as possible. Interest at 10% per annum
is calculated on the remaining balance each quarter. The regular quarterly payment that
is required to pay out the loan is closest to
A $527.50 B $573.10 C $584.92 D $600 E $630.64
7 Paula borrows $12 000 from a bank, to be repaid over 5 years. Interest of 12% per
annum is charged monthly on the amount of money owed. If Paula makes regular
monthly payments of $266.90, then the amount she owes at the end of the second year
is closest to
A $2880 B $5590 C $6410 D $8040 E $9120
8 Ayush invests $12 000 in an annuity from which he receives a regular monthly payment
of $239. The annuity earns interest of 7.2% per annum, compounding monthly.
The balance of the annuity after three months is closest to
A $11 495 B $11 496 C $11 665 D $12 938 E $12 939
9 James invests $50 000 in an annuity from which he receives a regular monthly payment
of $925.30.
The balance of the annuity, in dollars, after n months, Jn , can be modelled by a
J0 = 50 000, Jn+1 = 1.0035Jn − 925.30
The balance of the annuity after six months is closest to
A $45 289 B $45 458 C $45 459 D $56 659 E $56 660

Questions 10–13 refer to the following amortisation table for a reducing balance loan.

0 0.00 0.00 0.00 40000.00
1 400.00 160.00 240.00 39760.00
2 400.00 159.04 240.96 39519.04
3 400.00 158.08 241.92 39277.12
10 The principal of this loan is

A $20 000 B $30 000 C $40 000 D $50 000 E $60 000
11 The periodic payment amount on this loan is

A $80 B $160 C $240 D $400 E $560
12 The principal reduction of the loan from the third payment is

A $158.08 B $159.04 C $240 D $240.96 E $241.92
13 Assuming that payments are made monthly and interest compounds monthly, the
annual interest rate on the loan is
A 0.4% B 0.48% C 4% D 4.8% E 16%
14 Amir borrows $1500 in a reducing balance loan at a rate of 3.6% per annum,
She makes regular repayments of $383.45 each quarter for three quarters followed by a
final payment.
To pay out her loan fully, her final payment is
A $0.01 B $0.10 C $383.35 D $383.45 E $383.55
15 Monthly withdrawals of $220 are made from an account that has an opening balance of
$35 300, invested at 7% per annum, compounding monthly. The balance of the account
after 1 year is closest to
A $32 660 B $33 500 C $35 125 D $35 211 E $40 578
16 Tilly invests $5000 in an account that pays interest at the rate of 3.9% per annum,
She makes an additional payment of $1200 each year.
The number of years that it will take the investment to first reach a balance of $20 000
is
A 1 B 2 C 9 D 10 E 13

Review
17 Twenty years ago, Oscar invested $65 000 in an account earning interest at the rate of
2.8% per annum, compounding monthly.
After 10 years, he made a one-off payment of $20 000 to the account.
For the next 10 years, the account earned interest at the rate of 3.2% per annum,
compounding monthly. The balance of the account today is closest to
A $65 000 B $85 975.39 C $105 975.39 D $113 719.51 E $145 879.51
18 The monthly payment on an interest-only loan of $175 000, at an interest rate of 5.9%
per annum, compounding monthly, is closest to
A $198 B $397 C $860 D $1117 E $2581
19 A scholarship will be set up to provide an annual prize of $400 to the best Mathematics
student in a school. The scholarship is paid for by investing an amount of money into a
perpetuity, paying interest of 3.4% per annum, compounding annually. The amount that
needs to be invested to provide this scholarship is closest to
A $400 B $800 C $1176 D $11 764 E $11 765
20 Pham invests $74 000 from which he will receive a regular monthly payment.
The perpetuity has a compound interest rate of 4.8% per annum and compounds
monthly. The amount that Pham will receive from the perpetuity in the first two years
is closest to
A $296 B $3233 C $3552 D $7104 E $77 756.64

1 Josie is considering borrowing $250 000 to buy a house. A home loan at her bank will
charge interest at the rate of 4.8% per annum, compounding monthly. Josie will make
monthly payments of $1800.
Let Vn be the value of Josie’s loan after n months.
a Write down a recurrence model for the value of Josie’s loan after n months.
b After 12 months, how much would Josie owe on this loan? Round your answer to
the nearest cent.
c After how many months is the loan first below $200,000?
d If Josie chose an interest-only loan for the first year:
i what would her monthly payments be?
ii how much interest in total would she pay in the first year?
iii how much would she owe after the 12th payment?
2 Samantha inherited $150 000 from her aunt. She decides to invest this money into an
account paying 6.25% per annum interest, compounding monthly.
a If Samantha deposited her money into a perpetuity, what monthly payment would
she receive?
b If Samantha deposited her money into an annuity and withdrew $1000 per month,
how much would she have in the account after 1 year?
c If Samantha deposited her money into an annuity and withdrew $2000 per month,
after how many months is the investment first below $100,000?
d If Samantha deposited her money into an annuity and withdrew $4000 per month for
each month until the final month:
i How many regular payments of $4000 would she receive?
ii What would be the value of her last withdrawal?
3 A loan of $10 000 is to be repaid over 5 years with 19 equal quarterly payments of
$656.72 followed by a final payment. Interest is charged at the rate of 11% per annum
Find:
a the final payment. Round your answer to the nearest cent.
b the sum of all repayments, to the nearest dollar.
c the total amount of interest paid, to the nearest dollar.
4 The Andersons were offered a $24 800 loan to pay for a new car. Their loan is to be
repaid in equal monthly payments of $750, except for the last month when less than
this will be required to fully pay out the loan. The interest rate is 10.8% per annum,
a Find the number of months needed to repay this loan.
b Calculate the amount of the final payment. Round your answer to the nearest cent.
c Calculate the total interest that is paid on the loan.
5 Elsa borrowed $100 000 at 9.6% per annum, compounding quarterly. The loan was to
be repaid over 25 years with 99 equal quarterly payments followed by a final payment
that is as close to the regular payment as possible.
a How much of the first quarterly payment went towards paying off the principal?
b Elsa inherits some money and decides to terminate the loan after 10 years by paying
what is owing in a lump sum. How much will this lump sum be?
6 Helene won $750 000 in a lottery. She decides to place the money in an investment
account that pays 4.5% per annum interest, compounding monthly.
a How much will Helene have in the investment account after 10 years?
b After 10 years, Helene withdraws the money from the investment account and
places it in an annuity. The annuity pays 3.5% per annum, compounding monthly.
Helene receives $6000 per month from the annuity. For how many months will she
receive $6000?
c Helene’s accountant suggests that rather than purchase an annuity she places
the money in a perpetuity so that she will be able to leave some money to her
grandchildren. If she places $1,100,000 into a perpetuity that pays 3.6% per annum
compounding monthly, how much is the monthly payment that Helene will receive?

Chapter
9
Revision
Revision:
Recursion and financial
modelling

Revision 466 Chapter 9 Revision: Recursion and financial modelling
9A Exam 1 style questions

1 A sequence of numbers is generated by the recurrence relation shown below.
A0 = 5, An+1 = 2An − 1
What is the value of A3 ?
A 2 B 5 C 11 D 17 E 33
2 Consider the recurrence relation below.

T 0 = 6000, T n+1 = T n + 270
The recurrence relation could be used to model a
A simple interest investment of $6000 with an annual interest rate of 0.045%.
B simple interest investment of $6000 with an annual interest rate of 0.45%.
C simple interest investment of $6000 with an annual interest rate of 4.5%.
D compound interest investment of $6000 with an annual interest rate of 0.45%.
E compound interest investment of $6000 with an annual interest rate of 4.5%.
3 The value of an investment, in dollars, after n months, Vn , can be modelled by the

recurrence relation shown below.
V0 = 12 000, Vn+1 = RVn
This investment earns compound interest at the rate of 7.2% per annum, compounding
monthly.
What is the value of R in this recurrence relation?
A 0.006 B 0.072 C 1.006 D 1.072 E 1.72
4 The value of a commercial fridge, purchased for $9000, is depreciated by 10% per
annum using a reducing balance method.
Recursive calculations can determine the value of the fridge after n years, Vn .
Which one of the following recursive calculations is not correct?
A V0 = 9000
B V1 = 0.9 × 9000
C V2 = 0.9 × 8100
D V3 = 0.9 × 7290
E V4 = 0.9 × 6560
5 Lorraine has taken out a personal loan of $14 000.

Interest for this loan compounds quarterly and Lorraine makes no repayments.
After one year she owes $14 686.98.
The effective annual rate of interest for the first year of Lorraine’s loan is closest to
A 4.7% B 4.8% C 4.9% D 5.0% E 5.1%

9A Exam 1 style questions 467
Revision
6 Luke has purchased a caravan for $75 000.
He depreciates the value of the caravan using the reducing balance method.
For the first two years of reducing balance depreciation, the annual depreciation rate
was 12%.
Luke then changed the annual depreciation to d per cent.
After three more years of reducing balance depreciation, the value of the caravan was
$41638.56.
The changed depreciation rate, d per cent, is closest to
A 9% B 9.5% C 10% D 10.5% E 11.5%
7 Tran invests $375 000 in an account that pays interest at the rate of 3.6% per annum,
He makes additional payments of $400 each month into this account.
The value of Tran’s account, in dollars, after n months, T n , can be modelled by the
recurrence relation shown below.
T n = 375 000, T n+1 = 1.003T n + 400
The balance of Tran’s account first exceeds $650 000 at the end of month
A 15 B 16 C 143 D 144 E 145
8 Carlos has borrowed $54 000 to go on a holiday.

Interest on this loan is charged at the rate of 7.9% per annum, compounding monthly.
Carlos intends to repay the loan over five years with 60 monthly repayments.
After 59 equal repayments of $1092, he finds that a small adjustment to the final
repayment is required to fully repay the loan to the nearest cent.
Compared to the 59 earlier repayments, the final repayment will be
A $17.81 lower
B $25.12 lower
C $17.81 higher
D $25.12 higher
E $1109.81 higher
9 Consider the recurrence relation shown below

V0 = 175 000, Vn+1 = 1.002Vn − 350
This recurrence relation could be used to determine the value of
A a perpetuity with an interest rate of 2.4% per annum, compounding monthly.
B an annuity with an interest rate of 2% per annum, compounding annually.
C an annuity investment with an interest rate of 2% per annum, compounding
monthly.
D an item depreciating at a flat rate of 2.4% per annum of the purchase price.
E a compound interest investment earning interest at the rate of 0.24% per annum, of
the purchase price.
10 Witter has a reducing balance loan.

Six lines of the amortisation table for Witter’s loan are shown below.
13 402 148.19
14 9500.00 1742.64 7757.36 394 390.83
15 9500.00 1709.03 7790.97 386 599.86
16 9500.00 1675.27 7824.73 378 775.13
17 9500.00 1609.79 7890.21 370 844.92
18 9500.00 1576.26 7923.74 362 961.18
The interest rate for Witter’s loan changed after one of these repayments had been
made.
The first repayment with the lower interest rate was repayment number
A 14 B 15 C 16 D 17 E 18
11 Dina invested $225 000 in an annuity.

This investment earned interest at the rate of 5.2% per annum, compounding quarterly.
Immediately after the interest has been added to the account each quarter, Dina
withdraws a payment of $3800.
A recurrence relation that can be used to determine the value of Dina’s investment after
n quarters, Dn , is
A D0 = 225 000, Dn+1 = 0.948Dn − 3800
B D0 = 225 000, Dn+1 = 0.987Dn − 3800
C D0 = 225 000, Dn+1 = 1.0052Dn − 3800
D D0 = 225 000, Dn+1 = 1.013Dn − 3800
E D0 = 225 000, Dn+1 = 1.052Dn − 3800
12 Alicia invested some money in a perpetuity from which she receives a payment of
$675.10 each quarter.
The perpetuity pays interest of 4.3% per annum, compounding quarterly.
How much money did Alicia invest in the perpetuity?
A $6280 B $15 700 C $18 840 D $62 800 E $188 400
13 Geoff has a current balance of $317 922.75 in his superannuation account.

Geoff’s employer deposits $420 into this account every fortnight.
This account earns interest at the rate of 3.1% per annum, compounding fortnightly.
The account will first exceed $720 000 after
A 26 years B 235 fortnights
C 236 fortnights D 394 fortnights
E 395 fortnights

9B Exam 2 style questions 469
Revision
14 Lucy borrows $180 000 in an interest-only loan.
Interest is calculated and paid monthly.
If Lucy pays $1080 per month, then the annual interest rate is
A 0.06% B 0.6% C 0.72% D 6% E 7.2%
15 Maya borrowed $32 000 to buy a car and was charged interest at the rate of 10.4% per
For the first year of the loan, Maya made monthly repayments of $380.
For the second year of the loan, Maya made monthly repayments of $510.
The total amount of interest that Maya paid over this two-year period was closest to
A $3050 B $3268 C $4362 D $6318 E $10 680
9B Exam 2 style questions

1 Abigail invests some money into a savings account that earns interest compounding
quarterly.
The interest is calculated and paid at the end of each quarter.
Let An be the balance of Abigail’s savings account, in dollars, after n quarters.
The recursive calculations below show the balance of Abigail’s savings account for the
first two quarters.
A0 = 8500
A1 = 1.013 × 8500 = 8610.50
A2 = 1.013 × 8610.50 = 8722.44
a How much money did Abigail initially invest?
b How much interest in total did Abigail earn by the end of the second quarter?
c Write down a recurrence relation, in terms of A0 , An+1 and An that can be used to
model the balance, in dollars, of Abigail’s savings account after n quarters.
d What is the annual interest rate?
e Abigail plans to use her savings to buy a car that will cost $10 000. At the end of
which quarter will Abigail first have enough money to buy the car?

2 Tina spent $25 000 on a new coffee machine for her cafe.
The value of the coffee machine will depreciate by $0.50 per hour of use.
The recurrence relation below can be used to model the value of the coffee machine,
Vn , after n years.
V0 = 25 000, Vn+1 = Vn − 936
a Use recursion to show that the value of the coffee machine after three years is
$22 192.
b Tina uses the coffee machine all 52 weeks of the year for the same number of hours
each week.
For how many hours each week is the coffee machine used?
3 Yash has a reducing balance loan where he makes monthly payments.

Five lines of the amortisation table for Yash’s loan are shown below.
0 0 0.00 0.00 260 000.00
1 2150.00 1170.00 980.00 259 020.00
2 2150.00 1165.59 984.41 258 035.59
3 2150.00 1161.16 988.84 257 046.75
4 2150.00 A B C
a What amount did Yash originally borrow?

Interest is calculated monthly and Yash makes monthly payments.
b Show that the interest rate for this loan is 5.4% per annum.
c Write down the values of A, B and C.
4 Millie takes out a reducing balance loan of $240 000. The interest rate for the loan is
3.6% per annum, compounding fortnightly.
Millie pays $1350 per fortnight and decides to do so for all payments except the final
payment which will be lower.
a How many of Millie’s payments will be exactly $1350?
b After seven years of repayments, Millie decides to pay the remaining balance of her
loan.
How much will Millie need to pay?
c In which of the years will Millie pay the most interest?

Revision
5 Hugh bought a motorscooter to deliver food for a local restaurant.
He paid $4450 for the scooter.
The value of Hugh’s scooter depreciated by a fixed amount for each evening shift that
he completed.
After 20 evening shifts, the value of the scooter decreased by $50.
a What was the value of Hugh’s scooter after 20 evening shifts?
b Write a calculation that shows that the value of Hugh’s scooter depreciated by $2.50
per evening shift.
c The value of Hugh’s scooter after n evening shifts can be determined using a rule.
Complete the rule below by writing the appropriate numbers in the boxes provided.
Hn = − ×n
d Using the rule, find the value of Hugh’s scooter after 200 evening shifts.
e The value of the scooter continues to depreciate by $2.50 per evening delivery shift.
After how many shifts will the value of Hugh’s scooter first fall below $3000?
6 A government grant of $65 000 was invested in an annuity to provide a monthly

payment to a local artist.
The annuity pays interest at the rate of 3.54% per annum, compounding monthly.
After four months the annuity had a balance of $43592.50 remaining.
a What is the value, in dollars, of the monthly payment to the artist?
b After four months, the artist realises that she will not complete the artwork within
the originally set twelve months.
To extend the time that the annuity will last, the artist will receive a reduced
payment of $2000 per month.
The annuity will end with one final monthly payment that will be smaller than
$2000.
Calculate the number of months that the artist will receive $2000 per month.

Chapter
10
Matrices
Chapter questions
I What is a matrix?
I How is the order of a matrix defined?
I How are the positions of the elements of a matrix specified?
I How do we use matrices to represent information and solve practical
problems?
I How do we use matrices to represent a network diagram?
I What are the rules for adding and subtracting matrices?
I How do we multiply a matrix by a scalar?
I What is the method for multiplying a matrix by another matrix?
I How do we form and use permutation, communication and dominance
matrices?
I How can your CAS calculator be used to do matrix operations?
Matrix algebra was first studied in England in the middle of the nineteenth century.
Matrices are now used in many areas of science and business: for example, in physics,
medical research, encryption and internet search engines.
In this chapter we will show how addition and multiplication of matrices can be defined
and how matrices can be used to describe the relationship between people, businesses
and sporting teams.
In Chapter 11 we will see how matrices can be used to represent networks.

10A What is a matrix? 473
10A What is a matrix?

Learning intentions
I To be able to state the order of a given matrix.
I To be able to describe the location of an element in a matrix.
I To be able to determine the transpose of a matrix.
I To be able to define and recognise diagonal, symmetric and triangular matrices.
I To be able to define and recognise identity matrices.
A matrix (plural matrices) is a rectangular array or table of numbers or symbols, arranged

in rows and columns. We form a matrix from data in the following way.
The table of data shown below displays the heights, weights, ages and pulse rates of eight
students.
 
Name Height Weight Age Pulse rate 173 57 18 86
 
179 58 19 82
Mahdi 173 57 18 86  
 
Dave 179 58 19 82 167 62 18 96
 
Jodie 167 62 18 96 195 84 18 71
D =  
Simon 195 84 18 71 173 64 18 90
 
184 74 22 78
 
Kate 173 64 18 90
 
Pete 184 74 22 78 175 60 19 88
 
Mai 175 60 19 88 140 50 34 70
Tran 140 50 34 70
If we extract the numbers from the table and enclose them in square brackets, we form a
matrix. We might call this matrix D (for data matrix). We use capital letters A, B, C, etc. to
name matrices.
Rows and columns

Rows and columns are the building blocks of matrices. Col. 3
We number rows from the top down: row 1, row 2, etc. 173 57 18 86
Columns are numbered from the left across: column 1,
Row 2 179 58 19 82
column 2, etc.
167 62 18 96
195 84 18 71
Order of a matrix D=
173 64 18 90
In its simplest form, a matrix is just a rectangular array
(rows and columns) of numbers. 184 74 22 78
175 60 19 88
The order (or size) of matrix D is said to be 8 × 4, read
140 50 34 70
‘8 by 4’ because it has eight rows and four columns.

474 Chapter 10 Matrices
Order of a matrix
Order of a matrix = number of rows × number of columns
The numbers, or entries, in the matrix are called elements.

The number of elements in a matrix is determined by its order. For example, the matrix D
has order 8 × 4 and the the number of elements in matrix D is 32 (8 × 4 = 32).
Row matrices
Matrices come in many shapes and sizes. For example, from this same set of data, we could
have formed the matrix called K

K = 173 64 18 90
This matrix has been formed from just one row of the data: the data values for Kate.
Because it only contains one row of numbers, it is called a row matrix (or row vector). It is
a 1 × 4 matrix: one row by four columns. It contains 1 × 4 = 4 elements.
Column matrices
Equally, we could form a matrix called H (for height 
 173 

matrix). This matrix is formed from just one column of 
 179 

the data, the heights of the students. 
 167 

 
Because it only contains one column of numbers, it is  
 195 
called a column matrix (or column vector). This is H =   
 173 
an 8 × 1 matrix: eight rows by one column. It contains  
184
 
8 × 1 = 8 elements. 



 175 
 
140
Example 1
State the order of each of the following matrices.
   
  1 1 4 4 5
1 5    
  4 4 3 4 6
a 3 0 b 1 5 8 9 0 c   d 
   
  4 4 3 2 1
7 6    
9 9 1 0 7
  
Solution
a 3 rows and 2 columns. b 1 row and 5 columns c 4 rows and 1 column.

Order is 3 × 2 Order is 1 × 5 Order is 4 × 1
d 4 rows and 4 columns.
Order is 4 × 4

Switching rows and columns: the transpose of a matrix

If you switch the rows and columns in a matrix you have what is called the transpose of the
matrix.
 
1 2  
  1 3 5
For example, the transpose of the matrix 3 4 is   .
  2 4 6
5 6
The transpose of a row matrix is a column matrix and vice versa.
 
1
 
For example, the transpose of the matrix 2 is 1 2 3 .
 
3
The symbol we use to indicate the transpose of a matrix is T .
 T  
1 2   1
  1 3 5 T  
Thus, 3 4 =   and 1 2 3 = 2.

 
2 4 6
  
 
5 6 3
Note: The transpose of a 3 × 2 matrix is a 2 × 3 matrix because the rows and columns are switched.
Example 2 The transpose of a matrix

 
7 4
a Write down the transpose of  .
8 1

 T
0 1 2
b Write down the matrix   .
3 2 5


c If A = 2 0 1 , write down the matrix AT .
a The transpose of the matrix is obtained by
 
7 8
switching (interchanging) its rows and columns.
 
4 1
 
 
b The symbol T is an instruction to transpose the  T 0 3
0 1 2  
matrix.   = 1 2

3 2 5
  

2 5

 
2
T  
c The symbol T is an instruction to transpose AT = 2 0 1 = 0
 
matrix A. 1

Square matrices
As a final example, we could form a matrix we call M (for
 
173 57 18 86
males). This matrix contains only the data for the males. As
 
179 58 19 82
this matrix has four rows and four columns, it is a 4 × 4 matrix. M = 
 
195 84 18 71
It contains 4 × 4 = 16 elements.  
184 74 22 78

A matrix with an equal number of rows and columns is called a
square matrix.
Example 3 Matrix facts
For each of the matrices below, write down its type, order and the number of elements.
Solution
Matrix Type Order No. of elements


1 5 1
 Square matrix 3×3 9
 rows = columns 3 rows, 3 cols. 3×3=9
A = 2

2 4
 
2 −1 6
 
1
 
B = 0 Column matrix 3×1 3
3×1=3
 
1 single column 3 rows, 1 col.

C= 3 1 0 5 −3 1 Row matrix 1×6 6
single row 1 row, 6 cols. 1×6=6
Diagonal, symmetric and triangular matrices

Some square matrices occur so often in practice that they have their own names.
Diagonal matrices
   
1 2 3 4 1 2 3 4
   
5 6 7 8 5 6 7 8
A square matrix has two diagonals: 
   
9 
 0 1 2 9 0 1 2
  
3 4 5 6 3 4 5 6

In practice, the diagonal going downwards from left to right in the matrix (coloured red)
turns out to be more important than the other diagonal (coloured blue), so we give it a
special name: the leading diagonal.
A square matrix is called a diagonal matrix if all of the elements off the leading diagonal are
zero. The elements on the leading diagonal may or may not be zero.

 
  1 0 0 0
  2 0 0  
2 0
0 1 0 0 6 0 0
  
The matrices opposite are all diagonal matrices:   
  0

0 1 0 0 0

  
0 0 3  
0 0 0 6
Identity matrices
Diagonal matrices in which each element in the diagonal is 1 are of special importance.
They are called identity or unit matrices and have their own name and symbol (I).
Every order of square matrix has its
 
  1 0 0 0
own identity matrix, three of which are   1 0 0  
1 0
 I = 0 1 0 I = 0 1 0 0
  
shown opposite. I =    
0 1 0 0 1 0
 
 
0 0 1 
0 0 0 1

Symmetric matrices
A symmetric matrix is a square matrix that is
 
  1 2 4 6
unchanged by transposition (switching rows and   2 3 4  
2 3
3 1 5 2 1 5 7
  
columns). In a symmetric matrix, the elements  
  4

3 1 5 3 8
 
above the leading diagonal are a mirror image of the
  
4 5 3  
elements below the diagonal. Three are shown. 6 7 8 5
Triangular matrices
Triangular matrices come in two types:
1 An upper triangular matrix is a square matrix in which all elements below the leading
diagonal are zeros.
2 A lower triangular matrix is a square matrix in which all elements above the leading
diagonal are zeros.
Examples of triangular matrices are shown.  

1 0 0 0

1 2 3  
  3 2 0 0
0 4 5  
  
  6 5 4 0
0 0 6  
0 9 8 7
upper triangular matrix lower triangular matrix

Example 4 Types of matrices
Consider the following square matrices.

   
    1 0 0 1 3 5
1 0 1 2    
    0 4 0 3 4 7
   
0 1 0 3
   
   
0 0 3 5 7 2
Write down:
a the upper triangular matrices b the identity matrix
c the diagonal matrices d the symmetric matrices.
 
    1 0 0
1 2 1 0  
a All the elements below the leading   ,   and 0 4 0
 
0 3 0 1
  
diagonal are 0. 
0 0 3

b Elements in the leading diagonal are all  

1 and the other elements 0. 1 0
 
0 1
 
c All elements other than those in the 


 1 0 0

leading diagonal are zero. 1 0  
 , 0 4 0
  
0 1

 
0 0 3
d The matrix must be its own transpose. 
1 3 5 



1 0 0

3 4 7 , 1 0 and

     
0 4 0
     
  0 1  
5 7 2 0 0 3
Some notation
In some situations, we talk about a matrix and its elements without having specific numbers
in mind. We can do this as follows.
For the matrix A, which has n rows and m columns, we write:
a11 a12 a13 ... a1m a23

a21 a22 a23 ... a2m
A = a31 a32 a33 ... a3m
... ... ... ... ... row number column number
an1 an2 an3 ... anm
Thus:
a21 represents the element in the second row and the first column
a12 represents the element in the first row and the second column
a22 represents the element in the second row and the second column
amn represents the element in the mth row and the nth column.

Example 5 Identifying the elements in a matrix
For the matrices A and B, opposite, write down the

   
 1 5 3  3
values of:
   
A = −1 0 4 B = 2
a a12 b a21 c a33 d b31 .    
2 −2 6 1
a a12 is the element in the first row and the second a12 = 5
column of A.
b a21 is the element in the second row and the first a21 = −1
column of A.
c a33 is the element in the third row and the third a33 = 6
column of A.
d b31 is the element in the third row and the first b31 = 1
column of B.
In some instances, there is a rule connecting the value of each matrix with its row and
column number. In such circumstances, it is possible to construct this matrix knowing this
rule and the order of the matrix.
Example 6 Constructing a matrix given a rule for its ijth term
A is a 3 × 2 matrix. The element in row i and column j is given by ai j = i + j.

Construct the matrix.
1 The matrix is square and will have the form: j=1 j=2
 
  i=1 a11 a12 
a11 a12   
  i=2 a21 a22 
A = a21 a22   
i = 3 a31 a32
 
a31 a32
Use the rule where

a11 = 1 + 1 = 2 a12 = 1 + 2 = 3
ai j = i + j
a21 = 2 + 1 = 3 a22 = 2 + 2 = 4
to generate the elements one by one. For
a31 = 3 + 1 = 4 a32 = 3 + 2 = 5
example a32 = 3 + 2 = 5
2 Write down the matrix. 
2 3

 
A = 3 4
 
4 5

Entering a matrix into a CAS calculator

Later in this chapter, you will learn about matrix arithmetic: how to add, subtract and
multiply matrices. While it is possible to carry out these tasks by hand, for all but the
smallest matrices this is very tedious. Most matrix arithmetic is better done with the help of
a CAS calculator. However, before you can perform matrix arithmetic, you need to know
how to enter a matrix into your calculator.
CAS 1: How to enter a matrix on the TI-Nspire CAS

 
2 3 0
Enter the matrix A =   and determine its transpose (AT ).
1 4 2

Steps
1 Press ctrl + N . Select Add Calculator.
2 Press t and use the cursor arrows to
highlight the matrix template shown. Press ·.
Note: Math Templates can also be accessed by pressing
/+b>Templates.
3 Use the arrow to select the Number of rows

required (number of rows in this example is 2).
Press e to move to the next entry and repeat for

the Number of columns (the number of columns
in this example is 3).
Press e to highlight OK; press ·.

4 Type the values into the matrix template. Use
e to move to the required position in the matrix
to enter each value. When the matrix has been
completed, press e to move outside the matrix,
press / h, followed by A. Press ·. This
will store the matrix as the variable a.
 
2 3 0
5 When you type A (or a) it will paste in the matrix  . Press · to display.
1 4 2

6 To find aT , type in a (for matrix A) and then b>Matrix & Vector>Transpose

> enter as shown.
Note: Superscript T can also be accessed from the symbols palette (/ k).

CAS 1: How to enter a matrix using the ClassPad

 
2 3 0
Enter the matrix A =   and determine its transpose (AT ).
1 4 2

Steps
1 a Open the Main ( ) application Press to display
the soft keyboard.
b Select the keyboard.
2 Tap the 2 × 2 matrix icon, followed by the 1 × 2
matrix icon. This will add a third column and create a
2 × 3 matrix.
3 Type the values into the matrix template.
Note: Tap at each new position to enter the new value or use the
black cursor key on the hard keyboard to navigate to the new position.
4 To assign the matrix the variable name A:

a move the cursor to the very right-hand side of the matrix
b tap the variable assignment key followed
by h A
c press to confirm your choice.
Note: Until it is reassigned, A will represent the matrix as defined
above.
5 To calculate the transpose matrix AT :

a type and highlight A (by swiping with the stylus)
b select Interactive from the menu bar, tap Matrix-
Create and then tap trn.

482 Chapter 10 Matrices 10A
Skill-
sheet Exercise 10A
Order of a matrix
Example 1 1 State the order of each of the following matrices.
     
   2 6
 18 3 4 4
1 5 9      
a   b 7 6 12 c  3 4 d  7  e 0 3 3
3 0 4      
11 8 1 0 3 3
     
2 Write down the order of the following matrices:

 
1
   
1 1 2 2
a   b   c 2 2 3
2 1 0 3
  
 
4
3 How many elements are there in a

a 2 × 6 matrix? b 3 × 5 matrix? c 7 × 4 matrix?
4 A matrix has 12 elements. What are its possible orders? (There are six.)
The transpose of a matrix

Example 2 5 Determine the transpose of each matrix
     
1 2 3 9 1 0 7
a   b   c  
0 3 5 8 9 1 5
   
Types of matrices and their elements

Example 3 6 For each of the following matrices, state whether it is a column, row or square matrix,
and state the order and the number of elements.
 
  3
3 2  
a  
 

 b 5   c 7 11 2 1
2 3  
6
 
Example 4 7 Consider the following square matrices.

   
    1 0 0 1 3 5
1 2 3 0    
    0 1 0 0 4 7

2 1
 
0 5
   
 
0 0 1 0 0 2
  
Identify:
a the upper triangular matrices b the identity matrix
c the diagonal matrices d the symmetric matrices.

10A 10A What is a matrix? 483
Example 5 8 Complete the sentences below that relate to the following matrices.
 
   2 6   
3     4 3 1
   0 1 −1 0   
A = 2 1 1 0 2 B = 2 C =   D =   E = 0 −1 0
−1 2 1 3
      
1   2 0 4
4 −4
 
a The square matrices are and .

b Matrix B has rows.
c The row matrix is .
d The column matrix is .
e Matrix D has rows and columns.
f The order of matrix E is × .
g The order of matrix A is × .
h The order of matrix B is × .
i The order of matrix D is × .
j There are elements in matrix E.
k There are elements in matrix A.
l a14 = m b31 = n c11 = o d41 =
p e22 = q d32 = r b11 = s c12 =
Constructing a matrix given a rule for its ijth term

Example 6 9 B is a 3 × 2 matrix. The element in row i and column j is given by bi j = i × j.
10 C is a 4 × 1 matrix. The element in row i and column j is given by ci j = i + 2 j.

11 D is a 3 × 2 matrix. The element in row i and column j is given by di j = i − 3 j.

12 E is a 1 × 3 matrix. The element in row i and column j is given by ei j = i + j2 .

13 F is a 2 × 2 matrix. The element in row i and column j is given by fi j = (i + j)2 .

Entering a matrix into a CAS calculator and determining the transpose

14 Enter the following matrices into your calculator and determine the transpose.
     
1 0 3  4 −4 2
a B =   b C =   c E = 1 −1 2 d F =  
2 −2 1 −2 6 3

484 Chapter 10 Matrices 10A

15 Consider the following four matrices.
       
5 0 5 0 5 0 0 0
 
5
 0 0 0 5 0 5 0
    
0 5 7 5 0 5 0 0 0 7 0 0 0 5 7 5
   
    
0 0 5 0 0 0 5 0 0 0 7 0 0 8 5 0
       
0 0 0 5 0 0 0 5 0 0 0 5 9 0 0 5
     
How many of these matrices are diagonal matrices?

A 0 B 1 C 2 D 3 E 4
16 The element in row i and column j of matrix A is ai j . A is a 3 × 3 matrix. It is

constructed using the rule mi j = 3i + 2 j. A is
         
5 3 5  1 3 5 5 9 10  5 7 9  5 8 11
         
A 1 4 7  B 0 4 7 C 7 11 12 D  8 10 12 E 7 10 13
         
6 7 15 0 0 2 9 13 14 11 13 15 9 13 15
         
 
5 8 11
 
17 Consider the matrix A, where A = 7 10 13. The element in row i and column j of
 
9 12 15
matrix A is ai j . The elements in matrix A are determined by the rule
A ai j = 4 + j B ai j = 2i + 3 j C ai j = i + j + 3
D ai j = 3i + 2 j E ai j = 2i − j + 2
 
1 3 5 0
 
3 4 1 0
18 The matrix   is an example of

5 1 2 0
 
0 0 0 0

A a diagonal matrix. B an identity matrix. C a symmetric matrix.
D an upper triangular matrix. E a lower triangular matrix.
19 The element in row i and column j of matrix A is ai j . The elements in matrix A are
determined using the rule ai j = 2i + j. Matrix A could be
   
  3 3 4 5
3 4    
A   B 3 4 5 C 4 D [5] E 5 6 6
5 4    
6 7 8 7
   

10B Using matrices to represent information 485
10B Using matrices to represent information

Learning intentions
I To be able to represent data given in a table by a matrix.
I To be able to represent a network diagram with a matrix.
At the start of this chapter we used a matrix to store numerical information in a data table.
Matrices can also be used to carry codes that encrypt credit-card numbers for internet
transmission or to carry all the information needed to solve sets of simultaneous equations.
A less obvious application is using matrices to represent network diagrams.
Using a matrix to represent data tables

The numerical information in a data table is frequently presented in rows and columns. As
such, it is a relatively straightforward process to convert this information into matrix form.
Example 7 Representing information in a table by a matrix
The table opposite shows the three Gym membership

types of membership of a local
Gender Weights Aerobics Fitness
gym and the number of males and
females enrolled in each. Construct Males 16 104 86
a matrix to display the numerical Females 75 34 94
information in the table.
1 Draw a blank (2 × 3) matrix.
W A F
Label the rows M for male and F for female.  
M  
Label the columns W for weights, A for aerobics  
F
and F for fitness.
2 Fill in the elements of the matrix row by row,
W A F
starting at the top left-hand corner of the table.  
M  16 104 86
 
F 75 34 94
Example 8 Entering a credit card number into a matrix
Convert the 16-digit credit card number: 4454 8178 1029 3161 into a 2 × 8 matrix, listing
the digits in pairs, one under the other. Ignore the spaces.

1 Write out the sequence of numbers. 4454 8178 1029 3161
Note: Writing the number down in groups of four (as on
the credit card) helps you keep track of the figures.
2 Fill in the elements of the matrix row by row, 
4 5 8 7 1 2 3 6

starting at the top left-hand corner of the table.  
4 4 1 8 0 9 1 1
 
Using matrices to represent network diagrams

A less obvious use of matrices is to represent the information contained in network
diagrams. Network diagrams consist of a series of numbered or labelled points joined in
various ways. They are a powerful way of representing and studying things as different as
friendship networks, airline routes, electrical circuits and road links between towns.
Example 9 Representing a network diagram by a matrix
Represent the network diagram shown opposite as a 4 × 4 A

matrix M, where the: B
matrix element = 1 if the two points are joined by a line

matrix element = 0 if the two points are not connected. C
1 Draw a blank 4 × 4 matrix, labelling the rows and A B C D
columns A, B, C, D to indicate the points. A 
 

 
B  


C 

 

D
2 Fill in the elements of the matrix row by row, A B C D

starting at the top left-hand corner: A 0

1 0 1 

m11 = 0 (no line joining point A to itself)  
B 1 0 1 1 
m12 = 1 (a line joining points A and B)  
C 0 1 0 0 
m13 = 0 (no line joining points A and C)  
D 1 1 0 0
m14 = 1 (a line joining points A and D)
and so on until the matrix is complete.
Note: If a network contains no ‘loops’ (lines joining points to themselves) the elements in the leading
diagonal will always be zero. Knowing this can save a lot of work.

10B 10B Using matrices to represent information 487
Example 10 Interpreting a matrix representing a network diagram
The diagram opposite shows the roads connecting four towns: town A B
A, town B, town C, and town D. This diagram has been represented
C
by a 4 × 4 matrix, M. The elements show the number of roads
D
between each pair of towns.
a In the matrix M, m24 = 1. What does this tell
A B C D
us? 
A  0 1 0 0 

b In the matrix M, m34 = 3. What does this tell
 
M= B  1 0 2 1 
us?  
C  0 2 0 3 
c In the matrix M, m41 = 0. What does this tell  
D 0 1 3 0
us?
d What is the sum of the elements in row 3 of
matrix M and what does this tell us?
e What is the sum of all the elements of matrix
M and what does this tell us?
Explanation
a There is one road between town B and town D.
b There are three roads between town D and town C.
c There is no road between town D and town A.
d 5: the total number of roads between town C and the other towns in the network.
e 14: the total number of different ways you can travel between towns.
Note: For each road, there are two ways you can travel; for example, from town A to town B (a12 = 1)
and from town B to town A (a21 = 1).
Exercise 10B
Representing a table of data in matrix form

Example 7 1 The table opposite gives the Household Residents TVs Computers
number of residents, TVs and
A 4 2 1
computers in three households.
B 6 2 3
C 2 1 0
Use the table to:
a construct a matrix to display the numbers in the table. What is its order?
b construct a row matrix to display the numbers in the table relating to household B.
What is its order?
c construct a column matrix to display the numbers in the table relating to computers.
What is its order? What does the sum of its elements tell you?

488 Chapter 10 Matrices 10B
2 The table opposite gives the Car sales Small Medium Large
yearly car sales for two car
Honest Joe’s 24 32 11
dealers.
Super Deals 32 34 9
Use the table to:
a construct a matrix to display the numbers in the table. What is its order?
b construct a row matrix to display the numbers in the table relating to Honest Joe’s.
What is its order?
c construct a column matrix to display the numerical information in the table relating
to small cars. What is its order? What does the sum of its elements tell you?
3 Four exporters A, B, C and D sell refrigerators (R), dishwashers (D), microwave

ovens (M) and televisions (T ). The sales in a particular month can be represented by a
matrix.
R D M T
 
A 120 95 370 250
 
B
430 380 950 900
 
 
C  60 50 150 100
 
D 200 100 470 50
a What is the order of the matrix?
b Construct a row matrix to display the exports of B. What is its order? What does the
sum of the elements of this matrix tell you?
c Construct a column matrix to display the numerical information in the table relating
to microwave ovens. What is its order?
4 At a certain school there are 200 girls and 110 boys in Year 7. The numbers of girls
and boys in the other year levels are 180 and 117 in Year 8, 135 and 98 in Year 9,
110 and 89 in Year 10, 56 and 53 in Year 11, and 28 and 33 in Year 12. Summarise this
information in a matrix.
Example 8 5 Convert the 16-digit credit card number 3452 8279 0020 3069 into a 2 × 8 matrix. List
the digits in pairs, one under the other. Ignore any spaces.
6 The statistics for five members of a basketball team are recorded as follows:
Player A points 21, rebounds 5, assists 5
Player B points 8, rebounds 2, assists 3
Player C points 4, rebounds 1, assists 1
Player D points 14, rebounds 8, assists 6
Player E points 0, rebounds 1, assists 2
Express this information in a 5 × 3 matrix.

10B 10B Using matrices to represent information 489
Representing the information in a network diagram in matrix form

Example 9 7 Represent each of the following network diagrams by a matrix A using the rules:
matrix element = 1 if points are joined by a line
matrix element = 0 if points are not joined by a line.
a 1 b 1 c
2 2 1
2
3 3
3
4 4
8 The diagram opposite shows the roads interconnecting three town 1 town 2
towns: town 1, town 2 and town 3. Represent this diagram with
a 3 × 3 matrix where the elements represent the number of roads town 3
between each pair of towns.
Example 10 9 The network diagram opposite shows a friendship network girl 1 girl 2
between five girls: girl 1 to girl 5.
This network has been represented by a 5 × 5 matrix, F, using girl 3
the rule:
element = 1 if the pair of girls are friends girl 5 girl 4
element = 0 if the pair of girls are not friends.

1 2 3 4 5
 
a In the matrix F, f34 = 1. What does this tell us? 1 0 1 0 0 0
 
b In the matrix F, f25 = 0. What does this tell us? 2 1 0 1 0 0
 
c What is the sum of the elements in row 3 F= 3 0 1 0 1 1
 
of the matrix and what does this tell us? 4 0 0 1 0 1
 
d Which girl has the least friends? The most friends? 5 0 0 1 1 0
10 a The diagram below shows a ‘food web’ for polar bears (P), seals (S ) and
cod (C).
The matrix F below has been set up to represent the information in the diagram.
Polar bear
P S C
 
P 0 1 1 
eats eats
F = S  0 0 1 
 
 
C 0 0 0
Seal Cod
eats
i What does the ‘1’ in column C, row P, of matrix F represent?
ii What does the column of zeroes in matrix F represent?

490 Chapter 10 Matrices 10B
b The diagram below shows a ‘food web’ for polar bears (P), seals (S ) cod (C) and
zooplankton (Z).
Complete the matrix W to represent the information in the diagram.
P S C Z Polar bear
 
P  
 
S 
  eats eats
W = 


C  
 
Z
Seal Cod
eats
eats eats
zooplankton

11 The matrix below shows the number of boys and girls at a large school.
Boys Girls
 
Year 10  157 163 
 
Year 11  148 154 
 
Year 12 139 145
How many girls are there in Year 11?
A 157 B 148 C 154 D 139 E 145
12 A small chain of delicatessans with shopnames Allbright, Bunchof, Crisp, Delic and
Elite (A, B, C, D, E) each sell the products Feta, Goatmilk, Haloumi, Insalata and Jam
(F, G, H, I, J). The number of weekly sales of each product at each of the shops is
shown in the matrix below.
 A B C D E 
F  34 40 52 106 27 
 
 
G  42 154 38 55 68 
P = H  136 145 11 44 77 
 
 
I  136 147 43 72 111
 
 
J 139 140 66 56 145
Find which product had the highest weekly sales at any single one of the shops. The
name of this product and the shop is
A Feta at Allbright B Goatmilk at Bunchof C Jam at Elite
D Insalata at Bunchof E Haloumi at Allbright

10C Matrix arithmetic: addition, subtraction and scalar multiplication 491
10C Matrix arithmetic: addition, subtraction and

scalar multiplication
Learning intentions
I To be able to establish when two matrices are equal.
I To be able to recognise when matrix addition and subtraction are defined and to
peform these operations.
I To be able to undertake scalar multiplication.
I To be able to identify zero matrices.
I To be able to undertake addition, subtraction and multiplication by a scalar with a
CAS calculator.
I Using the operations of addition, subtraction and scalar multiplication of matrices in
practical situations.
Equality of two matrices

Equal matrices have the same order and each corresponding element is identical in value. It
is not sufficient for the two matrices to contain an identical set of numbers; they must also be
in the same positions.
For example:
   
1 2 1 2
  is equal to   because the corresponding elements are equal.
3 4 3 4
   
1 2 4 3
  is not equal to   because the numbers are in different positions.
3 4 2 1
   
Example 11
Given that
   
2 a   2 10
  =  
8 b+1 20 7
find the value of a and the value of b.
Solution
a = 10 and b + 1 = 7 which implies b = 6.
Matrix addition and subtraction

Adding and subtracting matrices
If two matrices are of the same order (have the same number of rows and columns), they can
be added (or subtracted) by adding (or subtracting) their corresponding elements.

Example 12 Adding two matrices

   
2 3 0 1 2 3
A =    B =    Find A + B.
1 4 2 2 −2 1
 
1 As the two matrices have the same 
 2 3 0 
 
 1 2 3 

order, 2 × 3, they can be added. A + B =   +  
1 4 2 2 −2 1
  
 
 2 + 1 3+2 0 + 3 
=   
1 + 2 4 + (−2) 2 + 1
2 Add corresponding elements.  

 3 5 3 
=  
3 2 3

Likewise, if we have two matrices of the same order (same number of rows and columns),
we can subtract the two matrices by subtracting their corresponding elements.
Example 13 Subtracting two matrices

   
 2 3 0   1 2 3 
A =   B =   Find A − B.
1 4 2 2 −2 1
 
1 As the two matrices have the same
   
order, 2 × 3, they can be subtracted.  2 3 0   1 2 3 
A − B =   −  
1 4 2 2 −2 1
  
 
 2 − 1 3 − 2 0 − 3 
=  
1 − 2 4 − (−2) 2 − 1

2 Subtract corresponding elements. 

 1 1 −3 

=   
−1 6 1

Multiplying matrices by a number (scalar multiplication)

Mutiplying a matrix by a number has the effect of multiplying each element in the matrix by
that number.
Multiplying a matrix by a number is called scalar multiplication, because it has the effect
of scaling the matrix by that number. For example, multiplying a matrix by 2 doubles each
element in the matrix.

Example 14 Scalar multiplication

   
 2 3 0   4 −4 
A =   C =    Find 3A and 0.5C.
1 4 2 −2 6
 
   
Multiplying a matrix by a number has the 2 3 0 6 9 0
3   =  
effect of multiplying each element by that 1 4 2 3 12 6
  
number.   
 4 −4  2 −2

0.5   =  
−2 6 −1 3
  
The zero

matrix
       
 1 2   1 2   1 2   1 2   0 0 
If X =   and Y =  , then X − Y =   −   =   = O.
3 4 3 4 3 4 3 4 0 0
        
A matrix of any order with all zeros is known as a zero matrix. The symbol O is used to
represent a zero matrix. The matrices below are all examples of zero matrices.
 
 0   
     0 0 
 0 0 
 , O =  0  , O =  0 0 
   
O = [0], O =   
0 0  0 
  
 
  0 0
0
Example 15 The zero matrix

   
 6 0   9 0 
G =    H =    Show that 3G − 2H = O.
−4 2 −6 3
 
Solution
       
 6 0   9 0   18 0   18 0 
3G − 2H = 3 ×   − 2 ×   =   −  
−4 2 −6 3 −12 6 −12 6
   
   
 18 − 18 0 − 0   0 0 
=   =  
−12 − (−12) 6 − 6 0 0
  
∴ 3G − 2H = O

Using a CAS calculator to perform matrix addition,

subtraction and scalar multiplication
For small matrices, it is usually quicker to add, subtract or multiply a matrix by a number
(scalar multiplication) by hand. However, if dealing with larger matrices, it is best to use a
CAS calculator.
CAS 2: How to add, subtract and scalar multiply matrices using the
TI-Nspire CAS
   
 2 3 0   1 0 3 
If A =   and B =  , find:
1 4 2 2 −2 1
  
a A+B b A−B c 3A − 2B
Steps
2 Enter and store the matrices A and B into
your calculator.
a To determine A + B, type a + b.
Press · to evaluate.
b To determine A − B, type a − b.
c To determine 3A − 2B, type 3a − 2b.
CAS 2:How to add, subtract and scalar multiply matrices with the
ClassPad
   
 2 3 0   1 0 3 
If A =   and B =  , find:
1 4 2 2 −2 1
 
a A+B b A−B c 3A − 2B
Steps
1 Enter the matrices A and B into your calculator using the h keyboard.

2 a To calculate A + B, type A + B and then

press to evaluate.
b To calculate A − B, type A − B and then
press to evaluate.
c To calculate 3A − 2B, type 3A − 2B and then
press to evaluate.
Example 16 Processing data using addition, subtraction and scalar multiplication
The sales data for two used car dealers, Honest Joe’s and Super Deals, are displayed
below.
2014 2015
Car sales Small Medium Large Small Medium Large
Honest Joe’s 24 32 11 26 38 16
Super Deals 32 34 9 35 41 12
Solution
a Construct two matrices, A and B, to  
24 32 11
 
26 38 16
represent the sales data for 2014 and A =   B =  
32 34 9 35 41 12

2015 separately.
b Construct a new matrix C = A + B. C = A+B
What does this matrix represent?   
24 32 11 26 38 16

=    + 
  
32 34 9 35 41 12

 
50 70 27
=  
67 75 21

Matrix C represents the total sales for 2014

and 2015 for the two dealers.
c Construct a new matrix, D = B − A. D = B−A
What does this matrix represent?  
2 6 5
=   
3 7 3

Matrix D represents the increase in sales

from 2014 and 2015 for the two dealers.

496 Chapter 10 Matrices 10C
d Both dealers want to increase their  

26 38 16
2015 sales by 50% by 2016. Construct E = 1.5B = 1.5 ×   
35 41 12

a new matrix E = 1.5B. Explain why  
this matrix represents the planned sales 1.5 × 26 1.5 × 38 1.5 × 16
=  
figures for 2016. 1.5 × 35 1.5 × 41 1.5 × 12

 
 39 57 24 
=  
52.5 61.5 18
Forming the scalar product 1.5B multiplies
each element by 1.5. This has the effect of
increasing each value by 50%.
Skill-
sheet Exercise 10C
Equality of two matrices

Example 11 1 Given that
   
4 a + 2  3 11
  =  
6 b−1 10 6
   
find the value of a and the value of b.
Matrix addition, subtraction and scalar multiplication

Example 12 2 The questions below relate to the following six matrices. Computations will be quicker
Example 13 if done by hand.
 
0 1 4
A= 1 3 B= 3 1 C =  
3 2 1

     
 0 1 1 0  0 1 4
D =   E =   F =  
−1 2 2 −1 3 2 1
  
a Which matrices are equal?

b Which matrices have the same order?
c Which matrices can be added or subtracted?
d Compute each of the following, where possible.
i A+B ii D + E iii C − F iv A − B v E−D
Example 14 vi 3B vii 4F viii 3C + F ix 4A − 2B x E+F
   
 8 0  2 0 
Example 15 3 Let G =   and H = 
1 .
 Show that 2G − 8H = O.
−4 2 −1
 
2

10C 10C Matrix arithmetic: addition, subtraction and scalar multiplication 497
4 Evaluate each of the folowing giving your answer as a single matrix.

       
1 2 4 3 1 2 4 3
a   +   b   −  
4 3 1 2 4 3 1 2
     
   
1 2 4 3
c    + 2 × 
   d 1 −1 + −1 1
4 3 1 2

       
0 1 0 1
e   +   f 3   + 2  
1 0 1 0
       
0 1 2 4 3
g 3   − 2   h   +  
1 0 1 1 2

5 Use a calculator to evaluate the following.

       
1 2 4 3 1.2 0.2 0.4 4
a 2.2 ×   − 1.1 ×   b   − 3.5 ×  
4 3 1 2 4.5 3.3 1 2

   
1 2 1  0 1 −1    
    1 2 1
4 −1 2 1 0
c 5 × 4 0 1 − 2 ×  2 0 1  d 0.8 ×   + 0.2 ×  
1 0 −1 2 1 0 1 2
    
0 1 0 0.5 0 −2
   
     
4 2  x y  10 8
6 Find the values of x, y, z and w in the following. 2   +   =  
3 5 z w 12 6
Applications of matrix addition, subtraction and scalar multiplication
Example 16 7 The number of DVDs sold in a company’s city, suburban and country stores for each
3-month period in a year is shown in the table.
DVD sales (thousands)

Store location Jan–March April–June July–Sept Oct–Dec
City 2.4 2.8 2.5 3.4
Suburban 3.5 3.4 2.6 4.1
Country 1.6 1.8 1.7 2.1
a Construct four 3 × 1 matrices A, B, C, and D that show the sales in each of the
three-month periods during the year.
b Evaluate A + B + C + D. What does the sum A + B + C + D represent?
8 The numbers of females and males enrolled in three different gym programs for 2014
and 2015, Weights, Aerobics and Fitness, are shown in the table.
2014 2015
Gym membership Weights Aerobics Fitness Weights Aerobics Fitness
Females 16 104 86 24 124 100
Males 75 34 94 70 41 96
498 Chapter 10 Matrices 10C
a Construct two matrices, A and B, which represent the gym memberships for 2014
and 2015 separately.
b Construct a new matrix C = A + B. What does this matrix represent?
c Construct a new matrix D = B − A. What does this matrix represent? What does the
negative element in this matrix represent?
d The manager of the gym wants to double her 2015 membership by 2018. Construct
a new matrix E that would show the membership in 2018 if she succeeds with her
plan. Evaluate.

       
−4 2  4 2 6 2 4 2
9   +   −   +   =
−3 1 −3 −1 3 1 3 1
       
         
−2 2  0 4 −2 4 −2 0  0 2
A   B   C   D   E  
−3 1 −3 1 −6 0 −6 0 −6 0
    
   
 0 2 0 4
10 If M =    and N = 
  , then 2M − 2N =
−3 1 3 0

         
 0 0  0 −2  0 −4  0 4 0 2
A   B   C   D   E  
−9 2 −6 1 −12 2 12 −2 6 −1
    
11 The table below shows information about two matrices A and B.

Order Rule
A 3×3 ai j = i j
B 3×3 bi j = i + j
The sum A + B is
     
1 2 3 2 3 4 3 5 7 
     
A 2 4 6 B 3 4 5 C 3 4 5 
     
3 6 9 4 5 6 3 6 19
    
   
 3 2 7  3 5 7 
   
D  3 4 5  E 5 8 11
   
14 15 16 7 11 15
  

10D Matrix arithmetic: the product of two matrices 499
10D Matrix arithmetic: the product of two matrices

Learning intentions
I To be able to establish when multiplication of two matrices can be undertaken.
I To be able to undertake the multiplication of matrices.
I To be able to understand that the order of multiplication of matrices is important.
I To be able to apply multiplication of matrices to practical problems.
I To be able to use your CAS calculator to undertake multiplication of matrices.
I To be able to understand the role of ‘summing matrices’.
I To be able to work with calculations involving powers of matrices.
The process of multiplying two matrices involves both multiplication and addition. The
process can be illustrated using Australian Rules football scores.
An illustration of matrix multiplication

Two teams, the Ants and the Bulls, play each other in a game of Australian rules football.
At the end of the game:
the Ants had scored 11 goals 5 behinds
the Bulls had scored 10 goals 9 behinds.
Now calculate each team’s score in points:

one goal = 6 points
one behind = 1 point.
Thus we can write:

11 × 6 + 5 × 1 = 71 points
10 × 6 + 9 × 1 = 69 points
Matrix multiplication follows the same pattern.
Point Final
Goals Behinds values points
       
Ants score: 11 5   6  11 × 6 + 5 × 1
 = 71
 
  ×   = 
Bulls score: 10 9 1 10 × 6 + 9 × 1 69
     
The order of matrices and matrix multiplication

Look at the order of each of the matrices involved in the matrix multiplication below.
Point Final
Goals Behinds
values points
       
Ants score: 11 5   6  11 × 6 + 5 × 1 71
  ×   =   =  
Bulls score: 10 9 1 10 × 6 + 9 × 1 69
  
Order of matrices: 2×2 2×1 2×1

Thus, multiplying a 2 × 2 matrix by a 2 × 1 matrix gives a 2 × 1 matrix.
Two observations can be made here:

1 To perform matrix multiplication, the number of columns in the first matrix (2) needs
to be the same as the number of rows in the second matrix (2). For example, if there
were three columns in the first matrix, there would not be enough elements in the
second matrix to complete the multiplication. When this happens, we say that matrix
multiplication is not defined.
2 The final result of multiplying the two matrices is a 2 × 1 matrix. For each row in the first
matrix, there will be a row in the product matrix (there are two rows). For each column in
the second matrix, there will be a column in the product matrix (there is one column).
These observations can be generalised to give two important rules for matrix multiplication.
Rule 1: Condition for matrix multiplication to be defined

Matrix multiplication of two matrices requires the number of columnsin the first matrix to
equal the number of rows in the second matrix.
That is, if A is of order m × n and B is of order r × s, then the product AB is only defined
if n = r.
 
   1   
 2 0 2
, B =  3  and C =  2 0 , then:
    
For example, if A =   
3 1 4 1 3
  
 
2
 
  1
3 0 2  
AB =   × 3 is defined: columns in A(3) = rows in B(3)
  
3 1 4  
2
2×3 3×1
 
1  
  2 0
BC = 3 ×   is not defined: columns in B(1) , rows in C(2).
  1 3

2
3×1 2×2
Example 17 Is a matrix product defined?

   
 6 0  1 
A =   , B = 3 1 and C =  
−4 2 −1

Which of the following matrix products are defined?

a AB b BC c AC

a Write down the matrix product. Under  
 6 0
each matrix, write down its order AB =   3 1 ; not defined
−4 2

(columns × rows)
order: 2×2 1×2
b The matrix product is defined if the

 
 1 
number of columns in matrix 1 = the BC = 3 1   ; defined
−1
 
number of rows in matrix 2.
order: 1×2 2×1
c Write down your conclusion.  

 6 0
 
 1 
AC =     ; defined
−4 2 −1
  
order: 2×2 2×1
Once we know that two matrices can be multiplied, we can use the order of the two matrices
to determine the order of the resulting matrix.
Rule 2: Determining the order of the product matrix

If two matrices can be multiplied, then the product matrix will have the same number of
rows as the first matrix and the same number of columns as the second matrix.
That is, if A is of order m × n and B is of order n × s, then AB will be of order m × s.
 
  1
2 0 2
For example, if A =   and B = 3, then:
 
 
3 1 4

 
2
 
  1
2 0 2  
AB =   × 3 is defined and will be of order 2 × 1.
  
3 1 4  
2
2×3 3×1 2×1

equal
   
2 0 2 0 2
AB =   ×   is defined and will be of order 2 × 3.
1 3 3 1 4
  
2×2 2×3 2×3

equal

Example 18 Determining the order of a matrix product
   
 6 0
 , B = 3 1 and C =  1 
 
A = 
−4 2 −1
  
The following matrix products are defined. What is their order?

a BA b BC c AC
1 Write down the matrix product.

 6 0

a BA = 3 1   ; order of BA 1 × 2
 
Under each matrix, write down
−4 2
its order rows × columns.
order: (1 × 2) (2 × 2)
2 The order of the product matrix
 
 1 
b BC = 3 1   ; order of BC 1 × 1
 
is given by rows in matrix 1 ×
−1
columns in matrix 2.
order: (1 × 2) (2 × 1)
3 Write down the order.   
 6 0  1 
c AC =     ; order of AC 2 × 1
−4 2 −1
order: (2 × 2) (2 × 1)
Order of multiplication is important when multiplying matrices

You might have noticed in Example 17 that while the matrix product BA was defined,
the matrix product AB in Example 18 was not defined. Order is important in matrix
multiplication. For example, if we have two matrices, M and N, and form the products
MN and N M, frequently the products will be different. We will return to this point when in
the next section.
Determining matrix products

For large matrices, the process of matrix multiplication is complex and can be error prone
and tedious to do by hand. Fortunately, CAS calculators will do it for us, and that is
perfectly acceptable.
However, before we show you how to use a CAS calculator to multiply matrices, we will
illustrate the process by multiplying a row matrix by a column matrix and a rectangular
matrix by a column matrix by hand.
In terms of understanding matrix multiplication, and using this knowledge to solve problems
later in this module, these are the two most important worked examples in the chapter.

Example 19 Multiplying a row matrix by a column matrix (by hand)

 
2

 
Evaluate the matrix product AB, where A = 1 3 2 and B = 4.
 
1
1 Write down the matrix product 3×1
and, above each matrix, write down 1×3  
 2 
AB = 1 3 2  
 
its order. Use this information to  4 
determine whether the matrix product  
1
is defined and its order.
Defined: the number of columns in A
equals the number of rows in B.
The order of AB is 1 × 1.
 
2 To determine the matrix product:  2 
 
 
a multiply each element in the row 1 3 2  4 
 
matrix by the corresponding element 1
in the column matrix = [1 × 2 + 3 × 4 + 2 × 1]
b add the results
= [16]
c write down your answer.
∴ AB = [16]
Example 20 Multiplying a rectangular matrix by a column matrix (by hand)

   
1 0 2
Evaluate the matrix product AB, where A =    and B =  .

2 3 3
1 Write down the matrix product 2×2 2×1
and, above each matrix, write down 
 1 0
 
  2 
its order. Use this information to AB =    
2 3 3
determine whether the matrix product
is defined and its order. Defined: the number of columns in A
equals the number of rows in B.
The order of AB is 2 × 1.

2 To determine the matrix product:

    
a multiply each element in the row  1 0   2   1 × 2 + 0 × 3 
    =  
matrix by the corresponding element 
2 3

3
 
2×2+3×3

in the column matrix
 
b add the results  2 
=  
c write down your answer. 13

 
 2 
∴ AB =  
13

Using a CAS calculator to multiply two matrices

In principle, if you can multiply a row matrix by a column matrix, you can work out the
product between any two matrices, provided it is defined. However, because you have to do
it for every possible row/column combination, it soon gets beyond even the most patient and
careful person. For that reason, in practice we use technology to do the calculation for us.
We will illustrate how to use a calculator to multiply matrices by evaluating the matrix
product in the football score example given earlier.
CAS 3: How to multiply two matrices using the TI-Nspire CAS

   
11 5
 and D = 6, find the matrix CD.
 
If C = 
10 9 1
  
Steps
2 Enter and store the matrices C and D into your
calculator.
3 To calculate matrix CD, type c × d. Press · to
evaluate.
Note: You must put a multiplication sign between the c
and d.
CAS 3:How to multiply two matrices using the ClassPad

   
11 5
 and D = 6 find the matrix product CD.
 
If C = 
10 9 1
  
Steps
1 Enter the matrices C and D into your calculator.

2 To calculate C × D, type CD and then press to evaluate.
Applications of the product of two matrices

The summing matrix
A row or column matrix in which all the elements are 1 is called a summing matrix.
The matrices opposite are all examples of summing
 
  1
matrices. 1  
  1
1 1 1 1 1 1 1  
The rules for using a summing matrix to sum the  
1
 
1
 
rows and columns of a matrix follow.  
1
Using matrix multiplication to sum the rows and columns of a matrix

To sum the rows of an m × n matrix, post-multiply the matrix by an
n × 1 summing matrix.
To sum the columns of an m × n matrix, pre-multiply the matrix by a
1 × m summing matrix.
Example 21 Using matrix multiplication to sum the rows and columns of a matrix
Use matrix multiplication to generate a matrix that:

a displays
   
1 3 0 1 3 0
   
i row sums of the matrix 2 6 7 ii column sums of the matrix 2 6 7
   
3 0 1 3 0 1
 
 
2 5 −1 −3 4
b displays the column sums of the matrix   .
0 6 2 −2 3

a i To sum the rows of a 3 × 3 matrix,    
 1 3 0   1   4 

post-multiply a 3 × 1 summing     
 2 6 7   1  =  15 
matrix.
     
     
3 0 1 1 4

 
1 3 0
 
1 1 1 2 6 7 = 6 9 8

 
ii To sum the columns of a 3 × 3 3 0 1
matrix, pre-multiply a 1 × 3
summing matrix.
b To sum the columns of a 2 × 5 matrix,
 
 2 5 −1 −3 4 
1 1 
 
pre-multiply a 1 × 2 summing matrix. 
0 6 2 −2 3

= 2 11 1 −5 7
Example 22 shows a practical application of matrix multiplication.
Example 22 A practical application of matrix multiplication

 
 25  Walk Matrix E gives the energy in kilojoules consumed per minute when
E =  
40 Run
 walking and running.
T = [20 40] Matrix T gives the times (in minutes) a person spent walking and
Walk Run running in a training session.
Compute the matrix product TE and show that it gives the total energy consumed during
the training session.
Solution
 
25

T × E = 20 40   = [20 × 25 + 40 × 40] = [2100]
 
40
The total energy consumed is:
20 minutes × 25 kJ/minute + 40 minutes × 40 kJ/minute = 2100 kJ
This is the value given by the matrix product T E.
You could work out the energy consumed on the training run for one person just as quickly
without using matrices. However, the advantage of using a matrix formulation is that, with
the aid of a calculator, you could almost as quickly have worked out the energy consumed by
10 or more runners, all with different times spent walking and running.
Matrix powers
Now that we can multiply matrices, we can also determine the power of a matrix. This is
an important tool when we meet communication and dominance matrices in the next section
and transition matrices in the next chapter.

The power of a matrix

Just as we define
22 as 2 × 2,
23 as 2 × 2 × 2,
24 as 2 × 2 × 2 × 2 and so on,
we define the various powers of matrices as
A2 as A × A,
A3 as A × A × A,
A4 as A × A × A × A and so on.
Only square matrices can be raised to a power.
Example 23 Evaluating matrix expressions involving powers

     
1 0  −1 1 0 1
If A =  , B =   and C =  , determine:
2 −1 2 1 1 1
    
a 2A + B2 − 2C b (2A − B)2 − C 2 c AB2 − 3C 2
1 Write down the matrices.  
1 0 
 
−1 1
 
0 1
A =   , B =   , C =  
2 −1 2 1 1 1
    
2 Enter the matrices A, B and C into your

1 0 1 0
calculator. a
2 −1 2. −1.
−1 1 −1. 1.
b
2 1 2. 1.
0 1 0. 1.
c
1 1 1. 1.
5. −2.
2 · a + b2 − 2 · c
2. −1.
6. −1.
(2 · a − b)2 − c2
−1. 5.
0. −3.
a · b2 − 3 · c 2
3. −9.

508 Chapter 10 Matrices 10D
 
3 Type in each of the expressions as 5 −2
written, and press to evaluate. a 2A + B2 − 2C =  
2 −1

Write down your answer.  
 6 −1
b (2A + B)2 − C 2 =  
−1 5

 
0 −3
c AB − 3C = 
2 2  
3 −9

Note: For CAS calculators you must use a multiplication sign between a and b2 in the last example,
otherwise it will be read as variable (ab)2 .
Skill-
sheet Exercise 10D
Matrix multiplication
Example 17 1 The questions below relate to the following five matrices.
 
    2
3  0 1  
A= 1 3 B =   C = 1 0 −1 D =   E = 1
1 −1 2

 
0
a Which of the following matrix products are defined?
i AB ii BA iii AC iv CE
v EC vi EA vii DB viii CD
Example 19 b Compute the following products by hand.
Example 20 i AB ii CE iii DB iv AD
c Enter the five matrices into your calculator and compute the following matrix
expressions.
i AB ii EC iii AB − 3CE iv 2AD + 3B
2 Evaluate each of the following matrix products by hand.

 
    1
2 1  
a 0 2 ×   = b 1 0 ×   = c 2 0 1 3 =
×
   
0 2  
1
   
        1 0 1 2
1 2 1 1 1 4 3 
 ×   =  = 0 × 0 =
  
d   e    × 
  f 0 1
3 4 0 0 1 1 2
       
1 1 0 1


10D 10D Matrix arithmetic: the product of two matrices 509
3 Evaluate each of the following matrix products using a CAS calculator.

 
 0.5     
  1 2 4 3
a −1.5 × 2 4 6 =
  b    × 
   =
4 3 1 2
  
2.5
 
       
1 3 1  2 1  1 3 1  0 2 1
       
c 0 2 0 ×  1 −1 d 0 2 0 × −1 4 2 =
       
1 1 2 −1 2 1 1 2 −2 1 2
       
Using summing matrices to sum the rows and columns of matrices

 
 2 5
 
Example 21 4 For the matrix opposite, write down a matrix that can be used to: −1 1 .
 
9 3
a sum its rows b sum its columns
   
1 7 1 2
   
5 Show how the matrix 1 can be used to sum the rows of 1 2 2.
   
1 8 1 4
 
9 0 2
 
6 Show how the matrix 1 1 1 can be used to sum the columns of 1 7 3.
 
8 3 4
7 Use matrix multiplication to construct a matrix that:

 
2 4 1 7 8
 
1 9 0 0 2
 
a displays row sums of the matrix 3 4 3 3 5
 
2 1 1 1 7
 
 
 
5 3 6 7 9
 
4 5 1 2 1
 
b displays the column sums of the matrix 0 3 4 5 1.
 
4 2 1 7 9
 
Practical applications of matrix multiplication

Example 22 8 Six teams play an indoor soccer competition.
If a team:
wins, it scores two points  
2 Win
draws, it scores one point
 
P = 1 Draw
loses, it scores zero points.  
0 Lose
This is summarised in the points
matrix opposite.
W D L
 
4 1 0  Team 1
 
3 1 1  Team 2
 

The results of the competition are 3 0 2  Team 3
R =  
summarised in the results matrix. Work 1 2 2  Team 4

out the final points score for each team

1 1 3  Team 5

by forming the matrix product RP.  
0 1 4 Team 6
9 Four people complete a training session in which they walked,
 
25 Walk
jogged and ran at various times.
 
E = 40 Jog
The energy consumed in kJ/minute when walking, jogging or  
65 Run
running is listed in the energy matrix opposite.
The time spent in each activity (in minutes) W J R
by four people is summarised 
10

20 30  Person 1
in the time matrix opposite. Work out 
15 
20 25  Person 2
the total energy consumed by each T = 
 
20 20 20  Person 3
person, by forming the matrix product  
30 20 10 Person 4

T E.
10 A manufacturer sells three types of fruit straps, A B C
A, B and C, through outlets at two shops, Energy (E) 
25

34 19  E
and Nourishing (N). Q =  
30 45 25 N
The number of fruit straps sold per month at each
shop is given by the matrix Q.
a Write down the order of matrix Q.
The matrix P, shown opposite, gives  

 2.50  A
the selling price, in dollars, of each  
P =  1.80  B
type of fruit strap A, B and C.  
3.20 C
b i Evaluate the matrix M, where M = QP.
ii What information do the elements of matrix M provide?
c Explain why the matrix PQ is not defined.
11 Matrix X shows the number of cars of models a and b bought by four dealers A, B,
C, D. Matrix Y shows the cost in dollars of cars a and b. Find XY and explain what it
represents.
a b
 
A 3 1
   
B 2 2 26 000 a
X=   Y =  
C 1 4 32 000 b
 
 
D 1 1

10D 10D Matrix arithmetic: the product of two matrices 511
12 It takes John 5 minutes to drink a milk shake which costs $2.50, and 12 minutes to eat
a banana split which costs $3.00.
  
 5 12  1
a Find the product     and interpret the result in fast-food economics.
2.50 3.00 2
 
  
 5 12  1 2 0
b Two friends join John. Find      and interpret the result.
2.50 3.00 2 1 1
13 The final grades for Physics and Chemistry are made up of three components: tests,
practical work and exams. Each semester, a mark out of 100 is awarded for each
component. Wendy scored the following marks in the three components for Physics:
Semester 1 tests 79, practical work 78, exam 80

a Represent this information in a 2 × 3 matrix.
To calculate the final grade for each semester, the three components are weighted:
tests are worth 20%, practical work is worth 30% and the exam is worth 50%.
b Represent this information in a 3 × 1 matrix.
c Calculate Wendy’s final grade for Physics in each semester.
Wendy also scored the following marks in the three components for Chemistry:

d Calculate Wendy’s final grade for Chemistry in each semester.
Students who gain a total score of 320 or more for Physics and Chemistry over the two
semesters are awarded a Certificate of Merit in Science.
e Will Wendy be awarded a Certificate of Merit in Science?
She asks her teacher to re-mark her Semester 2 Chemistry exam, hoping that she will
gain the necessary marks to be awarded a Certificate of Merit.
f How many extra marks on the exam does she need?
Powers of matrices
 
2 1
Example 23 14 If A =    , determine A2 , A3 , A4 and A7 .
1 3

 
−1 1
15 If A =  , determine A4 , A5 , A6 and A7 .
1 2


 
2 0 2
 
16 If B = 1 2 3 use your calculator to find B3 .
 
1 3 1
     
2 1 −1 1 0 1 
17 If A =   , B =   and C =  , evaluate:
1 3 1 2 1 −2
    
a A + 2B − C 2 b AB − 2C 2 c (A + B + 2C)2
d 4A + 3B2 − C 3 e (A − B)3 − C 3

 
3
 

 
18 The matrix product 6 2 0 × 6 is equal to
 
9
     
18
1 3
   
 
   
A 30 B 12 C 39 D 3 × 2 2 0 × 2 E 2 × 3 2 0 × 2
     
0 3 0
     
 
1 4  
  1 5 2
19 Matrix P = 2 6 and matrix Q = 
  
2 1 6
  
1 2
Matrix R = P × Q. Element r31 is determined by the calculation
A 1×1+4×2 B 1×1+4×0 C 1×1+2×2
D 4×1+5×0 E 4×2+5×1
20 There are two types of chocolate boxes, Minty Chews and Orange Delight, available in
a shop. The cost, in dollars, to purchase a box is shown in the table below.
Chocolate Cost($)
Minty Chews 6
Orange Delight 8
Liam is doing all his Christmas shopping by buying chocolate boxes. He buys 7 boxes
of Minty Chews and 9 boxes of Orange Delight. The total cost in dollars of these
chocolates can be determined by which one of the following calculations?
 
7
A 7 × 6 8 B 7 9 × 6 8 C   × 6 8
9
   
6 6
D 7 9 ×   E 9 7 ×  
   
8 8

10E Matrix inverse, the determinant and matrix equations 513
10E Matrix inverse, the determinant and matrix

equations
Learning intentions
I To be able to detemine by multiplication when two matrices are inverses.
I To be able to determine the determinant of a 2 × 2 matrix.
I To be able to determine the inverse of a 2 × 2 matrix.
I To be able to use a CAS calculator to determine the inverse and determinant of an
n × n matrix.
I To be able to solve simple matrix equations.
The inverse matrix A−1

So far, you have been shown how to add, subtract and multiply matrices, but what about
dividing them? As you might expect matrix division, like matrix multiplication, is a more
complicated process than its equivalent process for dividing numbers.
The starting point for matrix division is the inverse matrix. You will see why as we
proceed.
The inverse matrix A−1

The inverse of a square matrix A is called A−1 .
The inverse matrix has the property AA−1 = A−1 A = I.
Having defined the inverse matrix, two questions immediately come to mind. Does the
inverse of a matrix actually exist? If so, how can we calculate it?
First we will demonstrate that at least some matrices have inverses. We can do this by
showing that two matrices, which we will call A and B, have the property AB = I and
BA = I, where I is the identity matrix. If this is the case, we can then say that B = A−1 .
Example 24 Demonstrating that two matrices are inverses

   
2 3  5 −3
Show that the matrices A =   and B =   are inverses.
3 5 −3 2
  
   
2 3  5 −3
1 Write down A and B. A =   B =  
3 5 −3 2
  

  
2 3  5 −3
2 Form the product AB and AB =    
3 5 −3 2
 
evaluate. You can use your
 
calculator to speed things up 2 × 5 + 3 × (−3) 2 × (−3) + 3 × 2
=  
if you wish. 3 × 5 + (−5) × 3 3 × (−3) + 5 × 2

 
1 0
=   
0 1

∴ AB = I
3 Form the product BA and  
 5 −3 2 3

evaluate. You can use your BA =    
  
−3 2 3 5

calculator here to speed things  
up if you wish. 5 × 2 + (−3) × 3 5 × 3 + (−3) × 5
=  
(−3) × 2 + 2 × 3 (−3) × 3 + 2 × 5

 
1 0
=  
0 1
∴ BA = I
4 Write down your conclusion. Because AB = I and BA = I, we conclude that A
and B are inverses.
   
2 3  5 −3
While Example 24 clearly demonstrates that the matrices A =    and B = 
  
3 5 −3 2

both have an inverse, many square matrices do not have inverses. To see why, we need to
introduce another new matrix concept, the determinant, and see how it relates to finding the
inverse of a square matrix. To keep things manageable, we will restrict ourselves initially to
2 × 2 matrices.
The determinant of a matrix

The determinant of a 2 × 2 matrix
 
a b
If A =  , then the determinant of matrix A is given by:
c d
a b
det(A) = =a×d−b×c
c d

Example 25 Finding the determinant of a 2 × 2 matrix
Find the determinant of the matrices:

     
2 3 2 3 2 4
a A =   b B =   c C =  
3 5 2 3 2 3
  
Solution
a b
1 Write down the matrix and use the rule det(A) = = a × d − b × c.
c d
2 Evaluate.
 
2 3 2 3
a A =   ∴ det(A) = =2×5−3×3=1
3 5 3 5
 
2 3 2 3
b B =   ∴ det(B) = =2×3−2×3=0
2 3 2 3
 
2 4 2 4
c C =   ∴ det(C) = = 2 × 3 − 2 × 4 = −2
2 3 2 3
From Example 25, we can see that the determinant of a matrix is a number that can take on
both positive and negative values as well as being zero. For a matrix to have an inverse, its
determinant must be non-zero.
How to determine the inverse of a 2 × 2 matrix

Normally you will use a calculator to determine the inverse of a matrix, but we need to
do the following example by hand to show you why some square matrices do not have an
inverse. To do this we first need to consider the rule for finding the determinant of a 2 × 2
matrix.
The rule for finding the inverse of a 2 × 2 matrix

 
a b
If A =  , then its inverse, A−1 , is given by
c d
   
1  d −b 1  d −b
A−1 =   =  
ad − bc −c a  det(A) −c a 
provided ad − bc , 0; that is, provided det(A) , 0.
The most important thing about this rule is that it shows immediately why you cannot
calculate an inverse for some matrices. These are the matrices whose determinant is zero.

Example 26 Using the rule to find the inverse of a 2 × 2 matrix
Find the inverse of the following matrices.

   
2 2 2 3
a A =   b B =  
3 4 2 3
1 Write down the matrix and use the rule a  
2 2
a b A =  
det(A) = =a×d−b×c 3 4

c d
to evaluate the determinant. 2 2
det(A) = =2×4−3×2=2
3 4   
1  d −b 1  4 −2
Use the rule  ∴A =
−1   =  
 det(A) −c a  2 −3 2 
1  d −b
A =
−1    
det(A) −c a   2 −1
=   
to evaluate A−1 . −1.5 1
2 Write down the matrix and use the b  
2 3
a b B =  
rule det(B) = = 2 3
c d
a × d − b × c. 2 3
∴ det(B) = =2×3−2×3=0
to evaluate the determinant. 2 3
det(B) = 0
∴ B does not have an inverse.
Using a CAS calculator to determine the determinant and

inverse of an n × n matrix
There are rules for finding the inverse of a square matrix of any size, but in practice we tend
to use a calculator. The same goes for calculating determinants, although the inverse and
determinant of a 2 × 2 matrix are often computed more quickly by hand.
CAS 4: How to find the determinant and inverse of a matrix using the
TI-Nspire CAS
 
1 2 3
 
If A = 4 1 0, find det(A) and A−1 .
 
2 0 2

Steps
1 Press
2 ctrl + N . Select Add Calculator.
3 Enter the matrix A into your calculator.
4 To calculate det(A), type det(a) and press · to
evaluate.
Note: det() can also be accessed using b>Matrix &
Vector>Determinant.
4 To calculate the inverse matrix

A−1 type a ∧ − 1 and press · to
evaluate. If you want to see the answer
in fractional form, enter as exact (a∧
–1) and press · to evaluate.
Note:
1 Long strings of decimals can be avoided by asking for an
exact inverse. Type in exact( a−1 ).
2 If the matrix has no inverse, the calculator will respond
with the error message Singular matrix.
CAS 4: How to find the determinant and inverse of a matrix using the
ClassPad
 
1 2 3
 
If A = 4 1 0, find det(A) and A−1 .
 
2 0 2
Steps
1 Enter the matrix A into your calculator.
Note: Change the status of the calculator to Standard for fractions
to be displayed. Tapping on Decimal will change the calculator to
Standard.
2 To calculate det(A):
a type and highlight A (by swiping with the stylus)
b select Interactive from the menu bar, tap Matrix-
Calculation, then tap det.
3 To calculate the inverse matrix A−1 :
a type A∧ –1
b press to evaluate.
Note: If the matrix has no inverse, the calculator will respond with
the message Undefined.

Solving matrix equations

In the following example the solution of simple matrix equations is illustrated. The
techniques are analagous to those we use in solving linear equations while noting that there
are some important differences such as multiplication of matrices is not commutative and to
take care that operations are defined on the matrices involved.
Example 27
   
1 2 1       2
  2 1 8 −1 2  
Let A = 2 1 3 B =    C =    D =   and E = 3
3 2 4 6 3
     
1 1 1 1
Solve each of the following matrix equations for X
a B+X =C b BX = C c XB = C
 
4
d BX = D e AX = E f BX +   = D
5
   
2 1 8 −1
a 1 Write down B and C. B =   C =  
3 2 4 6
  
2 Form the equation B + X = C and B+X =C

note that X must be a 2 × 2 matrix. 
2 1
 
8 −1

 + X = 
3 Calculate X = C − B.  
3 2 4 6
  
   
8 −1 2 1
X =    − 
  
4 6 3 2

 
6 −2
=  
1 4


b 1 Form the equation BX = C and note BX = C

that X must be a 2 × 2 matrix. B BX = B−1C
−1
2 Note that B−1 B = I, the 2 × 2 X = B−1C

identity matrix. Calculate X = B C. −1
  
 2 −1 8 −1
X =    
−3 2 4 6
 
 12 −8
∴ X =  
−16 15
c 1 Form the equation XB = C and note XB = C
that X must be a 2 × 2 matrix. XBB−1 = CB−1
2 Note that BB−1 = I, the 2 × 2 X = CB−1
identity matrix. Calculate X = CB−1 .   
8 −1  2 −1
X =    
4 6 −3 2
 
 
 12 −8
∴ X =  
−16 15
d 1 Form the equation BX = D and note BX = D
that X must be a 2 × 1 matrix. B−1 BX = B−1 D
2 Calculate X = B D. −1
X = B−1 D
  
 2 −1 2
X =    
−3 2 3
 
 
1
∴ X =  
0
e 1 Form the equation AX = E and note AX = E
that X must be a 3 × 1 matrix. A AX = A−1 E
−1
2 Calculate X = A−1 E. X = A−1 E

  
−2 −1 5  2
   
X =  1 0 −1 3
   
1 1 −3 1
 
−2
 
∴ X =  1 
 
2

520 Chapter 10 Matrices 10E
   
4 4
f 1 Form the equation BX +   = D and BX +   = D
5 5
note that X must be a 2 × 1matrix.  
4
 4 BX = D −  
2 Calculate X = B−1 D −  . 5
5   
 4
X = B D −  
−1  
5
  
 2 −1 −2
X =    
−3 2 −2
 
 
−2
∴X =  
2
Skill-
sheet Exercise 10E
Review of the properties of the identity matrix

1 a Write down the:
i 2 × 2 identity matrix ii 3 × 3 identity matrix iii 4 × 4 identity matrix.
 
1 2
b If A =  , show that AI = IA = A.
0 3
 
1 2 0

c If C = 3 1 0, show that CI = IC = C.

 
0 1 2

Demonstrating that one matrix is the inverse of the other

Example 24 2 Show that each of the following pairs of matrices are inverses by multiplying one by
the other. Use a calculator if you wish.
       
1 1  2 −1 2 2  2 −1
a   and   b   and  
1 2 −1 1 3 4 −1.5 1
       
9 7 −3 7  0 2  0 0.5
c   and   d   and  
4 3 4 −9 2 0 0.5 0
Calculating the determinant of a matrix

Example 25 3 Determine (by hand) the value of the determinant for each of the following matrices.
       
1 2 0 3 1 2 −1 2
a A =   b B =   c C =   d D =  
0 3 1 4 2 4 2 4
   

10E 10E Matrix inverse, the determinant and matrix equations 521
Calculating the inverse of a matrix

Example 26 4 Use a calculator to determine the inverse of each of the following matrices.
   
    1 1 0 2 1 0 
1.1 2.2 0.2 −0.1
a A =  b B =   c D = 1 1 0 d E = 0 1 −1
   

0 3.0 10 4    
0 1 1 0 0 1
   
Solving matrix equations

   
2 1 3       5
  3 2 −6 −1 3  
Example 27 5 Let A = 1 2 1 B =   C =   D =   and E = 7
4 3 5 6 4
     
2 2 2 9
Solve each of the following matrix equations for X
a B+X =C b BX = C c XB = C
 
7
d BX = D e AX = E f BX +   = D
6
   
1 2  5 6
6 Find the 2 × 2 matrix A such that A    = 
  .
3 4 12 14

7 Suppose that A, B, C and X are 2 × 2 matrices and that both A and B have inverses.
Solve the following for X:
a AX = C b ABX = C c AXB = C
d A(X + B) = C e AX + B = C f XA + B = A
    
3 2 1  x 15 000
     
8 If 2 2 1  y = 20 000 find x, y and z
     
1 1 1 z 10 000
9 Three crop sprays are manufactured by combining chemicals A, B and C as follows:
Spray P One barrel of spray P contains 1 unit of A, 3 units of B and 4 units of C.

Spray Q One barrel of spray Q contains 3 units each of A, B and C.
Spray R One barrel of spray R contains 2 units of A and 5 units of B.
To control a certain crop disease, a farmer requires 6 units of chemical A, 10 units of

chemical B and 6 units of chemical C. How much of each type of spray should the
farmer use?
10 A factory makes and assembles three products, P, Q and R, each requiring different
quantities of three components, a, b and c. The following matrix A represents the
required quantities of components for each product, and the matrix K represents the
daily production of components at the factory.

522 Chapter 10 Matrices 10E
P Q R
   
a  5 3 2  a 95
A = b  2 K = b 80
   
2 4  and
   
c 0 2 3 c 40
a Find the inverse of A.
b Assume that the factory uses all components that are produced. Find the rate of
assembly of P, Q and R at the factory, expressed as number of products per day.
11 Bronwyn and Noel have a clothing warehouse

Brad Flynn Lina
in Summerville. They are supplied by three  
Dresses  5
 6 10 
contractors: Brad, Flynn and Lina.  
Pants  3 4 5 
The matrix shows the number of dresses, pants  

and shirts that one worker, for each of the Shirts 2 6 5
contractors, can produce in a week.
The number produced varies because of the different equipment used by the
contractors.
The warehouse requires 310 dresses, 175 pants and 175 shirts in a week. How many
workers should each contractor employ to meet the requirement exactly?

 
 2 2
12 The determinant of the matrix   is
−1 1

A 4 B 0 C −4 D 1 E 2
 
1 −1
13 The inverse of the matrix   is
1 −2

       
 2 1  1 1   1 1 2 −1
A −1 B   C   D   E  
−1 −1 −1 −2 −1 2 1 −1
    
14 Consider the matrix equation

   
4 2 14 12
3 ×   + X =  
6 7 18 22

Which one of the following is the matrix X?

         
4 2 2 5 2 5 2 6 4 2
A   
 B   
 C    D    E   
6 7 6 7 0 2 0 1 6 7
  

10F Binary, permutation and communication matrices 523
10F Binary, permutation and communication

matrices
Learning intentions
I To be able to identify and work with binary matrices, permutation matrices and
communication matrices.
Binary matrices
The following matrices are examples of binary matrices.
 
    0 1 0
1 1 0  
1 0 1     0 0 1
 
1 0 1
   

1 0 0

Binary matrices are at the heart of many practical matrix applications, including analysing
communication systems and using the concept of dominance to rank players in sporting
competitions.
Permutation matrices
A permutation1 matrix is a square binary matrix in which there is only one ‘1’ in each row
and column.
The following matrices are examples of permutation matrices.
   
    0 1 0 1 0 0
0 1 1 0 
 
 
 
    1 0 0 0 0 1
1 0 0 1    
0 0 1 0 1 0
An identity matrix is a special permutation matrix. A permutation matrix can be used to
rearrange the elements in another matrix.
Example 28 Applying a permutation matrix

   
T  0 0 1
   
X is the column matrix X =  A. P is the permutation matrix P = 0 1 0.
   
R 1 0 0
a Show that:
 
 R 
 
i pre-multiplying2 X by P changes the matrix X to the matrix Y =  A 
 
T
 
ii pre-multiplying X by P2 leaves the matrix X unchanged.
b What can be deduced about P2 from the result in a ii?
1 The word ‘permutation’ means a rearrangement a group of objects, in this case the elements of a matrix,
into a different order.
ISBN 978-1-009-11041-9
2 When we form the matrix product A × © B,
Peter
we Jones et we
say that al 2023
are pre-multiplying by A. Cambridge University Press
a i Form the matrix product PX. To     
find the first entry in the resulting 0 0 1 T   R 
     
column matrix note that the 1 of the PX = 0 1 0  A =  A
     
permutation matrix is in the third 1 0 0 R T
column. of the first row. We obtain
0×T +0×A+1×R=R
To find the second entry in the  2    
0 0 1 T  T 
resulting column matrix note that the      
P2 X = 0 1 0  A =  A
1 of the permutation matrix is in the      
second column of the second row and 1 0 0 R R
so on.
ii Form the matrix product P2 X.
b To leave the matrix X unchanged, 
1 0 0

P2 must be an identity matrix.  
P2 is the identity matrix I = 0 1 0.
 
0 0 1
Example 29
   
 A  D
   
 B  C 
Find a permutation matrix that takes the column matrix  to the column matrix  
   
C   B 
   
D A
   
We want to move D from the fourth place     
0 0 0 1  A  D
to the first place. Place the 1 in the fourth      
0 0 1 0  B  C 
column of the first row.
   =  .


We want to move C from the third place to 0 1 0 0 C   B 
     
the second place. Place the 1 in the third 1 0 0 0 D A
column of the second row.
We want to move B from the second place

to the third place. Place the 1 in the second
column of the third row.
We want to move A from the first place
to the fourth place. Place the 1 in the first
column of the fourth row.

Inverses of permutation matrices

Every permutation matrix has an inverse and this inverse is the adjoint of the matrix. That is
if P is a permutation matrix then P−1 exists and
P−1 = PT
Example 30
 
0 0 1 0
 
0 0 0 1
A 4 × 4 permutation matrix P =  applied to a 4 × 1 column matrix A gives
 
0 1 0 0
 
1 0 0 0
 
 
 X 
 
 Y 
the column matrix  . Determine the matrix A.
 Z 
 
W
 
Solution
 
 X 
 
 Y 
PA =  
 
 Z 
 
W
Using your calculator.
     
 X   X  W 
     
 Y   Y   Z 
∴ A = P−1   = PT   =  
     
 Z   Z   X 
     
W W Y
Communication matrices
A communication matrix is a square binary matrix in which the 1s represent the links in a
communication system.

Example 31 Constructing a communication matrix
Eva, Wong, Yumi and Kim are students who are staying in a E
backpacker’s hostel. Because they speak different languages they
can have problems communicating. The situation they have to deal
W
with is that:
Eva speaks English only
Yumi speaks Japanese only K
Y
Kim speaks Korean only
Wong speaks English, Japanese and Korean.
Conveniently, this information can be summarised in a network diagram, as shown above.

In this diagram, the arrow linking Eva and Wong indicates that they can communicate
directly because they both speak English.
The task is to construct a communication matrix.
1 There are four people so a 4 × 4 matrix is needed.
Label the columns and rows E, W, Y and K. Receiver
2 Label the rows ‘Speaker’ and the columns ‘Receiver’. E W Y K
 
3 Designate each element as a ‘1’ or ‘0’ E  0 1 0 0 
 
according to the following rules: W  1 0 1 1 
S peaker

the element = 1 if two people can communicate  
Y  0 1 0 0 
directly because they speak the same language  
K 0 1 0 0
the element = 0 if two people cannot communi-
cate directly because they do not have a common
language.
The completed matrix is shown opposite.
There is little point in having a matrix representation of a communication system if we

already have a network diagram. However, several questions that are not so easily solved
with a network diagram can be answered using a communication matrix.
For example, we can see from the network diagram that Eva, who speaks only English,
cannot communicate directly with Yumi, who speaks only Japanese. We call this a one-step
communication link.
However, Eva can communicate with Yumi by sending a message via Wong, who speaks
both English and Japanese. In the language of communication systems we call this a
two-step communication link.
The power of the matrix representation is that squaring the communication matrix generates
a matrix that identifies all possible two-step communication links in a communication

network. For example, if we call the communication matrix C, we have:

E W Y K
     
 0 1 0 0   0 1 0 0   1 0 1 1  E
     
 1 0 1 1   1 0 1 1   0 3 0 0  W
C 2 =   ×   = 
 
 0 1 0 0   0 1 0 0   1 0 1 1  Y
     
0 1 0 0 0 1 0 0 1 0 1 1 K

We have the two-step link: Eva → Wong → Yumi.

That is, Eva can communicate with Yumi by using Wong as a go-between.
However, if we use the same process to help us interpret the ‘1’ in row E, column E, we will
find that it represents the two-step link Eva → Wong → Eva. This is not a very useful thing
to know. Two-step links that have the same sender and receiver are said to be redundant
communication links because they do not contribute to the communication between
different people.
Redundant communication links

A communication link is said to be redundant if the sender and the receiver are the same
person.
All of the non-zero elements in the leading diagonal of a communication matrix, or its
powers, represent redundant links in the matrix.
However, all of the remaining non-zero elements represent meaningful two-step
communication links.
For example, the 1 in row Y, column K E W Y K
represents the two-step communication link that 
 1 0 1

1  E
enables Yumi to send a message to Kim. 
 0 3 0 
0  W
C 2 = 
 
 1 0 1 1  Y
 
1 0 1 1 K

Finally, the total number of one and two-step links in a communication system, T , can be
found by evaluating T = C + C 2 .
E W Y K
     
 0 1 0 0   1 0 1 1   1 1 1 1  E
     
 1 0 1 1   0 3 0 0   1 3 1 1  W
T = C + C 2 =   +   = 
 
 0 1 0 0   1 0 1 1   1 1 1 1  Y
     
0 1 0 0 1 0 1 1 1 1 1 1 K


528 Chapter 10 Matrices 10F
Analysing communication matrices

A communication matrix (C) is a square binary matrix in which the 1s are used to
identify the direct (one-step) links in the communication system.
The number of two-step links in a communication system can be identified by
squaring its communication matrix.
The total number of one and two-step links in a communication system can be found
by evaluating the matrix sum T = C + C 2 .
These statements can be readily generalised to include the determination of three
(or more) step links by evaluating the matrices C 3, C 4, etc. However, unless the
communication networks are extremely large, most of the multi-step links identified will
be redundant.
Note: In all cases, the diagonal elements of a communication matrix (or its power) represent redundant
communication links.
Skill-
sheet Exercise 10F
Permutation matrices
1 Which of the following binary matrices are permutation matrices?
   
  0 1 0 0 1
1 0    
A =    B = 1 0 0 C = 1 0
1 1
    
0 0 1 0 1

Example 28 2 X is the row matrix: X = S H U T
 
 0 0 0 1 
 
 1 0 0 0 
P is the permutation matrix: P = 
 
 0 1 0 0 
 
0 0 1 0

a What does matrix X change to if it is post-multiplied by P?

b For what value of n does XPn first equal X?
   
 A  C 
   
 B  D
3 Find a permutation matrix that takes the column matrix   to the column matrix  
   
Example 29
C   A 
   
D B

10F 10F Binary, permutation and communication matrices 529
Inverses of permutation matrices

 
1 0 0 0
 
0 0 0 1
4 A 4 × 4 permutation matrix P =  applied to a 4 × 1 column matrix A gives
 
Example 30
0 0 1 0
 
0 1 0 0
 
 
 X 
 
 Y 
the column matrix  . Determine the matrix A.
 Z 
 
W
 
Communication matrices
Example 31 5 Freya (F), Lani (L) and Mei (M) are close friends who regularly L
send each other messages. The direct (one-step) communication
links between the friends are shown in the diagram opposite. M
a Construct a communication matrix C from this diagram. F
b Calculate C 2 .
c How many different ways can Mei send a message to Freya?
6 Four fire towers T 1, T 2, T 3 and T 4, can communicate T2
with one another as shown in the diagram opposite. In T4
this diagram an arrow indicates that a direct channel of
communication exists between a pair of fire towers. T3
T1
For example, a person at tower 1 can directly communicate with a person in tower 2
and vice versa.
The communication matrix C can also T1 T2 T3 T4
be used to represent this information.  
 0 1 0 0  T1
a Explain the meaning of a zero in the  
C =  1 0 0  T2
communication matrix.  
 0 1 0 1  T3
b Which two towers can  
communicate directly with T 2? 0 1 0 T4
c Write down the values of the two missing elements in the matrix.
The matrix C 2 is shown opposite. T1 T2 T3 T4
d Explain the meaning of the 1 in row 
 1 0 1 0 

T1
T 3, column T 1.  
C =  0
2 2 0 1  T2

e How many of the two-step  
 1 0 2 0  T3
communication links shown in the  
matrix C 2 are redundant? 0 1 0 1 T4
f Construct a matrix that shows the total number of one and two-step communication
links between each pair of towers.
g Which of the four towers need a three-step link to communicate with each other?

530 Chapter 10 Matrices 10F
Representing a large network diagram by a matrix

7 Construct a 5 × 5 matrix to represent the B
C
communication network diagram opposite. A
E D

8 The matrix below shows how five people, Adam(A), Bertie(B), Catherine(C), David(D)
and Evan (E), can communicate with each other.
Receiver
 A B C D E
 
A  0 1 0 0 1
 
B  1 0 0 0 1
 
S ender C  1 0 0 0 1
 
D  1 0 1 0 1
 
E  1 0 1 1 0
A ‘1’ in the matrix shows that the person named in that row can send a message
directly to the person named in that column. Adam wants to send a message to David.
This can be done through a sequence of communications formed from the five people.
Which of the following is a possible sequence of communications to get the message
from Adam to David?
A A, D B A, B, D C A, B, C, D D A, B, E, D E A, E, C, B, D
9 Matrix P is a 3 × 3 permutation matrix. Matrix Z is another matrix such that the matrix
product P × Z is defined. This matrix product results in the third row becoming the first
row, the second row becoming the third and the first row becoming the second row of
matrix Z. The permutation matrix P is
     
0 0 1 1 0 0 1 0 0
     
A 1 0 0 B 1 0 0 C 1 0 0
     
0 1 0 0 1 0 1 0 0
     
   
0 1 0 1 0 0
   
D 0 0 1 E 0 1 0
   
1 0 0 0 0 1
   

10G Dominance matrices 531
10G Dominance matrices
Learning intentions
I To be able to construct and interpret dominance matrices.
In many group situations, certain individuals are said to be dominant. This is particularly
true in sporting competitions. Problems of identifying dominant individuals in a group can
be analysed using the same approach we used to analyse communication networks.
For example, five players – Anna, Birgit, Cas, Di and Emma – played in a round-robin
tournament3 of tennis to see who was the dominant (best) player.
The results were as follows: B
Anna defeated Cas and Di C
Birgit defeated Anna, Cas and Emma

Cas defeated Di
Di defeated Birgit D
A
Emma defeated Anna, Cas and Di.
We can use a network diagram to display the results

graphically, as shown opposite. In this diagram, the E
arrow from B to A tells us that, when they played,
Birgit defeated Anna.
Both Birgit and Emma had three wins each so there is a tie. How can we resolve this
situation and see who is the best player? One way of doing this is to calculate a dominance
score for each player. We do this by a constructing a series of dominance matrices.
One-step dominances
The first dominance matrix, D, A B C D E Dominance
records the number of one-step  
A  0 0 1 1 0 2
dominances between the players.  
B 
 1 0 1 0 1 3
For example, Anna has a one-step 
D= C

 0 0 0 1 0 1
dominance over Cas because,  
D  0 1 0 0 0 1
when they played, Anna beat Cas.  
E 1 0 1 1 0 3
This matrix can be used to calculate a one-step dominance score for each player, by
summing each of the rows of the matrix. According to this analysis, B and E are equally
dominant with a dominance score of 3.
3 A round-robin tournament is one in which each of the participants play each other once.

Now let us take into account two-step dominances between players.
Two-step dominances
A two-step dominance occurs when a player beats another player who has beaten someone
else. For example, Birgit has a two-step dominance over Di because Birgit defeated Cas who
defeated Di.
Two-step dominances can be A B C D E Dominance
determined using the same  
A  0 1 0 1 0 2
technique used to obtain two-  
B  1 0 2 3 0 6
step links in a communication  
network. We simply square the D2 = C

 0 1 0 0 0 1
 
one-step dominance matrix. D  1 0 1 0 1 3
 
The two-step dominances for E 0 1 1 2 0 4
these players are shown in matrix D2 .
Reading across the ‘B row’.
The 1 in column A represents the two-step dominance ‘Birgit defeated Emma who
defeated Anna’.
The 2 in column C represents the two-step dominances ‘Birgit defeated Emma who
defeated Cas’ and ‘Birgit defeated Anna who defeated Cas’
The 3 in column D represents the three two-step dominances ‘Birgit defeated Emma
who defeated Di’, ‘Birgit defeated Anna who defeated Di’ and ‘Birgit defeated Cas who
defeated Di’.
In column E the 0 tells us that there are no two-step dominances for Birgit over Emma
even though there was a one-step dominance.
We can combine the A B C D E T otal
information contained in both  
A  0 1 1 2 0 4
D and D2 by calculating a  
B 2 0 3 3 1 9
new matrix T = D + D2 .

 
T = D+D = 2
C

 0 1 0 1 0 2
 
D  1 1 1 0 1 4
Using these total dominance scores:  
E 1 1 2 3 0 7
Birgit is the top-ranked player with a total
dominance score of 9
Eva is second with a total score of 7
Anne and Di are equal third with a total score of 4
Cas is the bottom-ranked player with a total score of 2.

10G Dominance matrices 533
Example 32 Determining dominance
Four people, A, B, C and D, have been asked to B

form a committee to decide on the location of a
new toxic waste dump.
C
From previous experience, it is known that: A
A influences the decisions of B and D

B influences the decisions of C D
C influences the decisions of no one
D influences the decisions of C and B.
a Use the graph to construct a dominance matrix that takes into account both one-step
and two-step dominances.
b From this matrix, determine who is the most influential person on the committee.
a Construct the one-step dominance A B C D One-step
matrix D.  
A  0 1 0 1  2
 
D= B  0 0 1 0  1
 
C  0 0 0 0  0
 
D 0 1 1 0 2
Construct the two-step A B C D T wo-step

dominance matrix D2 .  
A  0 1 2 0  3
 
D2 = B  0 0 0 0  0
 
C  0 0 0 0  0
 
D 0 0 1 0 1
Form the sum T = D + D2 . A B C D T otal

 
A  0 2 2 1  5
 
T = D + D2 = B  0 0 1 0  1
 
C  0 0 0 0  0
 
D 0 1 2 0 3
b The person with the highest total Person A is the most influential
dominance score is the most influential. person with a total dominance score
of 5.

534 Chapter 10 Matrices 10G
Exercise 10G
Dominance matrices
Example 32 1 The results of a competition between teams A, B, C and D are B

displayed opposite. An arrow from D to C indicates that team
D defeated team C.
A C
a Construct a dominance matrix showing one-step
dominance between the teams. Rank the teams according
to one-step dominances.
b Construct a dominance matrix showing two-step
dominances between the teams. Rank the teams, taking
D
into account both one-step and two-step dominances.
2 Five students play each other at chess. The dominance matrix shows the winner of each
game with a ‘1’ and the loser or no match with a ‘0’. For example, row 2 indicates that
B loses to A, D and E but beats C.
a Find the one-step dominance Losers
score for each student and use
A B C D E
these to rank them.  
A  0
 1 1 1 1 
b Calculate the two-step dominance 
 
B  0 0 1 0 0 
matrix. Winners  
C  0 0 0 0 0
c Determine the matrix T = D + D 2
 
 

and use this matrix to rank the D  0  1 1 0 0 
 
players. E 0 1 0 1 0
3 Five friends – Ann, Bea, Cat, Deb and Eve – competed in a A

round-robin tennis tournament. The results were as follows:
Ann defeated Cat and Deb
E B
Bea defeated Ann, Cat and Eve
Cat defeated Deb
Deb defeated Bea D C
Eve defeated Ann, Cat and Deb.
Using this information:
a Construct a one-step dominance matrix, D.
b Construct a two-step dominance matrix, D2 .
c Use the dominance scores from the matrix D + D2 to rank the five players.

10G 10G Dominance matrices 535
4 The following dominance matrix, M, gives the results of a series of squash matches
between five friends, where mi j = 1 if player i beat player j.
Ash Ben Carl Dot Elle
 
Ash 
 0 0 1 1 0 

Ben

 1 0 1 1 0 
M = Carl
 
 0 0 0 1 0 
 
Dot 
 0 0 0 0 1 

Elle 1 1 1 0 0

a How many matches were there?

b Describe the outcomes of the matches.
c Use the dominance scores from the matrix M to give a ranking of the players.
5 Five chess players – A, B, C, D and E – competed in a round-robin chess tournament.

The results were as follows:
A defeated B and D B defeated C and E C defeated A and D
D defeated B E defeated A, C and D.

a Construct a one-step dominance matrix, M.
b Use the dominance scores from the matrix M + M 2 to rank the players.
6 A committee of four people – A, B, C and D – will decide B C

on the location of a new toxic waste dump. From previous
experience, it is known that:
A influences the decisions of B and D
B influences the decisions of C A D
C influences the decisions of no one
D influences the decisions of B and C.

a Construct a matrix that takes into account both one-step and two-step dominance.
b From this matrix, determine who is the most influential person on the committee.
7 The following table gives the results of the first round of games at a chess club.
Game A vs B C vs D A vs D B vs C B vs D A vs C
Winner A C D B B A
a Create a one-step dominance matrix, D.

b Create a ranking of the four players using D and D2 .

536 Chapter 10 Matrices 10G
8 Four schools – A, B, C and D – compete in a round-robin hockey tournament. The

results are summarised by the following dominance matrix:
A B C D
 
A  0 1 1 0 
 
B  0 0 1 1 
M=  
C  0 0 0 1 

 
D 1 0 0 0
Create an ordering of the four schools using M, M 2 and M 3 .
9 Five teams competed in a football tournament. Each team A

played every other team once. The results are shown in the
diagram. (For example, the arrow pointing from A to C
E B
indicates that A defeated C.)
Create a rank order for the five teams by taking into
consideration both one-step and two-step dominance.
D C

10 Four teams, A, B, C and D, competed in a round-robin competition where each team
played each of the other teams once. There were no draws. The results are shown in the
matrix below.
loser
A B C D
A  0 x y 0 
 
 
B  0 0 1 1 
winner  
C  1 0 0 0 
 
D1 z 1 0 
The values of x, y and z are

A x = 0, y = 0, z = 0 B x = 0, y = 1, z = q C x = 1, y = 0, z = 0
D x = 1, y = 0, z = 1 E x = 1, y = 1, z = 1
11 There are five hens in a coup. Their owner calls them Alpha, Beta, Gamma, Delta and
Epsilon. There is a pecking order in the coop, and the following dominance matrix, M,
was formed by the owner:

10G 10G Dominance matrices 537
Alpha Beta Gamma Delta Epsilon

 
Alpha 
 0 0 0 1 0 

Beta

 1 0 1 0 1 

M = Gamma
 
 1 0 0 0 0 
 
Delta 
 0 1 1 0 0 

Epsilon 1 0 1 1 0
 
Based on the matrix M + M 2 , which of the following best describes the pecking order
in the coop?
A Beta, Epsilon, Delta, Alpha, Gamma B Beta, Epsilon, Gamma, Delta, Alpha
C Beta, Epsilon, Delta, Gamma, Alpha D Epsilon, Beta, Delta, Alpha, Gamma
E Epsilon, Beta, Delta, Gamma, Alpha
12 Four soccer teams, X, Y, Z and W, compete in a round-robin competition. In each

game, there is a winner and a loser. To decide the winner of the tournament, the sum of
the one-step dominance matrix, D, and the two-step dominance matrix, D2 , is found.
This sum is
 
1 2 1 2
 
1 1 2 2
D + D2 = 
 
2 2 1 1
 
2 1 2 1
 
Which one of the following is the correct one-step dominance for this tournament?
     
0 0 1 1 1 0 1 1
 
0 0 1 1
 
    
1 0 0 0 1 0 0 0 1 0 1 0
A   B   C  
0 1 0 1 0 1 0 1 0 1 0 1
     
0 1 0 0 0 1 0 1 0 1 0 1
     
   
0 0 1 1 1 0 1 1
   
1 0 1 0 1 0 0 1
D  E 
   
0 1 0 1 0 1 0 1
   
1 1 0 0 0 1 0 0
   

Review 538 Chapter 10 Matrices
Matrix A matrix is a rectangular array of numbers or symbols (elements)

Assign- enclosed in brackets (plural: matrices).
ment
Row matrix A row matrix contains a single row of elements.
Column matrix A column matrix contains a single column of elements.
Transpose The transpose of a matrix is obtained by interchanging its rows and

columns.
Square matrix A square matrix has an equal number of rows and columns.
Zero matrix A zero (null) matrix, O, contains only zeros.
Order The order (or size) of a matrix is given by the number of rows and
columns in that order.
Locating an The location of each element in the matrix is specified by its row and
element column number in that order.
Equal matrices Matrices are equal when they have the same order and corresponding
elements are equal in value.
Adding and Two matrices of the same order can be added or subtracted, by adding
subtracting or subtracting corresponding elements.
matrices
Scalar Multiplying a matrix by a number (scalar multiplication) multiplies

multiplication every element in the matrix by that number.
Matrix Matrix multiplication is a process of multiplying rows by columns.

multiplication To multiply a row matrix by a column matrix, each element in the row
matrix is multiplied by each element in the column matrix and the
results added. For example:
 
4
h i  
1 0 3 2 = [1 × 4 + 0 × 2 + 3 × 5] = [19]
 
5
Power of a matrix The power of a matrix is defined in the same way as the powers of
numbers: A2 = A × A, A3 = A × A × A, and so on.
Only square matrices can be raised to a power.
A0 is defined to be I, the identity matrix.

Review
Identity matrix An identity matrix, I, is a square matrix with 1s down the leading
diagonal and zeros elsewhere.
Determinant The determinant of a matrix, A, is written as det(A).

Only square matrices have determinants.
 
1 3
If A =   then det(A) = 1 3 = 1 × 4 − 3 × 3 = −5
3 4 3 4

For higher order matrices, a calculator is used to calculate the
determinant.
Inverse The inverse of a matrix, A, is written as A−1 and has the property that
AA−1 = A−1 A = I.
Only square matrices have inverses.
The inverse of a matrix is not defined if det(A) = 0.
A calculator is used to determine the inverse of a matrix.
Binary matrix A binary matrix is a matrix whose elements are either zeros or ones.
Permutation A permutation matrix is a square binary matrix in which there is only

matrix a single 1 in each row and column.
Communication A communication matrix is a square binary matrix in which the 1s

matrix represent direct (one-step) communication links.
Redundant A communication link is said to be redundant if the sender and the

communication receiver are the same people.
link
Round-robin A round-robin tournament is one in which each of the participants

tournament plays every other competitor once.
Dominance A dominance matrix is a square binary matrix in which the 1s

matrix represent one-step dominances between the members of a group.

Skills checklist
list
10A 1 I can state the order of a given matrix.
10A 2 I can find the transpose of a matrix.
10A 3 I can identify a matrix as being a square, column or row matrix.
10A 4 I can identify a square matrix as being an identity or diagonal or symmetric or

upper/lower triangular matrix.
10A 5 I can identify the aij term in a matrix A.
10A 6 I can construct a matrix given a rule for aij .
10A 7 I can enter a matrix into a CAS calculator.
See CAS 1, and Exercise 10A Question 13
10B 8 I can enter information from a table into a matrix.
10B 9 I can represent a network diagram by a matrix.
10B 10 I can interpret a matrix representing a network diagram.
10C 11 I can recognise when two matrices are equal and use this to solve problems.
10C 12 I can add two matrices of the same order together.

Review
10C 13 I can subtract one matrix from another when they have the same order.
10C 14 I can multiply a matrix by a scalar.
10C 15 I recognise the role of the zero matrix and can undertake operations using the
zero matrix.
10C 16 I can use addition, subtraction and scalar multiplication to process data.
10D 17 I can determine if the product of two given matrices is defined.
10D 18 II can determine the order of a matrix product.
10D 19 I can multiply a row matrix by a column matrix by hand.
10D 20 I can multiply a rectangular matrix by a column matrix by hand.
10D 21 I can use summing matrices to sum the rows or columns of a matrix.
10D 22 I can undertake multiplications of matrices to solve practical problems.
10D 23 I can evaluate matrix expressions involving powers.
10E 24 I can recognise that two matrices are inverses if their product is the identity
matrix.
10E 25 I can evaluate the determinant of a 2 × 2 matrix by hand.
10E 26 I can calculate the inverse of a 2 × 2 matrix by hand.

10E 27 I can calculate the inverse and determinant of a n × n matrix using CAS.
See CAS 4, and Exercise 10E Question 4
10E 28 I can solve simple matrix equations.
10F 29 I can use a permutation matrix to rearrange the elements of a column or row
matrix.
10F 30 I can construct a permutation matrix to rearrange the elements of a column or

row matrix in a given order.
10F 31 I can find the inverse of a permutation matrix.
10F 32 I can construct a communication matrix from information given in written form
or a diagram.
10F 33 I can construct a dominance matrix from information given in written form or a
diagram.
Multiple choice questions
The following matrices are needed for Questions 1 to 8.

 
        1
2 0 −1 3 1 −1  0 1  
U =   V =   W = 1 −1 X =   Y =   Z = 0
1 1 1 1 0 1 −1 2
    
1
1 The row matrix is:

A U B V C W D X E Z
2 The square matrices are:

A U and V B X and Y C Y and W D U and Y E U, V and X
3 The order of matrix X is:

A 2×2 B 2×3 C 3×2 D 3×3 E 6

Review
4 The following matrices can be added:
A U and V B V and W C X and Y D U and Y E none of the above
5 The following matrix product is not defined:

A WV B XZ C YV D XY E UY
6 −2Y =
         
0 −2  0 2  0 −2  0 1 −2 −1
A   B   C   D   E  
2 −4 −2 4 −2 4 −1 2 −4 2
 
7 The order of matrix product XZ is:

A 1×3 B 2×1 C 3×1 D 3×2 E 3×3
8 UT =
         
2 0 1 1 2 1 0 2 1 0
A    B    C    D    E   
1 1 2 0 0 1 1 1 1 2
    
 
 2 0 1 
 
9 In the matrix A =  4 −1 3, the element a23 =
 
−5 −4 7
A −4 B −1 C 0 D 3 E 4
   
 2 0 −1 0
10 2   −   =
−1 1 1 1
         
 5 0 5 0  3 0 6 0  5 0
A   B   C   D   E  
−4 0 4 0 −2 0 1 2 −3 1
   
 
3
 

1 2 3 × 2 =
 
11
 
1
 
  1 3
1 2  
A 10 B 12 C   D 2 2 E not defined
4 3
  
3 1
 
The following matrices are needed for Questions 12 to 16.

 
      1 0 2  
2 0 43 0.9 0.5   1 2
U =   V =   W =   X = 3 2 0 Y =  
1 1 45 0.1 0.5 2 4
  
1 1 2
 

12 The matrix that cannot be raised to a power is:

A U B V C W D X E Y
13 det(U) =
A −2 B 0 C 1 D 2 E 4
14 Y −1 =
       
 0.5 0  4 −2 1 0 1 −4 2 
A    B    C    D   E not defined
−0.5 1

−2 1

0 1
 8  2 −1
15 U −1 =
       
 0.5 0  4 −2 1 0 1 −4 2 
A   B   C   D   E not defined
−0.5 1

−2 1 0 1
 8  2 −1
16 The matrix product that is defined is:

A UX B XY C VW D UW E WX
   
1 2 2
17   ×   =
3 4 1
   
     

3  4  2 2
A 18 B 12 C   D   E  
7 10 6 4

18 X is a 3 × 2 matrix. Y is a 2 × 3 matrix. Z is a 2 × 2 matrix. Which of the following

matrix expressions is not defined?
A XY B YX C XZ − 2X D Y X + 2Z E XY − Y X
 
3
 
5
19 A = 1 1 1 1 and B =  .
2
 
4
 
The matrix expression that displays the mean of the numbers 3, 5, 2, 4 is:
1 1 1 1 1
A (A + B) B (A + B) C B D AB E BA
4 2 4 4 4
20 Consider the following four matrix expression.

         
3 1 3 1 4 2 3 4 2
3 1 + 4 2   + 4 2   +     +  
4 2 5 5 6 7 5 6 7
      
How many of these four matrix expressions are defined?

A 0 B 1 C 2 D 3 E 4

Review
21 If both A and B are m × n matrices, where m , n, then A + B is
A an m × n matrix B an m × m matrix C an n × n matrix
D a 2m × 2n matrix E not defined
     
 4 6  3 6 −6 −12
22 The matrix expression   +   +   is equal to
−2 1 −2 1 4 −1
    
   
0 0 1 0
A   B 0 C  
0 0 0 1

   
13 24 1 24
D   E  
0 1 0 3

23 The diagram opposite is to be represented by a matrix, A, where: 1

2
element = 1 if the two points are joined by a line
element = 0 if the two points are not connected. 3
4
The matrix A is:
     
1 0 0 0 1 0 0 0 1
       
2 0 0 1 0 0 1 1 0 1 2
A 1 2 3 4 B   C  D  E 
       
3 0 1 0 1 0 1 1 1 3 4

     
4 1 0 1 0 1 0 1 0
 
 
3 0
 
4 1
24 If matrix M =   then the transpose matrix M T =

7 2
 
9 6

   
0 3 0 3
         
1 4 1 4 3 4 7 9 0 1 2 6 3 4 0 1
A  B  C   D  E 
     
2 7 6 9 0 1 2 6 3 4 7 9
 
2 6 7 9

   
6 9 2 7
   
25 M is a 6 × 6 matrix. N is a 5 × 6 matrix. Which one of the following matrix expressions

is defined?
A N M − 2N B M(MN)−1 C M2 N D NT M E MT N

 
C 
 
 B 
26 Matrix A1 is the 4 × 1 column matrix  
 
 A 
 
D
A second 4 × 1 column matrix, A2 , contains the same elements as A1 , but the elements
are ordered from top to bottom in alphabetical order. Matrix A2 = P × A1 , where P is a
permutation matrix. Matrix P is
         
1 0 0 0 0 0 0 1 0 0 1 0 0 0 0 1 0 0 0 1
         
0 0 0 1 0 1 0 0 0 1 0 0 0 1 1 0 0 1 1 0
A   B   C   D   E  
0 1 0 0 1 0 0 0 1 0 0 0 0 1 1 1 0 1 0 0
         
1 0 0 0 0 0 0 1 0 0 0 1 1 0 1 0 1 0 0 0
         
27 Four teams, X, Y, Z and W, compete in a round-robin competition. In each game, there

is a winner and a loser. The sum of the one-step dominance matrix, D, and the two-step
dominance matrix, D2 , is found. This sum is
 X Y Z W
X  0 1 2 1 
 
 
Y  0 0 1 1 
D + D = 
2 
Z  1 1 0 1 
 
W  1 2 2 0 
In the first two games:

team W defeated team X team Z defeated team W.
Which one of the following is the correct one-step dominance for this tournament?
 X Y Z W  X Y Z W  X Y Z W
X  0 1 1 1  X  0 1 1 0  X  0 1 1 0 
     
     
Y  0 0 1 1  Y  0 0 1 1  Y  0 0 1 0 
A 
 
 B 
 
 C 
 
Z  1 0 0 1  Z  1 0 0 1  Z  0 0 0 1 
     
W  1 0 1 1  W  1 0 1 2  W  1 1 0 0 
X Y Z W X Y Z W
X  0 1 1 1  X  0 1 1 1 
   
 0  
Y 0 1 1  Y  0 0 1 1 
D 
 1  E  
Z  0 0 1  Z  1
 0 0 0 
 1  
W 0 1 1  W  1 0 1 0 
 
1 5 3 2
 
0 1 3 2
28 The matrix   is an example of a

0 0 1 4
 
0 0 0 1


Review
A symmetric matrix. B unit matrix. C triangular matrix.
D diagonal matrix. E communication matrix.
29 A, B, C, D and E are five intersections A B
joined by roads, as shown in the
diagram opposite. Some of these roads
D C
are one-way only.
E
The matrix opposite indicates the
direction that cars can travel along each From intersection
of these roads.
A B C D E
 
In this matrix:  0 0 0 0 0  A
 
the 1 in column A and

 1 0 0 0 0  B
 
row B indicates that cars  0 1 0 1 1  C T o intersection
 
can travel directly from  1 0 0 0 0  D
 
A to B 
0 0 1 1 0 E
the 0 in column B and row A
indicates that cars cannot travel
directly from B to A (either it is a
one-way road or no road exists).
Cars can travel in both directions between intersections:
A A and D B B and C C C and D D D and E E C and E

1 The following diagrams represent the road network joining several towns. Represent
each road by a matrix.
a town 1 town 2 b town 1 town 2
c town 1
town 2
town 3
town 3
town 4
2 Heights in feet and inches can be converted into centimetres using matrix
 
30.45 
multiplication. The matrix C =    can be used as a conversion matrix
2.54

(1 foot = 30.45 cm and 1 inch equals 2.54 cm).
a What is the order of matrix C?
Jodieh tellsius that her height is 5 feet 4 inches. We can write her height as a matrix
J= 5 4.
b What is the order of matrix J?

c Is the matrix product JC defined? Why?

d Evaluate the matrix product JC. Explain why it gives Jodie’s height in centimetres.
 
5 8
e Matrix H =   gives the heights in feet and inches of two other people.
6 1

Use the conversion matrix C and matrix multiplication to generate a matrix that
displays the heights of these two people in centimetres.
3 Books can be classified as fiction or non-fiction and come in either hardback or

paperback form. The table shows the number of book titles carried by two bookshops
in each of the categories.
Bookshop 1 Bookshop 2
Number of titles Hardback Paperback Hardback Paperback
Fiction 334 876 354 987
Non-fiction 213 456 314 586
a How many non-fiction paperback titles does bookshop 1 carry?

 
334 876 
b The matrix A =   displays the number of book titles available at
213 456

bookshop 1 in all categories. What is the order of this matrix?
c Write down a matrix equivalent to matrix A that displays the number of book titles
available at bookshop 2. Call this matrix B.
d Construct a new matrix, C = A + B. What does this matrix represent?
e The average cost of books is $45 for a hardback and $18.50 for a paperback. These
 
45.00 
values are summarised in the matrix E =  .
18.50

i What is the order of matrix E?

ii Construct the matrix product AE and evaluate.
iii What does the product AE represent?
f Bookshop 1 plans to double the number of titles it carries in every category. Write
down a matrix expression that represents this situation and evaluate.
4 Mathematics and Physics are offered in a first year university science course.
 
600 Mathematics
The matrix N =   lists the number of students enrolled in each
320 Physics
subject.
A B C D E
The matrix P = [0.15 0.225 0.275 0.25 0.10] lists the proportion of these
students expected to be awarded an A, B, C, D or E grade in each subject.
a Write down the order of matrix P.

Review
b Let the matrix R = NP.
i Evaluate the matrix R.
ii Explain what the matrix element R13 represents.
c Students enrolled in Mathematics have to pay an extra fee of $220, while students
enrolled in Physics pay an extra fee of $197.
i Write down a clearly labelled row matrix, called F, that lists these fees.
ii Show a matrix calculation that will give the total fees fees, L, paid in dollars by
the students enrolled in Mathematics and Physics. Find this amount.
5 In a simplified game of darts, the possible scores are 25, 50 or 75.

 
25
G is a column matrix that lists the possible scores
 
G = 50
In one game of 15 throws, Daniel achieved  
75
Eight ‘25 scores’
Six ‘50 scores’
One ‘75 score’
a Write a row matrix, N, that shows the number of each score that Daniel had.
b Matrix P is found by multiplying matrix N with matrix G so that P = N × G.
Evaluate matrix P.
c In this context, what does the information in matrix P provide?
6 A mining company operates three mines, A, B and C. Each of the mines produces three
types of minerals, p, q and r. Consider the following two matrices:
p q r
   
A 20 20 40 A  46 000
X = B  0 40 20 Y = B  34 000
   
and
   
C 60 40 60 C 106 000
Matrix X gives the number of tonnes of each of the minerals extracted per day from
each of the mines, and matrix Y gives the total revenue (in dollars) from selling the
minerals extracted from each of the mines on one day.
a Calculate the total number of tonnes of minerals produced by mine A.
b Calculate the total number of tonnes of mineral q produced.
c Calculate the total revenue of the three mines.
d In the matrix equation XA = Y
i What is the order of matrix A?
ii What do the elements of matrix A represent?
iii We know that A = X −1 Y. Find A.

Chapter
11
Transition matrices and
Leslie matrices
Chapter objectives
I How do you construct a transition matrix from a transition diagram and
vice versa?
I How do you construct a transition matrix to model the transitions in a
population?
I How do you use a matrix recurrence relation, S0 = initial state matrix,
Sn+1 = TSn , to generate a sequence of state matrices?
I How do you informally identify the equilibrium state or steady-state matrix
in the case of regular state matrices?
I How do you use a matrix recurrence relation S0 = initial state matrix,
Sn+1 = TSn + B to model systems that include external additions or
reductions at each step of the process?
I How do you use and interpret Leslie matrices to analyse population growth?
In this chapter we use matrices to model proportional change of the numbers in a

particular state to itself and other states from one time to the next.
For example, in the first example of this chapter, there are two states which in this case
are the towns in which rental cars finish up each day.
In sections A-E of this chapter we look at transition matrices which satisfy certain
conditions and describe the proportional change.
In section F of the chapter Leslie matrices are used to model population change in a
particular way.

11A Transition matrices - setting up a transition matrix 551
11A Transition matrices - setting up a transition

matrix
Learning intentions
I To be able to set up a transition matrix from a diagram or written information.
A car rental firm has two branches: one in Bendigo and one in Colac. Cars are usually rented
and returned in the same town. However, a small percentage of cars rented in Bendigo each
week are returned in Colac, and vice versa. The diagram below describes what happens on a
weekly basis.
B – Bendigo
20%
C – Colac
80% B C 90%
10%
What does this diagram tell us?

From week to week:
0.8 (or 80%) of cars rented each week in Bendigo are returned to Bendigo
0.2 (or 20%) of cars rented each week in Bendigo are returned to Colac
0.1 (or 10%) of cars rented each week in Colac are returned to Bendigo
0.9 (or 90%) of cars rented each week in Colac are returned to Colac.
The percentages (written as proportions) are summarised in the form of the matrix below.
Rented in
Bendigo Colac
 
Bendigo  0.8 0.1 
Returned to  
Colac 0.2 0.9
This matrix is an example of a transition matrix (T). It describes the way in which
transitions are made between two states:
state 1: the rental car is based in Bendigo.
state 2: the rental car is based in Colac.
Note: In this situation, where the total number of cars remains constant, the columns in a transitional matrix
will always add to one (100%). For example, if 80% of cars are returned to Bendigo, then 20% must be
returned to Colac.

552 Chapter 11 Transition matrices and Leslie matrices
Example 1 Setting up a transition matrix
The diagram gives the weekly 80%

return rates of rental cars at three
A – Albury
locations: Albury, Wodonga and B – Benalla
Benalla. Construct a transition
W W – Wodonga
15%
matrix that describes the week- 70% 10%
by-week return rates at each of
11%
the three locations. Convert the A 5% B
percentages to proportions. 20%
12% 77%
1 There are three locations from which
Rented in
the cars can be rented and returned:
A W B
Albury (A), Wodonga (W) and Benalla  
A  
(B). To account for all the possibilities,  
Returned to w  
a 3 × 3 matrix is needed. Construct a  
blank matrix labelling the rows and B
columns A, W and B, respectively.
Column labels indicate where the car
was rented. The row labels indicate
where the cars were returned to.
2 Complete the matrix by writing
A W B
each of the percentages (converted  
A  0.7 
to proportions) into the appropriate  
w  0.1 
locations. Start with column A and  
write in values for each row: 0.7 (70%), B 0.2
0.1 (10%) and 0.2 (20%).
3 Mentally check your answer by
A W B
summing columns; they should  
A  0.7 0.05 0.12
sum to 1.  
w  0.1 0.8 0.11
 
B 0.2 0.15 0.77
Example 2 Setting up a transition matrix
A factory has a large number of machines. Machines can be in one of two states:
operating or broken. Broken machines are repaired and come back into operation, and
vice versa. On a given day:
85% of machines that are operational stay operating
15% of machines that are operating break down
11A 11A Transition matrices - setting up a transition matrix 553
5% of machines that are broken are repaired and start operating again
95% of machines that are broken stay broken.
Construct a transition matrix to describe this situation. Use the columns to define the
situation at the ‘Start’ of the day and the rows to describe the situation at the ‘End’ of
the day.
1 There are two machine states: operating (O) or broken
Start
(B). To account for all the possibilities, a 2 × 2 transition
 O B 
matrix is needed. Construct a blank matrix, labelling the O  
rows and columns O and B, respectively. End  
B
2 Complete the matrix by writing each of the percentages

(converted to proportions) into the appropriate locations. O B
 
Start with column O and write in the values for each O 0.85 
 
row: 0.85 (85%) and 0.15 (15%). B 0.15
3 Mentally check your answer by summing the columns;

they should sum to 1. O B
 
O 0.85 0.05
 
B 0.15 0.95
Exercise 11A
Setting up a transition matrix from a transition diagram

Example 1 1 The diagrams below describe a series of transitions between the states indicated.
Construct a transition matrix that can be used to represent each of these diagrams. Use
columns to define the starting points. Convert the percentages to proportions.
a 60% b 30%
40% A B 45% 70% X Y 75%
55% 25%
c 70% d 45%
10% Y 15% 25% B 20%
X 15% 23% Z A 35% 20% C

30% 30%
60% 55% 45% 65%
22% 15%
554 Chapter 11 Transition matrices and Leslie matrices 11A
Example 2 2 A factory has a large number of machines which can be in one of two states, operating
(O) or broken down (B). It is known that that an operating machine breaks down by the
end of the day on 4% of the days, and that 98% of machines which have broken down
are repaired by the end of the day.
Complete the 2 × 2 transition matrix T to describe this.
Today
O B
 
O  
T =   Tomorrow
B

3 A large company has 1640 employees, 60% of whom currently work full-time (F) and
40% of whom currently work part-time (P). Every year 20% of full-time workers move
to part-time work, and 14% of part-time workers move to full-time work.
This year
F P
 
F  
T =   Next year
P


4 In a particular newsagent, the two top-selling newspapers are the Argus and the
Bastion. The transition diagram below shows the way shoppers at this newsagent
change their newspaper choice from today to tomorrow.
70%
30% A B 35%
65%
A transition matrix that provides the same information as the transition diagram is
A Today B Today C Today
A B A B A B
     
A 65% 30% A 30% 65% A 30% 65%
Tomorrow   Tomorrow   Tomorrow  
B 70% 35% B 70% 35% B 35% 70%
D Today E Today
A B A B
   
A 30% 35% A 65% 70%
Tomorrow   Tomorrow  
B 65% 70% B 35% 30%

11B Interpreting transition matrices 555
11B Interpreting transition matrices
Learning intentions
I To be able to interpret a transition matrix and a transition diagram.
Let us return to the car rental problem at the start of this section. As we saw then, the
following transition matrix, T , and its transition diagram can be used to describe the weekly
pattern of rental car returns in Bendigo and Colac.
Rented in B – Bendigo
 B C  C – Colac
0.80 0.10 B 20%
T = 
 
Returned to 80% B C 90%
0.20 0.90 C 10%
Using this information alone, a number of predictions can be made.

For example, if 50 cars are rented in Bendigo this week, the transition matrix predicts that:
80% or 40 of these cars will be returned to Bendigo next week (0.80 × 50 = 40)
20% or 10 of these cars will be returned to Colac next week (0.20 × 50 = 10).
Further, if 40 cars are rented in Colac this week, the transition matrix predicts that:
10% or 4 of these cars will be returned to Bendigo next week (0.10 × 40 = 4)
90% or 36 of these cars will be returned to Colac next week (0.90 × 40 = 36).
Example 3 Interpreting a transition matrix
The following transition matrix, T , and its transition diagram can be used to describe the
weekly pattern of rental car returns in three locations: Albury, Wodonga and Benalla.
A W B
A = Albury
80%
 
 0.7 0.05 0.12 A B = Benalla
W = Wodonga
 
T =  0.1 0.8 0.11 W Returned to
10%
 
0.2 0.15 0.77 B W 15%
70%
A 5% 11% B
20%
77%
12%
Use the transition matrix T and its transition diagram to answer the following questions.
a What percentage of cars rented in Wodonga each week are predicted to be returned to:
i Albury? ii Benalla? iii Wodonga?

556 Chapter 11 Transition matrices and Leslie matrices 11B
b Two hundred cars were rented in Albury this week. How many of these cars do we
expect to be returned to:
i Albury? ii Benalla? iii Wodonga?
c What percentage of cars rented in Benalla each week are not expected to be returned to
Benalla?
d One hundred and sixty cars were rented in Albury this week. How many of these cars
are expected to be returned to either Benalla or Wodonga?
Solution
a i 0.5 or 5% ii 0.15 or 15% iii 0.80 or 80%
b i 0.70 × 200 = 140 cars ii 0.20 × 200 = 40 cars iii 0.10 × 200 = 20 cars
c 11 + 12 = 23% or 100 − 77 = 23%
d 20% of 160 + 10% of 160 = 48 cars
Exercise 11B
Interpreting transition matrices

Example 3 1 When people go to the movies they buy either a bag of popcorn (P) or an ice cream (I).
Experience has shown that:
85% of people who buy popcorn this time will buy popcorn next time
15% of people who buy popcorn this time will buy an ice cream next time
75% of people who buy an ice cream this time will buy an ice cream next time
25% of people who buy ice cream this time will buy popcorn next time.
a Construct a transition matrix and transition diagram that can be used to describe this
situation. Use the models below.
This time
P I
 
  P
T =   Next time P I
I
b Eighty people are seen buying popcorn at the movies. How many of these are
expected to buy popcorn next time they go to the movies?
c Sixty people are seen buying an ice cream at the movies. How many of these are
expected to buy popcorn next time they go to the movies?
d On another occasion, 120 people are seen buying popcorn and 40 are seen buying
an ice cream. How many of these are expected to buy an ice cream next time they
attend the movies?

11B 11B Interpreting transition matrices 557
2 On Windy Island, sea birds are observed nesting at three sites: A, B and C. The
following transition matrix and accompanying transition diagram can be used to predict
the movement of these sea birds between these sites from year to year.
This year 10%
A B C 100% 80%
A B
 
 1.0 0.10 0.05 A
  5%
T =  0 0.80 0.05 B Next year
  5% 10%
0 0.10 0.90 C C 90%
a What percentage of sea birds nesting at site B this year were expected to nest at:
i site A next year? ii site B next year? iii site C next year?
b This year, 850 sea birds were observed nesting at site B. How many of these are
expected to:
i still nest at site B next year? ii move to site A to nest next year?
c This year, 1150 sea birds were observed nesting at site A. How many of these birds
are expected to nest at:
i site A next year? ii site B next year? iii site C next year?
d What does the ‘1’ in column A, row A of the transition matrix indicate?
3 A car insurance company finds that:

22% of car drivers involved in an accident this year (A) are also expected to be
involved in an accident next year
9% of drivers who are not involved in an accident this year (N) are expected to be
involved in an accident next year.
The transition diagram that can be used to describe this situation is shown below.
78%
This year
22% 91%
A N
A N
 
0.22 0.09 A
T =   Next year
0.78 0.91 N
9%
a In 2015, 84 000 drivers insured with the company were not involved in an accident.
i How many of these drivers were not expected to be involved in an accident in
2016?
ii How many of these drivers were expected to be involved in an accident in 2016?
b In 2015, 25 000 drivers insured with the company were involved in an accident.
i How many of these drivers were expected to be involved in an accident in 2016?
ii How many of these drivers were expected to be involved in an accident in 2017?
iii How many of these drivers were expected to be involved in an accident in 2018?

558 Chapter 11 Transition matrices and Leslie matrices 11B
4 Fleas can move beteen Now

three locations A, B and C.
A B C
The way a flea moves after  
0.60 0.10 0.70 A
5 seconds in a location can  
be exactly described by the T = 0.20 0.80 0.10 B After 5 seconds
 
transition matrix. 0.20 0.10 0.20 C
The move is not dependent on any previous move.
a If there are 30 fleas at location A at the beginning of the 5-second period, how many
fleas would you expect to
i stay at A ii go to B iii go to C
at the end of the 5-second period?
b If there were 60 fleas at each of the locations how many fleas would you expect to
have at
i A ii B iii C
after one 5-second period?
 
60
 
c Find the product T 60 and comment.
 
60
 
d At the conclusion of the first 5-second period there are 30 fleas at C.
i How many of these go to A in the next 5-second period?
ii How many of these go to B in the next 5-second period?
iii How many of these go to C in the next 5-second period?
 
60

 
e Evaluate the product 1 1 1 T 60 and comment.
 
60
 

This week
5 There are 120 students in a Year
S O F
12 class. Each week every student  
0.6 0.2 0.4 S
participates in one of three activities:  
T = 0.1 0.7 0.1 O Next week
Sport (S ), Outdoor Activities (O)  
or First Aid (F). 0.3 0.1 0.5 F
The activities that the children select

each week change according to the
transition matrix opposite.
From the transition matrix it can be concluded that:
A in the first week of the program, eighty students do Sport, twenty students children
do Outdoor activities and twenty Students do First Aid.
B at least 50% of the students do not change their activities from the first week to the
second week.
11B 11C Transition matrices – using recursion 559
C in the long term, all of the children will choose the same activity.
D Sport is the most popular activity in the first week
E 40% of the students will do First Aid each week.
6 Warren text messages a friend each Today

week day of this week. His friends A B C D E
are Arthur (A), Belinda (B), Connie  
0 1 0 0 0 A
(C), Danielle (D) and Eleanor (E).  
0 0 0 1 0 B
On Monday, Warren will send a text  
message to Connie. T = 1 0 0 0 0 C Tomorrow
 
Based on the transition matrix, the 0 0 0 0 1 D
 
order in which Warren will text 0 0 1 0 0 E
message each of his friends for the
next four days is:
A Arthur, Belinda, Connie, Danielle B Danielle, Belinda , Arthur, Connie

C Danielle, Belinda, Arthur, Eleanor D Eleanor, Arthur, Danielle, Belinda
E Eleanor, Danielle, Belinda, Arthur
11C Transition matrices – using recursion

Learning intentions
I To be able to use a matrix recurrence relation: S0 = initial state matrix, Sn+1 = T Sn , to
generate a sequence of state matrices.
I To be able to informally identify the equilibrium state or steady-state matrix in the
case of regular state matrices.
We return again to the car rental problem. The car rental firm now plans to buy 90 new cars.
Fifty will be based in Bendigo and 40 in Colac.
Given this pattern of rental car returns, the first question the manager would like answered is:
‘If we start with 50 cars in Bendigo, and 40 cars in Colac, how many cars will be
available for rent at both towns after 1 week, 2 weeks, etc?’
You have met this type of problem earlier when doing financial modelling (Chapter 8). For
example, if we invest $1000 at an interest rate of 5% per annum, how much will we have
after 1 year, 2 years, 3 years, etc?
We solved this type of problem by using a recurrence relation to model the growth in our
investment year-by-year. We do the same with the car rental problem, the only difference
being that we are now working with matrices.

Constructing a matrix recurrence relation

A recurrence relation must have a starting point.
 
50
In this case it is the initial state matrix: S 0 =  
40
Generating S 1
To find out the

 number  of cars in Bendigo and Colac after 1 week, we use the transition
0.8 0.1
matrix T =   to generate the next state matrix in the sequence, S 1 , as follows:
0.2 0.9
S1 = T S0
    
0.8 0.1 50 0.8 × 50 + 0.1 × 40
=     =  
0.2 0.9 40 0.2 × 50 + 0.9 × 40
 
44
or S 1 =  
46
Thus, after 1 week we predict that there will be 44 cars in Bendigo and 46 in Colac.
Generating S 2
Following the same pattern, after 2 weeks;

    
0.8 0.1 44 39.8
S 2 = T S 1 =     =  
0.2 0.9 46 50.2

Thus, after 2 weeks we predict that there will be 39.8 cars in Bendigo and 50.2 in Colac.
Generating S 3
After 3 weeks:
    
0.8 0.1 39.8 36.9
S 3 = T S 2 =    
   = 
  
0.2 0.9 50.2 53.1

Thus, after 3 weeks we predict that there will be 36.9 cars in Bendigo and 53.1 in Colac.
A pattern is now emerging. So far we have seen that:
S1 = T S0
S2 = T S1
S3 = T S2
If we continue this pattern we have:
S4 = T S3
S5 = T S4
or, more generally, S n+1 = T S n .
With this rule as a starting point, we now have a recurrence relation that will enable us to
model and analyse the car rental problem on a step-by-step basis.

11C Transition matrices – using recursion 561
Recurrence relation
S 0 = intial value, S n+1 = T S n
Let us return to the factory problem in Example 2.
Example 4 Using a recursion relation to calculate state matrices step-by-step
The factory has a large number of machines. The machines can be in one of two states:
operating (O) or broken (B). Broken machines are repaired and come back into operation
and vice versa.
At the start, 80 machines are operating and 20 are broken.
Use the recursion relation
where
   
80 0.85 0.05
S 0 =   and T =  
20 0.15 0.95

to determine the number of operational and broken machines after 1 day and after 3 days.
   
80 0.85 0.05
1 Write down a column matrix S 0 =   T =   
20 0.15 0.95

with S 0 representing the initial
operational state of the machines,
and the transition matrix.
    
0.85 0.05 80 69
2 Use the rule S n+1 = T S n to S 1 = T S 0 =     =  
0.15 0.95 20 31
   
determine the operational state
After 1 day, 69 machines are operational and
of the machines after one day by
31 are broken.
forming the product S 1 = T S 0 and
evaluate.
    
0.85 0.05 69 60.2
3 To find the operational state of the S 2 = T S 1 =     =  
0.15 0.95 31 39.8
   
machines after 3 days, we must     
first find the operating state of the 0.85 0.05 60.2 53.16
S 3 = T S 2 =    
   = 
  
machines after 2 days (S 2 ) and 0.15 0.95 39.8 46.84

use this matrix to find S 3 using After 3 days, 53 machines are operating and
S3 = T S2 . 47 are broken.

Calculator hint: In practice, generating    

matrices recursively is performed on your CAS
80 80.
  → s0  
calculator as shown opposite for the calculations  
20
 
20.
performed in Example 11.    
0.85 0.05 0.85 0.05
  → t  
0.15 0.95 0.15 0.95
   
 
69.
t.s0  
31.
 
   
69 60 2
   
31 39 8
   
and so on.
A rule for determining the state matrix of a system after

n steps
While we can use the recurrence relation:
to generate state matrices step-by-step, there is a more efficient method when we need to de-
termine the state matrix after a large number of steps.
If we follow through the process step-by-step we have:
S1 = T S0
S 2 = T S 1 = T (T S 0 ) = T 2 S 0
S 3 = T S 2 = T (T S 1 ) = T 2 S 1 = T 2 (T S 0 ) = T 3 S 0
Continuing the process
S4 = T 4 S0
S5 = T 5 S0
or more generally, Sn = Tn S0 .
We now have a simple rule for finding the value, S n , of the state matrix after n steps.
A rule for finding the state matrix after n steps

If the recurrence rule for determining state matrices is
S 0 = initial state matrix, S n+1 = T S n ,
the state matrix after n steps (or transitions) is given by S n = T n S 0 .
Let us return to the factory problem we analysed in Example 2.

Example 5 Determining the nth state of a system using the rule Sn = Tn S0
The factory has a large number of machines. The machines can be in one of two states:
operating (O) or broken (B). Broken machines are repaired and come back into operation
and vice versa.
Initially, 80 machines are operating and 20 are broken, so:
   
80 0.85 0.05
S 0 =   and T =  
20 0.15 0.95

Determine the number of operational and broken machines after 10 days.

1 Write down the transition matrix, T , and  
0.85 0.05
 
 S 0 = 80
 
initial state matrix, S 0 . Enter the matrices T = 
0.15 0.95 20
  
into your calculator. Use T and S .
2 To find out how many machines are in
Sn = T n S0
operation and how many are broken after
10 days, write down the rule S n = T n S 0 and ∴ S 10 = T 10 S 0
substitute n = 10 to give S 10 = T 10 S 0 .
3 Enter the expression T 10 S into your    
0.85 0.05 0.85 0.05.
calculator and evaluate.   → t  
0.15 0.95 0.15 0.95.
   
   
80 80.
  → s  
20 20.
   
 
30.9056
t10 · s  
69.0944
 
4 Write down your answer in matrix form and

 
30.9
then in words. S 10 = 
 
69.1
After 10 days, 31 machines will be
operational and 69 broken.
Using the inverse matrix of a transition matrix

In the above we have seen how to move from left to right in the sequence of state matrices
by multiplying by the transition matrix.
S 0 , S 1 , S 2 , . . . S n , S n+1 . . .
We can move from right to left through the transition states by using the inverse of the
transition matrix. In general, the inverse is not a transition matrix.
S n+1 = T S n and S n = T −1 S n+1

Example 6
 
0.6 0.3
We have a transition matrix T =  
0.4 0.7

 
25 587
and we know that the state matrix S 4 =   .
34 413

Determine S 3 and S 2 .
Solution
 
7
 −1
We know that S 4 = T S 3 . Hence S 3 = T S 4 . First T −1 −1
=  3 .
− 43 2
You should hold this in your calculator and then
S 3 = T −1 S 4 and S 2 = T −1 S 3
     
 7 −1 25 587  7 −1 25 290
=  34    =  34   
− 3 2 34 413 − 3 2 34 710
   
   
25 290 24 300
=   =  
34 710 35 700
Note: To calculate S 2 given S 3 we could have used:
2
S 2 = T −1 S 4
The steady-state solution

A second question a manager might like answered about the car rental is as follows.
‘Will the number of rental cars available from each location vary from week to week or
will they settle down to some fixed value?’
To investigate this question, we start by listing the state matrices from week 0 to week 15.
Week 0 1 2 3 4–11 12 13 14 15
               
50 44 39.8 36.9 30.3 30.2 30.1 30.1
State matrix         ...        
40 46 50.2 53.1 59.7 59.8 59.9 59.9
               
What you should notice is that, as the weeks go by, the number of cars at each of the locations
starts to settle down. We call this the steady- or equilibrium- state solution.
For the rental car problem, the steady-state solution is 30.1 (in practice, 30) cars at the
Bendigo branch and 59.9 (in practice, 60) cars at the Colac branch, which means the
numbers of cars at each location will not change from then on.

This can be seen more clearly in the graph below (the points have been joined to guide the eye).
initial value: steady-state value:
Bendigo 50 Colac 60
60
Number of cars
50 Colac
40 Bendigo
30
20 initial value: steady-state value:
10 Colac 40 Bendigo 30
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Weeks
In summary, even though the number of cars returned to each location varied from day to day,
the numbers at each location eventually settled down to an equilibrium or steady-state solution.
In the steady state, the number of cars at each location remained the same.
Important
1 In the steady state, cars are still moving between Bendigo and Colac, but the number
of cars rented in Bendigo and returned to Colac is balanced by the number of cars
rented in Colac and returned to Bendigo. Because of this balance, the steady state is
also called the equilibrium state.
2 For a system to have a steady state, the transition matrix must be regular and the
columns must add up to 1. A regular matrix is one whose powers never contain
any zero elements. In practical terms, this means that every state represented in the
transition matrix is accessible, either directly or indirectly from every other state.
A strategy for estimating the steady-state solution

In the car rental problem we found that, even though the number of cars returned to each location
initially varied from day to day, it eventually settled down so the number of cars at each location
remained the same.
Although we arrived at this conclusion by repeated calculations, we can arrive at the solution
much faster by using the rule S n = T n S 0 to find the nth state.
Estimating the steady state solution

If S 0 is the initial state matrix, then the steady-state matrix, S , is given by
S = T n S0
as n tends to infinity (∞).
Note: While in practice we cannot evaluate T n for n = ∞, we find that, depending on the circumstances,
large values of n can often give a very close approximation to the steady-state solution.

Example 7 Estimating the steady-state solution
For the car rental problem:

   
50 0.8 0.1
S 0 =   and T = 
   
40 0.2 0.9

Estimate the steady-state solution by calculating S n for n = 10, 15, 17 and 18.
1 Write down the transition matrix T
   
and initial state matrix S 0 . Enter the 0.8 0.1 50
matrices into your calculator. Use T T =    , S 0 =  

0.2 0.9 40
and S .
2 Use the rule S n = T n S 0 to write Sn = T n S0
down the expression for the nth state  
30.6
for n = 10. ∴ S 10 = T 10 S 0 =  
59.4

3 Enter the expression T 10 S into your  

30.565
calculator and evaluate. t10 · s  
59.435

4 Repeat the process for n = 15, 17  
and 18. 30.095
t15 · s  
59.905
 
30.047
t17 · s  
59.953
 
 
30.033
t18 · s  
59.967
 
5 Write down your answer in matrix      

30.1 30.0 30.0
form and then in words. This result S 15 =    , S 17 =    , S 18 =   
59.9 60.0 60.0
  
agrees with the graphical result
arrived at earlier. The estimated steady-state solution is 30 cars
based in Bendigo and 60 cars based in Colac.
Note: To establish a steady state to a given degree of accuracy, in this case one decimal place, at least
two successive state matrices must agree to this degree of accuracy.

11C 11C Transition matrices – using recursion 567
Skill-
sheet Exercise 11C
Calculating state matrices step-by-step and by rule

   
200 0.7 0.6
Example 4 1 For the initial state matrix S 0 =   and the transition matrix T =  : use the
400 0.3 0.4

recursion relation: S 0 = initial state matrix, S n+1 = T S n , to determine:
a S1 b S2 c S3
   
200 0.7 0.6
Example 5 2 For the initial state matrix S 0 =   and the transition matrix T = 
   : use the
400 0.3 0.4

recursion relation: S 0 = initial state matrix, S n+1 = T n S 0 , to determine:
a S5 b S7 c S 12
 
0.65 0.4
Example 6 3 We have a transition matrix T =   
0.35 0.6

 
5461
and we know that the state matrix S 5 =  .
4779
Determine S 4 and S 3 .
   
100 0.9 0.2
Example 7 4 For the initial state matrix S 0 =   and the transition matrix T =  :
200 0.1 0.8

a use the recursion relation: S 0 = initial state matrix, S n+1 = T S n , to determine:

i S1 ii S 2 iii S 3
b determine the value of T 5
c use the rule S n = T n S 0 to determine:
i S2 ii S 3 iii S 7
d by calculating S n = T n S 0 for n = 10, 15, 21 and 22, show that the steady-state matrix
 
200
is close to  .
100
   
100 0.7 0.4 0.1
   
5 For the initial state matrix S 0 = 200 and the transition matrix T = 0.2 0.1 0.3:
   
300 0.1 0.5 0.6
a use a recurrence relation to determine:
i S1 ii S 2 iii S 3
b use the relationship S n = T n S 0 to determine:
i S2 ii S 3 iii S 7
c by calculating S n = T n S 0 for n = 10, 15, 17 and 18, show that the steady-state matrix
 
247.1
 
is close to 129.4.
 
223.5 © Peter Jones et al 2023 Cambridge University Press
ISBN 978-1-009-11041-9
568 Chapter 11 Transition matrices and Leslie matrices 11C
Practical applications of transition matrices

6 Two fast-food outlets, Jill’s and Pete’s, are located in a small town.
In a given week:
80% of people who go to Jill’s return the next 20%
week
20% of people who go to Jill’s go to Pete’s the 80% J P 75%
next week
25% of people who go to Pete’s go to Jill’s the 25%
next week
75% of people who go to Pete’s return the next week.
a Construct a transition matrix to describe this situation. Call the matrix T .
b Initially, 400 people eat at Jill’s and 400 eat at Pete’s. Write down a column matrix
S 0 that describes this situation.
c How many of these people do we expect to go to Jill’s the next week? How many to
Pete’s?
d How many do we expect to go to Jill’s after 5 weeks? How many to Pete’s?
e In the long term, how many do we expect to eat at Jill’s each week? How many of
these people do we expect to eat at Pete’s?
7 Imagine that we live in a world in which people are either ‘happy’ or ‘unhappy’, but
the way people feel can change from day to day.
In this world:
90% of people who are happy today 10%
will be happy tomorrow
10% of people who are happy today 90% H U 40%
will be unhappy tomorrow
40% of people who are unhappy 60%
today will be unhappy tomorrow
60% of people who are unhappy today will be happy tomorrow.
a Construct a transition matrix to describe this situation. Call the matrix T .

b On a given day, out of 2000 people, 1500 are happy and 500 are unhappy. Write
down a column matrix, S 0 , that describes this situation.
c The next day, how many of these people do we expect to be ‘happy’ and how many
‘unhappy’?
d After 4 days, how many of these people do we expect to be ‘happy’ and how many
‘unhappy’?
e In the long term, how many people do we expect to be ‘happy’ and how many
‘unhappy’?

11C 11C Transition matrices – using recursion 569
8 In another model of this world, people can be ‘happy’, ‘neither happy nor sad’, or
‘sad’, but the way people feel can change from day to day.
The transition matrix opposite shows how H N S
people’s feelings may vary from day to day  
H 0.80 0.40 0.35
in this world, and the proportions of people  
T = N 0.15

0.30 0.40
involved.  
S 0.05 0.30 0.25
In the transition matrix, the columns define the situation today and the rows define the
situation tomorrow.
a On a given day, out of 2000 people, 1200 are ‘happy’, 600 are ‘neither happy nor
sad’ and 200 are ‘sad’. Write down a matrix, S 0 , that describes this situation.
b The next day, how many people do we expect to be happy?
c After 5 days, how many people do we expect to be happy?
d In the long term, how many of the 2000 people do we expect to be happy?

9 Students at a boarding school have a choices of two breakfast cereals, Crispies (C)
and Krunchies (K). The change in the percentage of students who have each cereal on
consecutive days is descibed by the transition matrix T shown below.
Today
C K
 
C 0.42 0.56
T =   Tomorrow
K 0.58 0.44
On Monday 25% of the students ate Crispies. What percentage of the students ate
Krunchies on Tuesday?
A 47.5% B 48% C 50.25% D 52.5% E 62%
10 A factory employs the same number of workers each day. The workers are allocated to
work with either machine A or machine B. The workers may be allocated to work on a
different machine from day to day, as shown in the transition matrix below.
A B
 
A 0.32 0.16
T =  
B 0.68 0.84
Machine A has 72 workers each day on it. Each day, the number of workers machine B
will be
A 24 B 36 C 72 D 288 E 306

570 Chapter 11 Transition matrices and Leslie matrices 11C
11 Consider the matrix recurrence relation below.
   
10  x 0.4 y 
   
S 0 = 20 , S n+1 = T S n where T = 0.6 z 0.4
   
30 0.1 0.2 w
 
14
 
Matrix T is a regular transition matrix. Given that S 1 = 26 which of the following is
 
20
true?
A x = 0.3, y = 0.1, z = 0.4, w = 0.5 B x = 0.3, y = 0.3, z = 0.4, w = 0.7
C x = 0.2, y = 0.7, z = 0.3, w = 0.3 D x = 0.2, y = 0.8, z = 0.3, w = 0.2
E x = 0.3, y = 0.6, z = 0.4, w = 0.4
11D Transition matrices – using the rule Sn+1 = TSn + B

Learning intentions
I To be able to use a matrix recurrence relation S0 = initial state matrix, Sn+1 = T Sn + B
to model systems that include external additions or reductions at each step of the
process.
To date, we have only considered matrix recurrence models of the form

S 0 = initial state matrix, S n+1 = T S n
This recurrence model can be used to model situations where the total number of objects in the
system, like cars, machines, people or birds, remains unchanged. For example, in the car rental
problem 90 cars are available for rental. But what happens if management wants to increase the
total number of cars available for rent by adding, say, an extra car at each location each week?
To allow for this situation we need to use the matrix recurrence relation:
S 0 = initial state matrix, S n+1 = T S n + B
where B is a column matrix.
The next example applies this model to the rental car problem.
Example 8 Determining the nth state of a system using the rule Sn = T n S0 + B
A rental starts with 90 cars, 50 located at Bendigo and 40 located at Colac.

Cars are usually rented and returned in the B – Bendigo
same town. However, a small percentage
C – Colac
of cars rented in Bendigo are returned 20%
in Colac and vice versa. The transition 80% B C 90%
10%
diagram opposite gives these percentages.
11D Transition matrices – using the rule Sn+1 = TSn + B 571
To increase the number of cars, two extra cars are added to the rental fleet at each location
each week. The recurrence relation that can be used to model this situation is:
     
50 0.8 0.1 2
S 0 =   , S n+1 = T S n + B where T =   and B =  
40 0.2 0.9 2
Determine the number of cars at Bendigo and Colac after:
a 1 week b 2 weeks.
a Use the rule S 1 = T S o + B to determine S 1 = T S0 + B
the state matrix after 1 week and write     
0.8 0.1 50 2
your conclusion. =     +  
0.2 0.9 40 2
     
44 2 46
=   +   =  
46 2 48
Thus, we predict that there will be 46 cars
in Bendigo and 48 cars in Colac.
b Use the rule S 2 = T S 1 + B to determine S 2 = T S1 + B
the state matrix after 2 weeks and write     
0.8 0.1 46 2
your conclusion. =     +  
0.2 0.9 48 2
     
41.6 2 43.6
=    +   = 
    
52.4 2 54.4

Thus, we predict that there will be 43.6

cars in Bendigo and 54.4 cars in Colac.
Unfortunately, the recurrence rule S n+1 = T S n + B does not lead to a simple rule for the state
matrix after n steps, so we need to work our way through this sort of problem step-by-step.
Using the inverse matrix of a transition matrix

In the above we have seen how to move from left to right in the sequence of state matrices by
applying S n+1 = T S n + B
S 0 , S 1 , S 2 , . . . S n , S n+1 . . .
We can move from right to left through the transition states by the observation
S n+1 = T S n + B and S n = T −1 (S n+1 − B)
Question 4 in the Exercise can be completed in this way.

572 Chapter 11 Transition matrices and Leslie matrices 11D
Skill-
sheet Exercise 11D
Using a recurrence rule to calculate state matrices

   
0.6 0.2 100
Example 7 1 For the transition matrix T =   and the state matrix S 0 =  :
0.4 0.8 100
  
a use the recurrence rule S n+1 = T S n to determine:

i S1 ii S 3  
10
b use the recurrence rule S n+1 = T S n + R, where R =  , to determine:
5
i S1 ii S 2
 
−20
c use the recurrence rule S n+1 = T S n − B, where B =  , to determine:
20

i S1 ii S 2
Practical application
2 On Windy Island, sea birds are observed nesting at three sites: A, B and C. The
following transition matrix and accompanying transition diagram can be used to predict
the movement of sea birds between these sites from year to year.
This year 10%
A B C 100% 80%
 
 1.0 0.10 0.05 A A B
5%
 
T =  0 0.80 0.05 B Next year
5% 10%
 
0 0.10 0.90 C
C 90%
 
10 000
 
Initially, 10 000 sea birds were observed nesting at each site, so S 0 = 10 000.
 
10 000
a Use the recurrence rule S n+1 = T S n to:
i determine S 1 , the state matrix after 1 year
ii predict the number of sea birds nesting at site B after 2 years.
b Without calculation, write down the number of sea birds predicted to nest at each of
the three sites in the long term. Explain why this can be done without calculation.
c To help solve the problem of having all the birds eventually nesting at site A, the
ranger suggests that 2000 sea birds could be removed from site A each year and
relocated in equal numbers to sites B and C.
The state matrix, S 2 , is now given by
S2 = T S1 + N

11D 11D Transition matrices – using the rule Sn+1 = TSn + B 573
     
10000 1.0 0.10 0.05 −2000
     
where S 1 = 10000 , T =  0 0.80 0.05 and N =  1000  .
     
10000 0 0.10 0.90 1000
Evaluate:
i S2 ii S 3 (assuming that S 3 = T S 2 + N) iii S 4 (assuming that S 4 = T S 3 + N).

3 The matrix S n+1 is determined from the matrix S n using the recurrence relation
S n+1 = T × S n − C, where
     
0.6 0.75 0.1 2000 1975
     
T = 0.2 0.2 0.1 , S 0 = 1000 , S 1 =  650  .
     
0.2 0.05 0.8 1000 1125
and C is a column matrix. Matrix S 2 =
         
 1860   1700   1710   1600   1650 
         
A  687.5  B  560.5  C  587.5  D 1725.5 E  550.5 
         
1452.5 1022.5 1202.5 1200.5 1032.5
         
4 Supporters of a football team attend home games. There are 3 areas, bays A, B and C,
where they sit. There is considerable moving of position from game to game and the
numbers attending the home games gradually decline as the year progresses. Let Xn
be the state matrix that shows the number of supporters in each bay n weeks into the
the season. The number of supporters in each location can be determined by the matrix
recurrence relation
Xn+1 = T Xn − D
where
This game
A B C
 
  70
 0.1 0.2 0.5  A  
  and D = 70
T =  0.3 0.7 0.2  B Next game  
  70
0.6 0.1 0.3 C
 
 9830 
 
If X3 = 11 130 then X2 =
 
7830
         
12370  4000  12 000  7500   79300 
         
A  9510  B 10 000 C 11 000 D  8600  E 245232
         
4650 15 000 5000 12 000 231011
         

11E Leslie matrices

Learning intentions
I To use and interpret Leslie matrices to analyse changes in population over time.
Leslie matrices are used to construct discrete models of population growth. In particular, they
are used to model changes in the sizes of different age groups within a population.
The general setting

Leslie matrices were developed by Patrick Holt Leslie (1900–1972) while he was working in
the Bureau of Animal Population at the University of Oxford. They are used by biologists and
ecologists to model changes over time in various animal populations.
Age groups First the population is divided into age groups. The time period for each age
group is the same length. Together they cover the life span of the population. For example,
in a study of a human population, we could use a time period of 10 years and consider eleven
age groups as follows:
Age group(i) 1 2 3 4 ··· 10 11

Age range (years) 0–10 10–20 20–30 30–40 ··· 90–100 100–110
Note: Only the females of the species are counted in the population, as they are the ones who give birth to
the new members of the population.
A Leslie matrix is a transition matrix that can be used to describe the way population changes
over time. It takes into account two factors for the females in each age group: the birth rate, bi ,
and survival rate, si , where i is the number of the age group.
Birth rates We ignore migration, and so the population growth is entirely due to new female
births. The birth rate, bi , for age group i is the average number of female offspring from a mother
in age group i during one time period. For example, average birth rate of women in age group 4
(20 − 30 years) might be 1.7 female children for the 10 year period.
Survival rates The survival rate, s i , for age group i is the proportion of the population in
age group i that progress to age group i + 1. Note that 0 ≤ s i ≤ 1.
For example, the survival rate for age group 2 might be 0.95, that is 95% of females in this
10 − 20 year age group would survive to progress to age group 3, 20 − 30 years.
Note: The survival rate of the last age group (100 − 110) is taken to be 0.
A simple example
We start with a simple example where the life span of the species is 9 years. We will divide the
population into three age groups. This means we use a time period of 3 years.

11E Leslie matrices 575
Age group(i) 1 2 3
Age range (years) 0–3 3–6 6–9
A Leslie matrix for three age groups is a 3 × 3 matrix of the form

 
b1 b2 b3 
 
L =  s1 0 0 
 
0 s2 0
Suppose that the survival rates are s1 = 0.6 and s2 = 0.3, and that the birth rates are b1 = 0,
b2 = 2.3 and b3 = 0.4. Then the Leslie matrix is
From age group i
1 2 3
 
0 2.3 0.4  1 birth rate
 
L = 0.6 0 0  2 survival rate To age group i + 1
 
0 0.3 0 3 survival rate
Life cycle transition diagram

The above Leslie matrix can be represented by a diagram which we will refer to as a Life cycle
transition diagram.
0.4
2.3
1 2 3
0.6 0.3
The population state matrix

The population state matrix is a column matrix that lists the number in each age group at a
given time.
The initial population state matrix is denoted by S 0 . Suppose that for our example, initially
the population has 400 females in each age group. We represent the initial population state
matrix S 0 , as a 3 × 1 column matrix as shown below.
Age group
 
400 1
 
S 0 = 400 2
 
400 3
We can now use the Leslie matrix, L, in combination with the initial state matrix S 0 to generate
the state matrix after one time period, S 1 , to find the size of each age group after one time period
(3 years) as follows:

    
0 2.3 0.4  400 1080
     
S 1 = LS 0 = 0.6 0 0  400 =  240
     
0 0.3 0 400 120
Thus after one time period, there are 1080 females in age group 1, 240 in age group 2 and 120
in age group 3 and the total population size has increased from 1200 (= 400 + 400 + 400) to
1440 (= 1080 + 240 + 120). Similarly, to find the number in each age group after two time
periods we calculate S 2 from S 1 as follows:
    
0 2.3 0.4  1080 600
     
S 2 = LS 1 = 0.6 0 0   240 = 648
     
0 0.3 0 120 72
Thus, after two time periods, there are 600 females in age group 1648 in age group 2 and 72 in
age group 3 and the over-all population size has decreased to 1320.
Finding the population matrix Sn after n time periods.
To speed up the process we can make use of the explicit formula for the state matrix S n after n
time periods. Notice that there is a pattern when calculating the population state matrices:
S 1 = LS 0
S 2 = LS 1 = L2 S 0
S 3 = LS 2 = L3 S 0
..
.
S n+1 = LS n = Ln S 0
In general, we can find the population matrix S n using the rule
S n = Ln S 0
Using this rule, to find S 3 , we have
 3    
0 2.3 0.4 400 1519.2

S 3 = L3 S 0 = 0.6 0 0  400 =  360
    
     
0 0.3 0 400 194.4
Continuing in this way, we can see the change over time in the total population and in the
distribution of the age groups.
Change in the population over time
Time period 0 1 2 3 4 5
Age 0–3 years 400 1080 600 1519.2 905.76 2139.70
Age 3–6 years 400 240 648 360.0 911.52 543.46
Age 6–9 years 400 120 72 194.4 108.00 273.46
Total 1200 1440 1320 2073.6 1925.28 2956.61

Leslie matrices
An m × m Leslie matrix has the form
 
b1 b2 b3 · · · bm−1 bm 
 
 s1 0 0 · · · 0 0 

 
 0 s
2 0 ··· 0 0 
L = 
 
 0 0 s3 · · · 0 0 
 . .. .. . . .. .. 
 . .
 . . . . . 
 
0 0 0 · · · sm−1 0
where:
m is the number of age groups being considered
s i , the survival rate, is the proportion of the population in age group i that progress to
age group i + 1
bi , the birth rate, is the average number of female offspring from a mother in
age group i during one time period.
Leslie matrix and its interpretation
From age group
1 2 3 4
 
 0 1.4 1.2 0.3 1
 
0.6 0 0 0  2
L =  To age group


 0 0.5 0 0  3
 
0 0 0.1 0 4
This is a Leslie matrix with 4 age groups. The corresponding life-cycle transition
diagram is shown here.
0.3
1.2
1.4
1 2 3 4
0.6 0.5 0.1
Recursive rules
The population matrix S n is an m × 1 matrix representing the size of each age group after
n time periods. This is calculated using a recursive formula
S 0 is the initial state matrix, S n+1 = LS n
or the explicit rule
S n = Ln S 0

Example 9 Determining state matrices and life cycle diagrams
Use the Leslie matrix and initial state matrix below to answer the following questions.
From age group
1 2 3 4
   
 0 1.8 2.6 0.1 1 1000
   
0.2 0 0 0  2  0 
L =  To age group S 0 = 
  
 0 0.4 0 0  3  0 
   
0 0 0.3 0 4 0
  
a Write down
i the birth rate for age group 2 ii the survival rate for age group 3
b Complete a life cycle diagram for this Leslie matrix.
c Evaluate the following population state matrices.
S 1 , S 5 and S 20
 
9.53
 
2.42
d Given that S 16 =  , determine S 17
 
1.22
 
0.16
a i The birth rate for age group 2 is Birth rate for age group 2 = 1.8
given in the matrix position, row 1,
column 2.
ii The survival rate for age group 3 is Survival rate for age group 3 = 0.3
given in the matrix position, row 4,
column 3.
b Survival rates 0.1
s1 = 0.2, s2 = 0.4, s3 = 0.3 2.6
1.8
Birth rates
1 2 3 4
b2 = 1.8, b3 = 2.6, b4 = 0.1 0.2 0.4 0.3
c S 1 = LS 0 Using a calculator.
S5 = L S0
     
5  0  149.76 3.84
     
S 20 = L20 S 0 200  26.4  0.97
S 1 =  , S =  , S =   
 0  5  16.64  20 0.49
     
0 8.64 0.19
     

d S 17 = LS 16 
7.54

(Further investigation would reveal that 
1.91

the population continues to decrease S 17 =  
0.97
over time.)  
0.37
 
Example 10 Entering information in a Leslie matrix and state matrix
Information about a population of female goats is given in the following table.
Age group (years) 0−1 1−2 2−3 3−4 4−5

Initial population 10 25 40 20 15
Birth rates 0 0.2 0.9 0 0
Survival rates 0.6 0.7 0.5 0.2 0
a Write down the initial population state matrix, S 0 .

b Using the information in the table above, write down a Leslie matrix to describe the
change in population of female goats.
c Construct a life cycle transition diagram for this Leslie matrix.
d Determine the number of 3 − 4 year old female goats in the population after 3 years.
Round your answer to the nearest whole number.
a Enter the initial population  
10
numbers into a 5 × 1 matrix.  
25
 
S 0 = 40
 
 
20
 
15
 
b Enter the birthrates and survival 

 0 0.2 0.9 0

0
rates into a 5 × 5 matrix. 
0.6 0 
 0 0 0
L =  0 0.7 0

 0 0
 
 0 0 0.5 0 0
 
0 0 0 0.2 0

c Survival rates
0 0
s1 = 0.6, s2 = 0.7, s3 = 0.5, 0.9
s4 = 0.2 0.2
1 2 3 4 5
Birth rates 0.6 0.7 0.5 0.2
b2 = 0, 2, b3 = 0.9, b4 = 0,
b5 = 0
 3    
 0 0.2 0.9 0 0 10  8.7 
     
0.6 0 0 0 0 25 10.17

d S 3 = L3 S 0 0 40 = 17.22
 0 0.7 0     
 0
     
 0 0 0.5 0 0 20  2.1 
     
0 0 0 0.2 0 15 1.75
There are two goats in the 3-4 year old age group
in this population.
Long term (limiting) behaviour of population numbers

The following examples demonstrate numerical techniques for modelling the use of Leslie
matrices.
Example 11 Limiting behaviour for Leslie matrices
Consider the following Leslie matrix L and initial population matrix S 0 :

   
0 4 4  1000
   
L = 0.25 0 0 and S 0 =  0
   
0 0.5 0 0
a Find
i S5 ii S 10 iii S 50

Premultiply each of these state matrices by 1 1 1 to calculate the total populations
at each of these stages and comment.
b Determine S 25 and S 26 . Divide each age group population for S 26 by the correspond-
ing age group population for S 25 and show that S 26 ≈ 1.1915S 25 and comment.
Solution
a
     
1000  2500  2 777 063
     
i S 5 =  250 ii S 10 = 531.25 iii S 50 = 582 688.05
     
62.5 218.75 244521.18
 

Finding the total populations according to this model

i 1 1 1 S 5 = [1312.5]. The population is approximately 1312 after 5 years.

ii 1 1 1 S 10 = [3250]. The population is approximately 3250 after 10 years.

iii 1 1 1 S 50 = [3 604 272.2]. The population is approximately 3 604 272
after 50 years.
A numerical investigation reveals that the population increases without bound.
b We calculate S 25 and S 26 :
   
34 781.25 41 441.41
   
S 25 = L25 S 0 =  7297.85 , S 26 = L26 S 0 =  8695.31
   
3062.50 3648.93
Then we can find the rate of increase in each age group during the 26th time period:
41 441.41 8695.31 3648.93
≈ ≈ ≈ 1.1915
34 781.25 7297.85 3062.50
This suggests that the age-group proportions have stabilised after the 25 time periods.
The long-term growth rate is approximately 1.19. (That is, after a certain stage, the
population is increasing by 19% each time period.)
Note: Try different entries in S 0 to see if you get the same behaviour. The long-term growth rate is
largely dependent on the Leslie matrix L.
Limiting behaviour of Leslie matrices

Often we will find that, after a long enough time, the proportion of the population in
each age group does not change from one time period to the next. This happens if we can
find a real number k such that LS n+1 = kS n for some sufficiently large n. This does not
happen with every Leslie matrix as we see in Example 13.
Example 12 A Leslie matrix and state matrix with constant rate of increase

   
0 2.3 0.4  1600
   
L = 0.6 0 0  and S 0 =  800 .
   
0 0.3 0 200
a Determine
i S 10 ii S 14 iii S 15
b The rate of increase of the population is a constant and each of the age group
populations increase in the same way. Find this rate by comparing S 14 and S 15 .
c Confirm the ratio of the age group populations stays constant for S 0 , S 1 and S 10 at
8:4:1

Solution
a 
9906.78
 
20542.70
 
 24651.24

     
i S 10 = 4953.39 ii S 14 = 10271.35 iii S 15 =  12325.62
     
1238.35 2567.84 3081.4043
By comparing S 14 and S 15 ,
24651.24 12325.62 3081.4043
= = ≈ 1.2
20542.70 10271.35 2567.84
we find that the growth rate is 1.2.
c 8 : 4 : 1 = 1600 : 800 : 200 ≈ 9906.78 : 4953.39 : 1238.35
b
Example 13 Periodic, increasing and decreasing rates of change

   
0 0 b3  1000
  
L = 0.25 0 0  and S 0 =  0
 
   
0 0.5 0 0
Investigate the long-term behaviour of the population if:
a b3 = 8 b b3 = 4 c b3 = 10
Solution
a Let b3 = 8. Use your calculator to store the matrices L and S 0 . Then compute:
     
 0  0 1000
    
S 1 = LS 0 = 250 , S 2 = L2 S 0 =  S 3 = L3 S 0 =  0

0 ,
     
0 125 0
The population will continue to cycle through these three states; this is because L3 = I.
b Let b3 = 4. Then a numerical investigation suggests that the population decreases over
the long term:
     
 0  0  0 
     
S 1 = LS 0 = 250 , S 5 = L5 S 0 =  0  , S 50 = L S 0 = 0
50  

     
0 62.5 0.0019
c Let b3 = 10. Then a numerical investigation suggests that the population increases over
the long term:
     
 0  0   0 
     
S 1 = LS 0 = 250 , S 5 = L5 S 0 =  0  ,
 S 50 = L50 S 0 =  0 

     
0 156.25 4440.89
Note: A population can increase, decrease, become constant or oscillate.

11E 11E Leslie matrices 583
Skill-
sheet Exercise 11E
Determining state matrices and life cycle diagrams
Example 8 1 Use the Leslie matrix and initial state matrix below to answer the following questions.
From age group
1 2 3 4
   
 0 1.9 2.1 1.1 1 100
   
0.7 0 0 0  2 100
L =  To age group S 0 = 
 
 0 0.5 0  100
 0  3  

0 0 0.6 0 4 100
  
a Write down
i The birth rate for age group 2 ii The survival rate for age group 3
b Complete the life cycle diagram for this Leslie matrix.
c Evaluate the following population state matrices.
 
2613
 
1200
d Given that S 7 =   determine S 8 . Give your values correct to the nearest whole

 485 
 
168
number.
2 Complete the life cycle diagram corresponding to each of the following Leslie
matrices:
   
 0 2.9 3.1 2.1    0 0 3 8
   0 0 0.42   
0.8 0 0 0    0.4 0 0 0
a   b 0.6 0 0 
 c  
 0 0.7 0 0   
 
  0 0.5

 0 0
  0 0.75 0  
0 0 0.5 0 0 0 0.25 0

3 Construct the Leslie matrix corresponding to each life cycle diagram.

a 2.4 b 3
1.3 2.3
1 2 3 1 2 3
0.7 0.6 0.6 0.3
c 0.6
2.6
1.4
1 2 3 4
0.5 0.4 0.05

584 Chapter 11 Transition matrices and Leslie matrices 11E
Entering information in a Leslie matrix and state matrix
Example 9 4 Information about a population of female kangaroos in a particular area is given in the
following table.
Age group (years) 0−4 4−8 8 − 12 12 − 16 4−5

Initial population 15 20 30 15 10
Birth rates 0 0.2 0.9 1.1 0
Survival rates 0.8 0.9 0.7 0.8 0

b Write down the Leslie matrix.
c Complete the life cycle diagram for this Leslie matrix.
d Determine the population state matrix after
i one year, (S 1 ) ii after 5 years, (S 5 ).
e Determine the number of 4 − 8 year old female kangaroos in the population after 5
years.
f State the initial total population.

g Use multiplication of state matrices by the matrix 1 1 1 1 1 to find the total
population after
i one year, ii after 5 years iii after 10 years.
h It is conjectured that the population is increasing by about 10% per annum.
Calculate each of the following and comment
i 1.1 × 90 ii 1.15 × 90 iii 1.110 × 90.
5 Information about a population of female locusts is given in the following table.
Stage Eggs Nymphs Adults

Initial population 0 0 50
Birth rates 0 0 1000
Survival rates 0.02 0.05 0

b Write down the Leslie matrix.
c Construct a time life-cycle transition diagram for this Leslie matrix.
d Determine the population state matrix after
i one year, (S 1 ) ii after 3 years, (S 3 ) iii after 4 years, (S 4 ).

e If the Initial population is now:
Stage Eggs Nymphs Adults

Initial population 50 100 50
find the populations of each after

i one year (S 1 ) ii after 3 years (S 3 ). iii after 4 years (S 4 ).
Limiting behaviour for Leslie matrices
Example 10 6 Consider the following Leslie matrix L and initial population state matrix S 0 :
   
0 2 1  204
   
L = 0.5 0 0 and S 0 =  96
   
0 0.25 0 23
a Find

Premultiply each of these state matrices by 1 1 1 to calculate the total
populations at each of these stages and comment.
b Determine S 20 and S 21 . Divide each age group population for S 21 by the
corresponding age group population for S 20 and show that S 21 ≈ 1.057S 20 and
comment.
Example 11 7 A Leslie matrix that models a certain population of female animals is

 
0 2.5 1 
 
L = 0.6 0 0
 
0 0.25 0
where the animals have a maximum life span of 9 years, and the population has been
divided into three age groups of 3 years each.
a Assume that each age group initially consists of 400 females. What is the number of
females in each age group after:
i 3 years ii 6 years iii 9 years?
 
767
 
b Now assume that the initial population is 1200 and S 0 = 362 . Find S n after
 
71
 
i 3 years ii 6 years iii 9 years.

586 Chapter 11 Transition matrices and Leslie matrices 11E
c Calculate
i 1.27S 0 ii 1.272 S 0 iii 1.273 S 0
Compare these answers to the answers of part b
Example 12 8 Consider the following Leslie matrix L and initial population matrix S 0 :
   
 0 0 12 1200
   
L =  14 0 0  and S 0 =  0
   
0 31 0 0
a Find:
i LS 0 ii L2 S 0 iii L3 S 0
b Comment on these results in terms of the population behaviour. Try using a different
initial population matrix S 0 .
c Now investigate for each of the following Leslie matrices. Comment on population
increase or decrease.
   
 0 0 6  0 0 15
  
i L =  41 0 0 ii L =  41

   0 0 

0 13 0 0 1
0
  
3
9 A Leslie matrix that models a certain population of female insects is

 
0 3 2 2
 
0.5 0 0 0
L = 
 
0 0.5 0 0 
 
0 0 0.1 0
where the insects have a maximum life span of 4 months, and the population has been
divided into four age groups of 1 month each.
Assume that each age group initially consists of 400 females. What is the number of
females in each age group after:
a 1 month b 2 months c 3 months?
10 For a certain species of fish, we consider three age groups each of one year in length.
These fish reproduce only during their third year and then die. Assume that 20% of fish
survive their first year and that 50% of these survivors make it to reproduction age. The
initial population consists of 1000 newborns.
a Investigate what happens for each of the following values of b3 :
i b3 = 10 ii b3 = 15 iii b3 = 6
b For b3 = 20, determine the long-term growth rate and the proportion of fish in each
age group.


11 The Leslie matrix for a certain endangered species is:
 
0.9 2.5 0.4
 
L = 0.3 0 0 
 
0 0.45 0
Some of the species were moved into a sanctuary. The initial female population in the
sanctuary is given by
 
130
 
S 0 =  40 
 
20
The best estimate of the total female population after 7 years is
A 1000 B 1500 C 2000 D 2500 E 3000
 
0 2 b
 
12 The Leslie matrix L = c 0 0 satisfies the matrix equation
 
0 d 0
   
16 16
   
L  4  =  4 
   
2 2
The values of b, c and d are
1 1 1 1 1 1
A b = 2, d = , c = B b = 2, d = , c = C b = 4, d = ,c =
4 2 2 4 4 2
1 1 1 1
D b = 4, d = , c = E b = 2, d = , c =
2 4 4 2
13 A population of birds is modelled by using the Leslie matrix
 
 0 2 1.5 
 
L = 0.44 0 0 
 
0 0.55 0
The growth has reached the point where the rates of growth of the different age groups
 
1000
 
of the population are constant and the state matrix at this point is S k =  400 . The rate
 
200
of growth per time period is
A 10% B 11% C 12%
D 13% E 14%

Review 588 Chapter 11 Transition matrices and Leslie matrices
State matrix A state matrix S n is a column matrix whose elements represent the
Transition nth state of a dynamic system defined by a recurrence relation of the
Assign-
ment matrixes
form: S 0 = initial state, S n+1 = T S n . Here T is a square matrix called a
transition matrix.
Steady-state The steady-state matrix, S, represents the equilibrium state of a

matrix system. For regular matrices, this equilibrium state of a system can be
estimated by calculating T n S 0 for a large value of n.
Leslie matrices An m × m Leslie matrix has the form

 
b1 b2 b3 · · · bm−1 bm 
 
 s1 0 0
 ··· 0 0 

 0 s
2 0 ··· 0 0 
L = 
 
 0 0 s3 ··· 0 0 
 . .. .. .. .. 
 . ..
 . . . . . . 
 
0 0 0 · · · sm−1 0
where:
• m is the number of age groups being considered
• si , the survival rate, is the proportion of the population in
age group i that progress to age group i + 1
• bi , the birth rate, is the average number of female offspring from a
mother in age group i during one time period.
The population matrix S n is an m × 1 matrix representing the size
of each age group after n time periods. This is calculated using a
recursive formula
S 0 is the initial state matrix, S n+1 = LS n
or the explicit rule
S n = Ln S 0
Skills checklist
list
11A 1 I can set up a transition matrix from a diagram.

Review
11A 2 I can set up a transition matrix from a written information.
11B 3 I can interpret a transition matrix and a transition diagram.
11C 4 I can use a recurrence relation to calculate state matrices step by step
11C 5 I can use a recurrence relation Sn+1 = Tn S0 to determine the nth state.
11C 6 I can use the inverse of a transition matrix.
11C 7 I can estimate steady state solution for suitable transition matrices.
11D 8 I can use the matrix recurrence relation S0 = initial state matrix, Sn+1 = TSn + B.
11E 9 I can determine state matrices and construct life cycle diagrams in situations
modelled by Leslie matrices.
11E 10 I can enter information into a Leslie matrix from written information.
11E 11 I can use numerical techniques to consider the limiting behaviour of Leslie
matrices.
11E 12 I can identify the properties of a Leslie matrix and the state matrices when
there is a constant rate.
Multiple choice questions

1 The transition matrix that can be used to represent 25%
the information in the diagram shown is:
75% A B 95%
5%

From From
 A B   A B 
A To A 0.75 0.25 B To A 0.75 0.05
   
B 0.05 0.95 B 0.25 0.95
From From
A B
 A B 
C To A 0.75 0.25 D To
 
  A 0.75 0.95
 
B 0.95 0.05 B 0.25 0.05
From
A B
E
 
To A 0.25 0.05
 
B 0.75 0.95
2 The transition matrix that can be used to represent 60%

the information in the diagram shown is:
X Y Z 10% Y 35%
  75%
A To X 0.75 0.05 0.30
 X 5% 20% Z

Y 0.10 0.60 0.20
  15%
Z 0.15 0.35 0.50 50%
30%
X Y Z X Y Z
 
B To X 0.75 0.10 0.15
 
C To X 0.75 0.10 0.15
   
Y 0.60 0.05 0.35 Y 0.10 0.05 0.35
   
Z 0.50 0.30 0.20 Z 0.50 0.30 0.20
X Y Z X Y Z
   
D To X 0.75 0.05 0.15 E To X 0.75 0.05 0.15
   
Y 0.10 0.60 0.20 Y 0.15 0.35 0.50
   
Z 0.15 0.35 0.50 Z 0.10 0.60 0.20
The following information is needed for Questions 3 to 8.

   
0.6 0.5 100
3 A system is defined by a transition matrix T =   with S 0 =  .
0.4 0.5 200
For this system, S 1 =
         
 60  140 160 166 200
A   B   C   D   E  
200 160 140 144 100
        

Review
4 For this system, T 2 is:
       
0.36 0.25 0.56 0.55 0.6 0.5 1.2 1.0
A   B   C   D  
0.16 0.25 0.44 0.45 0.4 0.5 0.8 1.0
    
E not defined
5 For this system, S 3 is closest to:

         
160 166.6 166.7 640 400
A   B    C    D    E   
140 133.4 133.3 560 800
   
6 For this system, the steady-state matrix is closest to:

         
166.5 166.6 166.7 166.8 166.9
A   B   C   D   E  
133.5 133.4 133.3 133.2 133.1
    
 
10
7 If L1 = T S 0 + B, where B =  , then L1 equals:
20
         
 70  150 170
 176 210
A   B   C   D   E  
220 180 160 164 120
    
 
10
8 If P1 = T S 0 − 2B, where B =  , then P1 equals:
20
         
140 170 180 170 180
A   
 B   
 C   D    E  
100 100 100 160 180
 
9 A system of state matrices S n is defined by the matrix equation S n+1 = GS n where

 
 0 −0.5
G =   .
1.5 0.5

 
10
If S 1 =   , then S 2 equals:
20
         
−12.5 −10 10 10 15
A   B   C   D   E  
−2.5 25 20 25 30
   
0.5 0.6
 is a transition matrix. S 5 = 22 is a state matrix.
 
10 T = 
0.5 0.4 18
  
If S 5 = T S 4 , then S 4 equals:
         
18 20 21.8 22 18.2
A   B   C   D   E  
22 20 18.2 18 21.2
   

11 A large population of birds lives on a remote island. Every night each bird settles at either
location A or location B.
On the first night the number of birds at each location was the same. On each subsequent
night, a percentage of birds changed the location at which they settled.
The movement of birds between the two locations is described A B
by the transition matrix T shown opposite. Assume this pattern  
A 0.8 0
of movement continues. T=  
B 0.2 1
In the long term, the number of birds that settle at location A will:
A not change B gradually decrease to zero C gradually increase

D eventually settle at around 20% of the island’s bird population
E eventually settle at around 80% of the island’s bird population

Two politicians, Rob and Anna, are the only
Number of people
candidates for a forthcoming election. At the
planing to vote
beginning of the election campaign, people
Candidate for candidate
were asked for whom they planned to vote.
Rob 5692
The numbers were as per the table.
Anna 3450
During the election campaign, it is expected 24%

that people may change the candidate that
they plan to vote for each week according to
75% Rob Anna 76%
the transition diagram shown.
25%
12 The total number of people who are expected to change the candidate that they plan to vote
for 1 week after the election campaign begins is:
A 828 B 1423 C 2251 D 4269 E 6891
13 The election campaign will run for 10 weeks. If people continue to follow this pattern of
changing the candidate they plan to vote for, the expected winner after 10 weeks will be:
A Rob by about 50 votes B Rob by about 100 votes
C Rob by fewer than 10 votes D Anna by about 100 votes
E Anna by about 200 votes

Review
1 The Diisco (D) and the Spin (S) are two large music venues in the same city. They both
open on the same Saturday night and will open on every Saturday night.
The matrix A1 opposite is the attendance matrix for the first
 
500 D
Saturday. This matrix shows the number of people who attended A1 =  
240 S

the first Diisco and the number of people who attended the Spin.
The number of people expected to attend the second Saturday for each venue can be
determined using the matrix equation
A2 = GA1
This Saturday
D S
−0.4 D Next Saturday
 
1.2
where G is the matrix G =  
0.2 0.6 S
a i Determine A2 , the attendance matrix for the second Saturday.

ii What was the total attendance on the second Saturday.
Assume that the attendance matrices for successive Saturdays can be determined as
follows:
A3 = GA2 , A4 = GA3 , and so on such that An+1 = GAn
b Determine the attendance matrix (with the elements written correct to the nearest whole
number) for the eighth Saturday.
c Describe the way in which the number of people attending the Diiscoo is expected to
change over the next 80 or so Saturdays.
Suppose instead that 500 people attend the first Diiscoo, and 490 people attend the Spin.
d Describe the way in which the attendance at both venues changes if attendance follows
this prediction.
2 Suppose that the trees in a forest are classified into three age groups: young trees
(0–15 years), middle-aged trees (16–30 years) and old trees (more than 30 years). A time
period is 15 years, and it is assumed that in each time period:
10% of young trees, 20% of middle-aged trees and 40% of old trees die
surviving trees enter into the next age group; old trees remain old
dead trees are replaced by young trees.


Y M O
 
Y  


M 

 

O
3 The following table represents a study of a particular population of marsupials, which has
been divided into eight age groups. The table gives the initial population, birth rate and
survival rate for each age group.
Age group 1 2 3 4 5 6 7 8
Initial population 0 100 100 50 0 0 0 0
Birth rate 0 0.1 0.9 0.2 0 0 0 0
Survival rate 0.98 0.95 0.95 0.9 0.7 0.5 0.1 0
a Write down the Leslie matrix for this population.

b Calculate S2 and S3 .
c Estimate the long-term growth rate of the population.
4 The growth of algae in a particular lake is being studied to protect the ecology from a
disastrous algal bloom. The algae can live for up to four days. So the population is divided
into four age groups of one day each. The fertility rates and survival rates are being
monitored so that the population can be modelled using a Leslie matrix.
At the beginning of the study in late winter (day 0), it was observed that the algae
concentration in the lake was 3200 cells per millilitre of water, with equal numbers in each
age group. The fertility rates on the four days of life were 0.2, 0.5, 0.6 and 0.4 respectively.
The survival rate for each of the first three days of life was 0.7.
a Write down a Leslie matrix to represent this particular model.
b Find the population matrix for cells per millilitre of water on day 20, correct to three
c Find the population matrix for cells per millilitre of water on day 21. Hence find the
rate of change in the algae concentration per day at this stage.
d With the coming of spring on day 21, the fertility rates increased to 0.3, 0.6, 0.7 and 0.5;
the survival rate remained unchanged. Find the population matrix after a further
three weeks (i.e. on day 42).
e With the arrival of warmer weather on day 42, the fertility rates increased to 0.3, 0.7,
0.8 and 0.5; the survival rate increased to 0.85. Suppose that an algal bloom is declared
if the concentration of algae reaches 100 000 cells per millilitre of water. Using trial
and error, find the day of the study on which an algal bloom was declared.

Chapter
12
Revision
Revision: Matrices

Revision 596 Chapter 12 Revision: Matrices
12A Exam 1 style questions: Matrices

 
1 4  
  2 3 4
1 Matrix A =2 3 and B = 
  Matrix Q = A × B. The element in row i and
1 0 6
  
6 2
column j of matrix Q is qi j . Element q33 is determined by the calculation
A 6×2+2×1 B 2×4+3×2 C 1×4+4×6
D 6×3+2×0 E 6×4+2×6
2 Matrices P and W are defined below.

   
0 1 0 0 D
   
0 0 1 0  O
P =  W =  
   
0 0 0 1  G 
   
1 0 0 0 S
  
Pn W = W for n =
A 1 B 2 C 3 D 4 E 5
3 The element in row i and column j of matrix A is ai j . A is a 3 × 3 matrix. It is

constructed using the rule ai j = 2i + 4 j. A is
     
 6 10 14  6 10 14  6 10 8 
     
A  8 12 16 B 10 14 18 C 10 12 14
     
10 14 18 8 12 16 14 16 18
     
   
14 16 18  6 10 14
   
D  6 10 8  E 10 12 18
   
10 12 14 14 14 20
   
 
0 0 1 0
 
0 1 0 0
4 O T S P 
  =
1 0 0 0
 
0 0 0 1


A S T O P B P O S T C T O P S

D O P S T E T O S P
 
1
 
0
 

 
5 1 0 1 0 0 × 0 equals
 
1
 
1
A [0] B [1] C [2] D [3] E [5]
12A Exam 1 style questions: Matrices 597
Revision
6 Matrix A has four rows and three columns.
Matrix B has three rows and four columns.
Matrix C = B × A has:
A two rows and three columns B three rows and two columns
C three rows and three columns D four rows and two columns
E four rows and three columns
 
14
 
 3 
7 Matrix N =  . The matrix P is the permutation matrix such that the 4 × 1 matrix
12
 
2
 
M = P × N has the smallest value at the top and the elements are in increasing order as
you go down the matrix. The matrix P is
     
0 0 1 0 0 1 0 0 0 0 1 0
     
0 1 0 0 0 0 1 0 0 1 0 0
A   B   C  
1 0 0 0 1 0 0 0 0 0 0 1
     
0 0 0 1 0 0 0 1 1 0 0 0
     
   
0 0 0 1 0 0 0 1
   
0 1 0 0 0 1 0 0
D  E 
   
0 0 1 0 1 0 0 0
   
1 0 0 0 0 0 1 0
   
8 Matrix X is of order p × q and matrix Y is of order q × r. The matrix products XY −1

and X −1 Y are both defined:
A for no values of p, q or r B when p = r
C when p = q = r only D when p = q only
E for all values of p, q and r
9 The diagram opposite is to be represented by a 2 3

matrix A, where:
element = 1 if the two points are joined by a line
element = 0 if the two points are not connected. 1 4
The matrix A is:
     
1 0 0 1 1 0 0 1 1
       
2 0 0 1 1 0 1 1 0 1 2
A 1 2 3 4 B   C   D   E 
    
3 1 1 0 1 0 1 1 1 3 4

     
4 1 1 1 0 1 0 1 0
 

10 The order of three matrices: A, B and C are respectively

A:3×5 B:3×4 C :4×5
The transpose of matrix A, for example, is written as AT . What is the order of the
product (C T × BT ) × A?
A 4×3 B 3×4 C 4×5 D 5×5 E 3×3
11 A and B are n × n matrices. Which of the following is not always true?

A (AB)T = BT AT B (A − B)(A + B) = A2 − B2
C (AT )T = A D A+B= B+A
E A × kB = kAB where k is a real number.
12 A is a 5 × 5 matrix.
B is a 8 × 5 matrix.
Which one of the following matrix expressions is defined?
A ABT B A3 B C BA + 3A
−1 2
D A(BA) E A − BA
The following information is needed for Questions 13 to 15.

   
100 0.1 0.8
The recurrence relation S 0 =  , S n+1 = T S n , where T =  , can be used to
100 0.9 0.2
generate a sequence of matrices. In this sequence:
13 S1 =
         
 90  100 110 120 140
A   B   C   D   E  
110 100 90 80 60
    
14 S 5 is closest to:
         
 90   93.1  95.5 107.9 106.9
A   B   C   D   E  
110 106.9 104.5 92.1 93.1
    
15 The steady-state matrix is closest to:

         
 93.0  93.6  94.1 106.4 107
A   B   C   D   E  
107.0 106.4 105.9 93.6 93
    
 
0 0 0 1
 
1 0 0 0
16 Consider the permutation matrix P =  The row matrix X =
 
0 1 0 0
 
0 0 1 0
 

B O A R is permuted by finding the product XP2 . The result is


Revision
A O A R B B A O R B C A R B O

D A B R O E B O A R
People in a suburb are very fitness conscious but get sick of going to the same gymnasium. It
is found that they go to four gymnasiums A, B, C and D but there are changes in attendance
every month. It is found that following transition matrix can be used to predict the number
of people at each of the four gymnasiums each month.
This month
A B C D 
300

 
0.30 0.25 0.1 0.1  
A 300
S 0 = 
 
T = 0.45

0.5 0.15 0.1 B Next month 300
   
0.13 0.2 0.45 0.2 C
300
  

0.12 0.05 0.3 0.6 D
17 Three hundred people go to gym C in September. How many of these people are in
gym A in October.
A 280 B 290 C 30 D 320 E 330
18 This transition matrix predicts that, in the long term, the people
A go only to A B go only to B
C go to only A and C D go only to B and D
E continue to visit all four gyms
19 After 12 months the total number of customers at gymnasiums B and C is closest to

A 500 B 600 C 620 D 650 E 700
20 A town has two hardware shops: Fairtrade (F) and Bungles (B). The percentage of
shoppers at each shop changes from day to day, as shown in the transition matrix T.
Today
F B
 
F 65% 70%
Tomorrow  
B 35% 30%
On a particular Monday, 35% of shoppers went to Fairtrade. The matrix recursion
relation S n+1 = T S n is used to model this situation. The percentage of shoppers who go
to Fairtrade on Wednesday of the same week is closest to
A 65% B 66% C 67% D 68% E 69%

21 The matrix gives the results of a table tennis round robin competition between five
players: A, B, C, D and E. A ‘1’ indicates a win of ’row’ over ‘column’.
Loser
A B C D E
 
 0 0 0 1 0 A
 
 1 0 1 0 1 B
 
Winner  1
 0 0 0 0 C
 
 0 1 1 0 0 D
 
1 0 1 1 0 E
When the sum of the one-step and two-step dominances is used to rank the players in
this competition, the ranking is:
A B, E, D, A, C B B, E, C, D, A
C B, E, D, C, A D E, B, D, A, C
E E, B, D, C, A
22 A taxi company has two depots at A and B. They always keep x taxis at A and z taxis at
B. The transition matrix T shows how the taxis change their nightly location.
Today
A B
 
A 65% 70%
B 35% 30%
On a particular night 14 taxis came from depot B to depot A. How many taxis in the
fleet?
A 24 B 36 C 42 D 60 E 80
23 The matrix S n+1 is determined from the matrix S n using the recurrence relation
S n+1 = T × S n − C, where
     
0.3 0.8 0.4 80 58
     
T = 0.6 0.1 0.3 , S 0 = 60 , S 1 = 41
     
0.1 0.1 0.3 20 17
and C is a column matrix. Matrix S 2 is equal to
         
58 35 14.85 30.5 59.5
         
A 41 B 25 C  8.95  D 26.5 E 41.5
         
17 12 4.2 15.5 15
         

Revision
24 A group of people travel every weekday but they have a choice of the type of transport
they use: train (T), bus (B) or Car (C). They change from day to day according to the
transition matrix.
Today
T B C
 
T 65% 70% 50%
 
Tomorrow B 20% 10% 25%
 
C 15% 20% 25%
On Monday 30% take a car and 70% take the train and no one takes the bus. What is
the percentage of people who are not expected to change their mode of transport from
Tuesday to Wednesday?
A 50.45% B 46% C 48.45% D 45.975% E 100%
25 The transition matrix that can be used to 45%

represent the information in the diagram
shown is:
30% B 25%
A B C
  30% 20%
A To A 0.30 0.30 0.55 25% 35%
A C
 
B 0.25 0.45 0.25
  45%
C 0.45 0.35 0.20
55%
A B C A B C
 
B To A 0.30 0.30 0.55
 
C To A 0.75 0.10 0.15
   
B 0.60 0.05 0.35 B 0.60 0.05 0.35
   
C 0.1 0.30 0.20 C 0.50 0.30 0.20
A B C A B C
   
D To A 0.75 0.05 0.15 E To A 0.30 0.30 0.55
   
B 0.10 0.60 0.20 B 0.15 0.35 0.50
   
C 0.45 0.35 0.20 C 0.45 0.35 0.20

12B Exam 2 style questions: Matrices

 
1.316 US dollar rate
 
1 At some stage, the matrix C = 1.818 Euro rate
 
0.167 HK dollar rate
could be used to convert US dollars (US$), European euro (e) and Hong Kong dollars
(HK$) into Australian dollars (A$).
a What is the order of matrix C?
You return from an overseas trip with US$102, e262
and HK$516.

We can write this information as the matrix H = 102 262 516 .
b What is the order of matrix H?
c Is the matrix product HC defined? Why?
d i Evaluate the matrix product HC.
ii What does the matrix product represent and why?
 
125 216 54 
 
e Matrix M =  0 34 453  gives the amounts in US dollars, euros and HK

 
0 356 0

dollars that three other people want to change into Australian dollars. The rows
represent people. The columns indicate the amounts of each currency they have.
Use the conversion matrix C and matrix multiplication to generate a matrix that
displays the amounts of Australian currency that each person will receive.
2 Lake Blue and Lake Green are two small lakes connected by a channel. This enables
fish to move between the two lakes on a daily basis. Research has shown that each day:
67% of fish in Lake Blue stay in Lake Blue
33% of fish in Lake Blue move to Lake Green
72% of fish in Lake Green stay in Lake Green
28% of fish in Lake Green move to Lake Blue.
a Construct a transition matrix, T , of the form:
From
Blue Green
To
 
Blue  
 
Green
to describe this situation.

b Today there are currently 4000 fish in Lake Blue and 6000 fish in Lake Green. Write
down a column matrix, S 0 , that describes this situation.
c How many fish do you expect to be in each lake tomorrow?
d How many fish do you expect to be in each lake in 3 days’ time?
e In the long term, how many fish do you expect to be in each lake?

12B Exam 2 style questions: Matrices 603
Revision
3 The life cycle of a type of insect can be descibed by a Leslie matrix. The stages of life
which are used are
Egg E Juvenile J Young adult Y adult A
From stage of life
E J Y A
 
 0 0 20 30  E
 
0.5 0 0 0  J
L =  To next stage of life
 
 0 0.1 0 0  Y
 
0 0 0.05 0 A

 
897 E
 
438 J
The initial population is described by the 4 × 1 state matrix S 0 =  
 43  Y
 
2 A
 
The time period in this model is one week and the state matrix after n weeks can be
determined by S n+1 = LS n .
a From the Leslie matrix complete the life-cycle diagram.
E J Y A
b How many eggs are produced in total each week?
c How many insect eggs are there after one week?
d How many weeks pass before there are more than 1000 eggs?
e What percentage of young adult insects become adult insects each week?
f How many insects of every type (including eggs) are there after?
i 8 weeks ii 9 weeks
Give answers correct to the nearest whole number.
g It is known that after some weeks the rate of increase per week of the entire
population (including eggs) is very close to being a constant. Use the results of f to
give an estimate of this rate per week as a percentage correct to the nearest percent.
(That is, in the form a%, where a is a whole number.)
4 The following transition matrix, T , is used to help predict attendance at a weekly club
meeting
This meeting
attend not attend
 
 0.80 0.30  attend
T =   Next meeting
0.20 0.70 not attend

S 1 is the attendance matrix for the first club meeting of the year.
 
110 attend
S 1 =  
40 not attend
S 1 indicates that 110 club members attended the first meeting and 40 club members did
not attend the first meeting.
a Use T and S 1 to:
i determine S 2 , the attendance matrix for the second meeting.
ii predict the number of club members attending the third meeting.
b Write down a matrix equation for S n in terms of T , n and S 1 .
c How many weeks does it take for the attendance to fall below 91?
d In the long term, how many club members are predicted to attend meetings?
5 A chemist wholesaler stocks three brands of hand sanitizer Cleanup (C), Loveneasy (L)
and Orama (O). The number of half litre bottles of these sanitizers sold in March 2022
is shown in matrix A below.
 
3000 C
 
A = 1500 L
 
2500 O
a i What is the order of matrix A?
ii The wholesaler expected that in April 2022 the sales of all three brands of
sanitizer would increase by 20%. She multiplied matrix A by a real number, k, to
determine the expected volume of sales for April. What is the value of k?
b A small chain of 4 chemists operate through this wholesaler and communicate with
each other rather inefficiently through an online facility. The communication links
are shown in this communication matrix receiver
A B C D
 
A  0 1 1 0 
 
B  1 0 0 0 
M = sender  
C  1 0 0 1 

 
D 0 1 0 0
the ‘1’ in row A, column B indicates that A can send information to B.
the ‘0’ in row D, column C indicates that D cannot send information to C.
i Which pairs of chemists can send information directly to each other?
ii D needs to send documents to C. What is the sequence of communication links
that will successfully get the information from D to C?

Revision
iii Matrix M 2 below shows the number of two-step communication links between
each pair of chemists
receiver
A B C D
 
A  2 0 0 1 
 
B
 0 1 1 0 
M 2 = sender 
 

C  0 2 1 0 
 
D 1 0 0 0
Communication from C to B in two different two-step communication links is
possible. List each two-step communication link for this pair.
c The total purchases of the three hand sanitizers, Cleanup (C), Loveneasy (L) and
Orama (O), change from month. Let T denote the transition matrix and S n represent
the state matrix describing the number of shoppers buying each brand n months
after June 2022.
this month
C L O
 
 0.7 0.8 0.5 C
 
T = 0.05 0.1 0.2 L next month
 
0.25 0.1 0.3 O
The initial state matrix S 0 below shows the number of shoppers at the four chemists
who bought each brand of hand sanitizer in June 2022.
 
200 C
 
S 0 = 150 L
 
250 O
i Calculate S 1 .
ii How many of the shoppers bought the same sanitizer in July as they did in June
2022?
iii Consider the shoppers who were expected to buy Cleanup in September 2022.
What percentage of these shoppers also bought Cleanup in August 2022?
d The shopping habits changed over the months. A rule to model this situation is
S n+1 = T × S n + B, where S n represents the state matrix describing the number of
shoppers n months after June 2021.
 
200 C
 
Here T is as above and S 0 = 150 L
 
250 O
 
405 C
 
If S 1 =  90  L find B.
 
165 O

6 In a cinema complex there are four cinemas A, B, C and D. They vary in the number
of seats, the standard of seating and other amenities. The number of empty (E) and
occupied seats (O) on a Friday afternoon is shown in the matrix below.
E O
 
A 30 60 
 
B 60 120
M=  
C 20 50
 
 
 
D 10 85
a What is the order of matrix M?
b What is the
i total number of seats
ii percentage of seats which are occupied on the Friday afternoon.
The cost of admission to each of the cinemas is given in the following matrix
A B C D
Q = $8.50 $12.60 $18.00 $25.80
c The total payments for this Friday afternoon can be determined by the calculation
T otal = Q × L, where L is a 4 × 1 matrix. Write the matrix L with its entries and
calculate the matrix T otal.
7 A commercial art school offers classes in pottery (P), sculpture (S), drawing (D) and
weaving (W). Students are allowed to change activities every month. In January 2022
the number of students in each class can be described by the state matrix and the
movement from one month to the next is described by the transition matrix T .
P S D W
   
30 P 0.6 0.2 0.4 0.1
   
30 S 0.2 0.3 0.1 0.6
S 0 =   T=  
30 D 0.1

0.2 0.2 0.2
   
30 W 0.1 0.3 0.3 0.1
 
Let S n be the state matrix n months after January 2022. Use S n = T n S 0

a Determine S 1 and S 2 .
b The number in each class stabilises quite quickly according to this model. Find
S 10 and give the numbers in each class (rounded to the nearest whole number) after
10 months.
c How many of those doing pottery moved to weaving from January to February?
d How many people stayed in the same activity from January to February?
e How many of those doing pottery moved to weaving from February to March?
f How many people stayed in the same activity from February to March?

Revision
8 A local take-away shop popular with students sells hamburgers (H), fish and chips
(F) and sandwiches (S ). The number of each item sold over three weeks is shown in
matrix M.
H F S
 
160 200 50 week 1
 
M = 180 210 55 week 2
 
210 240 80 week 3
a How many hamburgers were sold in these three weeks?
b What does the element m23 indicate?
c the total sales in dollars for three weeks for each of these items is given in the matrix
below.
 
H 8250
C = F 9100
 
 
S 2220
Determine the unit cost of each of these items.
d The matrix expression shown gives the total cost of all hamburgers and sandwiches
in these three weeks.
L×C
Matrix L is a 1 × 3 matrix. Write down matrix L.

Chapter
13
Graphs, networks and
trees: travelling and
connecting problems
Chapter objectives
I What is a graph?
I How do we identify the features of a graph?
I How do we draw a graph?
I How do we apply graphs in practical situations?
I How do we construct an adjacency matrix from a graph?
I How do we define and draw a planar graph?
I How do we identify the type of walk on a graph?
I How do we find the shortest path between two vertices of a graph?
I How do we find the minimum distance required to connect all vertices of a
graph?
In this chapter graphs and their use as networks representing connections between
objects will be introduced, in addition to exploring their properties and applications.
Problems involving networks will be investigated and you will learn unique algorithms
such as Djikstra’s and Prim’s to solve such problems.

13A Graphs and networks 609
13A Graphs and networks

Learning intentions
I To be able to define and identify a graph, vertex, edge and loop.
I To be able to find the degree of a vertex and the sum of degrees.
I To be able to describe the features of a graph.
I To be able to define and identify a planar graph and its faces.
I To be able to apply Euler’s formula and use it to verify if a graph is planar.
Representing connections with graphs

There are many situations in everyday life that involve connections between people or
objects. Towns are connected by roads, computers are connected to the internet and people
connect to each other through being friends on social media. A diagram that shows these
connections is called a graph.
Edges and vertices

Six people – Anna, Brett, Cora, Dario, Ethan and Frances – have connections on a social
media website. The graph shows these connections.
Anna is a friend of Brett, Ethan and Frances. Frances
Brett is a friend of Anna, Cora and Frances. Anna Brett
Cora is a friend of Brett, Dario and Ethan. Cora

Ethan
Dario
Dario is a friend of Cora and Frances.
Ethan is a friend of Anna and Cora.
Frances is a friend of Anna, Brett and Dario.
The graph shows each of the people as a dot called a vertex. The vertices (plural of vertex)
are joined together by a line that indicates the social media friendship between the people.
The lines that join the vertices in the graph are called edges.
Degree of a vertex
Anna has three friends. The vertex representing Anna has Frances
three edges attached to it, connecting Anna to one of her Anna Brett
friends. The number of edges attached to a vertex is called
the degree of that vertex. Ethan Cora
Dario
The degree of the vertex representing Anna is odd, because there is an odd number of edges
connected to it. The degree of the vertex representing Dario is even because there is an even
number of edges connected to it.
In symbolic form, we can let the letter A represent the vertex for Anna. The degree of this
vertex can be written as deg(A). In this graph, deg(A) = 3.
610 Chapter 13 Graphs, networks and trees: travelling and connecting problems
Loops
Imagine that Ethan is able to add himself as a friend on the social media website.
The edge representing this connection would connect the Frances
vertex representing Ethan, E, back to itself. This type of Brett
Anna
edge is called a loop.
Ethan
Cora
A loop is attached twice to a vertex and so it will contribute (E)
Dario
two degrees. So deg(E) = 4.
Edges, vertices and loops

A graph consists of vertices joined by edges.
The number of edges attached to a vertex is called the degree of the vertex. The
symbolic form for the degree of vertex A is deg(A).
A loop connects a vertex to itself. Loops contribute two degrees to a vertex.
Describing graphs
Graphs that represent connections between objects can take different forms and have
different features. This means that there is a variety of ways to describe these graphs.
Simple graphs
Simple graphs do not have any loops. There are no C
duplicate or multiple edges either.
A D
E
Isolated vertex
A graph has an isolated vertex if there is a vertex B
that is not connected to another vertex by an edge.
C
A D
The isolated vertex in this graph is E, because it is
not connected to any other vertex by an edge. E
Degenerate graphs
Degenerate graphs have all vertices isolated. This means B
that there are no edges in the graph at all.
A
D
E
C

Connected graphs and bridges

A connected graph has every vertex connected to every B E G
other vertex, either directly or indirectly via other vertices.
A
The graph on the right is connected. A bridge is an edge D
F
in a connected graph that, if removed, will cause the graph C
to be disconnected. The graph on the right has a bridge
connecting vertex D to vertex E.
The graph on the right shows the bridge from vertex D to B E G
vertex E removed. There are now two separate sections of
A
the graph that are not connected to each other. D
F
C
Complete graphs
If there is an edge between every pair of vertices, the graph is B C
called a complete graph. Every vertex in the graph is connected
directly by an edge to every other vertex in the graph.
A D
Subgraphs
A subgraph is a part of a larger graph. All of the edges and B C
vertices in the subgraph must exist in the original graph.
A
If there are extra edges or vertices, the graph will not be a D
subgraph of the larger graph.
F
E
Graph 1
B B C
A A
D
B C
F F
E A
Graph 2 Graph 3 Graph 4
Graphs 2 and 3 above are subgraphs of Graph 4 above is not a subgraph of

graph 1. All of the vertices and edges in graph 1. There are two edges connecting
graphs 2 and 3 exist in graph 1. vertex A to vertex B, but in graph 1 there is
only one.

Example 1 Graphs
A connected graph is shown on the right. A D

C
a What is the degree of vertex C?
B
b Which vertices have a loop? F
E
c What is the degree of vertex F?
d A bridge exists between two vertices. Which vertices are they?
e Draw a subgraph of this graph that involves only vertices A, B and C.
a Count the number of times an edge The degree of vertex C is 4.
connects to vertex C. There are four deg(C) = 4.
connections.
b A vertex has a loop if an edge connects Vertex B and vertex F have loops.
it to itself.
c Count the number of times an edge The degree of vertex F is 5.
connects to vertex F. Remember that a deg(F) = 5
loop contributes two degrees.
d Look for an edge that, if removed, A bridge exists between vertex A and
would disconnect the graph. vertex C.
e There are a few possible answers for A
A C C A C
this question. Some are shown on the
right. B B
B
Equivalent (Isomorphic) graphs

All of the graphs shown in the diagram below contain exactly the same information. For
example, the edge between vertex E and C exists in all of them. The vertex A is connected to
B, D and C as well.
The location of the vertices and edges in B
the diagram are unimportant. As long as the C
A
connections are all represented accurately, the A
graph can be drawn in any way that you prefer. D E
B
The first of the graphs has some curved edges C
and the second has all straight edges. The third
has the vertices arranged in a straight line. D
E
The style that they are drawn in and the
position of the vertices relative to each
other is unimportant. It is important that A B C D E
the information contained in the graph – the
connections between the vertices – is correct.

All of these graphs are considered to be equivalent to each other because they all contain
identical information. Each has edges connecting the same vertices. Graphs that contain
identical information like this are called equivalent graphs or isomorphic graphs.
Planar graphs
The graph opposite has two edges that overlap. It is important to B
note that there is no vertex at the point of overlap of the edges. A
C
It can help to think of an edge as an insulated electrical wire. It
is quite safe to cross two such electrical wires because the wires D
E
themselves never touch and never interfere with each other. The
edges that cross over in this diagram are similar, in that they do not
intersect and do not interfere with each other.
If a graph has edges that cross, it may be possible to redraw A B
the graph so that the edges no longer cross. The edge between
vertices A and D has been moved, but none of the information in C
the graph has changed. Graphs where this is possible are called
E
planar graphs. If it is impossible to draw an equivalent graph D
without crossing edges, the graph is called a non-planar graph.
Example 2 Redrawing a graph in planar form
Show that this graph is planar by redrawing it so that B C

no edges cross.
A D
F E
1 Choose one of the edges that crosses over another B C
edge.
A D
F E
2 Remove it temporarily from the graph. B C
A D
F E
3 Redraw the edge between the same vertices but B C

without crossing over another edge.
A D
F E

Euler’s formula
Leonard Euler (pronounced ‘oiler’) was
C
one of the most prolific mathematicians B
D
of all time. He contributed to many areas
f1 f3
of mathematics and his proof of the rule f2
f4
named after him is considered to be the
beginning of the branch of mathematics A F
called topology. E
Faces
A planar graph defines separate regions of the paper it is drawn on. These regions are
enclosed spaces that you could colour in and these regions are called faces. An often-
forgotten face of a graph is the space outside of the graph itself, covering the infinite space
around it. This face is labelled f4 in the graph above.
The number of faces for a graph can be counted. In the graph shown above, there are four
faces, labelled f1 , f2 , f3 and f4 .
Euler’s formula
There is a relationship between the number of vertices, v, the number of edges, e, and the
number of faces, f, in a connected planar graph.
In words: number of vertices + number of faces = number of edges +2
In symbols: v + f = e + 2
Euler’s formula
For any planar graph:
v+ f =e+2
where v is the number of vertices, e is the number of edges and f is the number of faces
in the graph.
Example 3 Verifying Euler’s formula
For the graph shown on the right: B C

a redraw the graph into planar form E
b verfiy Euler’s formula for this graph.
A D

a Temporarily remove an B C
B C
edge that crosses another
E E
edge and redraw it so that
it does not cross another A D
A D
edge.
b Count the number of vertices, In the planar graph there are five vertices,
edges and faces. seven edges and four faces.
v+ f =e+2
5+4=7+2
9=9
Euler’s formula is verified.
Example 4 Using Euler’s formula
A connected planar graph has six vertices and nine edges. How many faces does the
graph have? Draw a connected planar graph with six vertices and nine edges.
a Write down the known values.
v=6 e=9
b Substitute into Euler’s formula and v+ f =e+2
solve for the unknown value.
6+ f =9+2
6 + f = 11
f = 11 − 6
f =5
This graph has five faces, labelled
f1 , f2 , f3 , f4 and f5 .
c Sketch the graph. D
C
Note: There are other possible graphs. E
B f f3 f F
f1 2 4
f5

616 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13A
Skill-
sheet Exercise 13A
Drawing and describing graphs

Example 1 1 This section of a road map can be considered town B
town A town C
a graph, with towns as vertices and the roads
connecting the towns as edges. town D
a Give the degree of: town H

i town A ii town B iii town H.
b What is the sum of the degrees of all the vertices of this graph?
c A bridge exists between two towns. Which towns are they?
d Draw a subgraph of this road map that contains only towns H, D and C.
2 Draw a graph that:

a has three vertices, two of which are odd
b has four vertices and five edges, one of which is a loop
c has six vertices, eight edges and one bridge
d has six vertices, two of which are odd, and contains a subgraph that is a triangle.
Equivalent graphs
3 In each question below, three graphs are isomorphic and the fourth is not. Identify the
graph which is not isomorphic to the others.
A
a i ii iii iv
B A
B B B
C C
A C A C
b i ii iii A B iv
A B C
A
C
B A
B C D
D C D D
B
c iA ii B C iii B iv B
C A C
E A
E D
C
D A
E D D E

13A 13A Graphs and networks 617
Drawing planar graphs

Example 2 4 Where possible, show that the following graphs are planar by redrawing them in a
suitable planar form.
a A B C b A B
C E
F E D D
c D d B
C E
A C
B F
A E D
Euler’s formula
Example 3 5 For each of the following graphs:
i state the values of v, e and f ii verify Euler’s formula.
a b c
d e f
6 For a planar connected graph, find:

a f, if v = 8 and e = 10 b v, if e = 14 and f = 4 c e, if v = 10 and f = 11.
Properties of graphs
Example 4 7 A connected planar graph has eight vertices and thirteen edges. Find the number of
faces of this graph.
8 A connected graph has five vertices and seven edges. Find the sum of the degrees of
the vertices.
9 Find the number of edges needed to make a complete graph with six vertices.

618 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13A

10 Consider the graph opposite.
The number of vertices with a degree of 4 is
A 1 B 2 C 3
D 4 E 5
11 A planar graph has four faces. The graph could have

A Seven vertices and seven edges B Seven vertices and four edges
C Seven vertices and five edges D Four vertices and seven edges
E Five vertices and seven edges
Use the following information for questions 12 and 13.
12 The number of faces in the graph above is

A 3 B 4 C 5 D 6 E 7
13 Consider the following five statements about the graph above:

The graph is planar. The graph contains a bridge.
It is a simple graph. The sum of degrees of the vertices is 16.
It is a complete graph.
How many of these statements are true?
A 1 B 2 C 3 D 4 E 5
14 Consider the graph shown opposite.

The minimum number of edges that must be added to
make this a complete graph is
A 1 B 2 C 6
D 8 E 12
15 The following graph with six vertices is a complete graph.

Edges are removed so that the graph will have the
minimum number of edges to remain connected. The
number of edges that are removed is
A 6 B 8 C 10 D 12 E 14

13B Adjacency matrices 619
13B Adjacency matrices

Learning intentions
I To be able to use an adjacency matrix to represent a graph.
Summarising the connections in a graph

A matrix can be used to summarise the information in a graph. A matrix that records the
number of connections between vertices of a graph is called an adjacency matrix.
A graph and the adjacency matrix for that graph are shown here.
B  A B C D E
 
A  0 1 1 0 1
C
 
A B 
 1 0 2 1 0

C  1 2 0 0 0
D
 
D  0 1 0 0 0
E  
E  1 0 0 0 0
The adjacency matrix has:

five rows and five columns, one for each vertex in the graph
row and column labels that match the vertices in the graph, A, B, C, D, E
a ‘0’ in the intersection of row A and column D because there is no edge connecting
A to D
a ‘0’ in the intersection of row A and column A because there is no edge connecting A to
itself; that is, there is no loop at vertex A
a ‘1’ in the intersection of row A column B because there is one edge connecting A to B
a ‘2’ in the intersection of row C and column B because there are two edges connecting
C to B.
The number of edges between every other pair of vertices in the graph is recorded in the
adjacency matrix in the same way.
Example 5 Drawing a graph from an adjacency matrix
Draw the graph that is represented by the  A B C D E

following adjacency matrix. A 
 
0 0 2 0 1 
 
B  0 0 2 1 0 
 
C  2 2 0 1 0 
 
D  0 1 1 1 0 
 
E  1 0 0 0 0 

1 Draw a dot for each vertex and label B
A to E. D
A
2 There is a ‘2’ in the intersection of
row A and column C. This means there
C
are two edges connecting vertex A and
E
vertex C. Add these to the graph.
3 Note the ‘1’ in the intersection of row D and column D. This shows that there is a loop
at vertex D.
4 Look at every intersection of row and column and add edges to the graph, if they do
not already exist.
Note: This graph is drawn as a planar graph, but this is not strictly necessary unless required by the
question.
Example 6
Construct an adjacency matrix that can be used to represent A B C
the graph opposite. This graph represents the ways that
three houses A, B and C are connected to three utility
outlets, gas (G), water (W) and electricity (E).
G W E
The convention used to enter the values is  A B C G W E 
the same as discussed above. A  0 0 0 1 1 1 
 
 
B  0 0 0 1 1 1 
 
C  0 0 0 1 1 1 
 
G  1 1 1 0 0 0 
 
W  1 1 1 0 0 0 
 
E  1 1 1 0 0 0 
The graph in Example 6 is called a bipartite graph as the set of vertices is separated into
two sets of objects Houses (A, B, C) and Utility outlets (G, W, E) with each edge connecting
a vertex in each set. You will meet bipartite graphs again in Chapter 14 when studying
allocation problems.
Adjacency matrices
The adjacency matrix A of a graph is an n × n matrix in which, for example, the entry in
row C and column F is the number of edges joining vertices C and F.
A loop is a single edge connecting a vertex to itself.
Loops are counted as one edge.

13B 13B Adjacency matrices 621
Exercise 13B
Writing adjacency matrices

1 For each of the following graphs, write down the adjacency matrix.
a B D b A B c A B
C D
A
C D
C
d A B e A f B
B E A
D
D C C
D F C
Drawing graphs from adjacency matrices

Example 5 2 Draw a graph from each of the following adjacency matrices.
a  A B C b  A B C D c  A B C D
A  0 1 1  A  1 0 0 0  A  0 1 2 1 
     
     
B  1 0 1  B  0 0 1 1  B  1 0 1 1 
     
C  1 1 0  C 
 0 1 1 0  C 
 2 1 0 0 
 
D 0 1 0 0  D 1 1 0 0 
3 The adjacency matrix on the right has a row and column  A B C
for vertex C that contains all zeros. What does this tell you A 

0 1 0 

 
about vertex C? B  1 0 0 
 
C  0 0 0 
4 Every vertex in a graph has one loop. What feature of the adjacency matrix would tell
you this information?
5 A graph has five vertices: A, B, C, D and E. It has no duplicate edges and no loops. If
this graph is complete, write down the adjacency matrix for the graph.

622 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13B

6 The map opposite shows the pathways between
A C
five buildings: A, B, C, D and E. An adjacency
matrix for the graph that represents this map is
B
formed. The number of zeros in this matrix is
A 9 B 10 C 11
D 12 E 13
E D
7 The adjacency matrix opposite shows the number  A B C D

of pathways between four points A, B, C and D. A 

0 3 0 2 

 
A graph that could be represented by the B  3 0 2 0 
 
adjacency matrix is C  0 2 1 1 
 
D 2 0 1 0 
A B B B C
B
A A
C C A
C
D D
D
D E B
B
A
A
C C
D
D
Use the following information to answer questions 8 and 9.

The map opposite shows the pathways between five schools: A
A, B, C, D and E.
E
An adjacency matrix for the graph that represents this map
is formed.
B
8 Of the 25 elements in the adjacency matrix, the
number ‘1’ appears D
A 7 times B 8 times C 9 times C
D 10 times E 11 times
9 Of the 25 elements in the adjacency matrix, the numbers ‘2’ or ‘3’ appear
A 6 times B 7 times C 8 times D 9 times E 10 times
13B 13B Adjacency matrices 623
10 A graph has four vertices A, B, C, and D.  A B C D

The adjacency matrix for this graph is shown A 

0 1 1 1 

 
opposite. Which one of the following statements B  1 0 0 1 
 
about this graph is not true? C  1 0 1 0 
 
D 1 1 0 0 
A The graph is connected. B The graph contains a loop.

C The graph contains multiple edges. D The graph is planar.
E The graph contains a bridge
11 The number of faces of the graph represented by the adjacency matrix above is
A 2 B 3 C 4 D 5 E 6
12 The map opposite shows all the road connections P

between six towns, P, Q, R, S , T and U. The road
U Q
connections could be represented by the adjacency
matrix
T
R
A P Q R S T U
 B P Q R S T U C
 
P Q R S T U

P  0 1 1 1 1 1  P  0 1 1 1 1 1 
 P  0
 1 1 2 1 2 
    
Q  0 1 1 1 1 1  Q  1 0 1 0 0 0 

Q  1 0 1 0 0 0     

R  1 1 0 1 1 1 
 R  0 1 1 1 1 1  R  1 1 0 2 1 1 
   
S  1

0 1 1 1 1 
 S  1 0 1 1 2 2  S  2 0 2 1 2 2 
   
  T  1 0 1 2 0 1  T  1 0 1 2 0 1 
T  1 0 1 1 0 1      
  U  1 0 1 2 1 0  U  2 0 1 2 1 0 
U  1 0 1 1 1 0 
D  P Q R S T U E P Q R S T U
  
P  0 1 1 2 1 2  P  0 1 1 2 1 2 
   
Q  1 0 1 0 0 0  Q  1 0 1 0 0 0 
   
R  1 1 0 2 1 1  R  1 1 0 2 1 1 
   
S  2 0 2 0 3 2  S  2 0 2 1 3 2 
   
T  1 0 1 3 0 1  T  1 0 1 3 0 1 
   
U  2 0 1 2 0 0  U  2 0 1 2 1 0 

13C Exploring and travelling

Learning intentions
I To be able to identify a walk as a trail, path, circuit or cycle.
I To be able to identify a walk as an Eulerian trail, Eulerian circuit, Hamiltonian path or
Hamiltonian cycle.
I To be able to use the degrees of the vertices to identify if an Eulerian trail or circuit is
possible.
Travelling
Graphs can be used to model and analyse problems involving exploring and travelling.
These problems include minimising the distance travelled or time taken between different
locations using different routes. For example, a courier driver would like to know the
shortest route to use for deliveries, and a tour guide would like to know the quickest route
that allows tourists to see a number of sights without retracing their steps.
To solve these types of problems, you will need to learn the language we use to describe the
different ways of navigating through a graph, from one vertex to another.
Walks, trails, paths, circuits and cycles

The different ways of navigating through A
graphs, from one vertex to another, are
described as walks, trails, paths, circuits and
B C D
cycles.
The graph opposite will be used to explain
and define each of these terms. E F G
Walks
A walk is a sequence of edges, linking successive A start
vertices in a graph.
A walk starts at one vertex and follows any route to B C D
finish at another vertex.
The red line in the graph opposite traces out a walk. end
This walk can be written down by listing the vertices in E F G
the order they are visited: A−C−A−D−G.

13C Exploring and travelling 625
Trails
A trail is a walk with no repeated edges. A end
The red line in the graph opposite traces out a trail. This
start
trail can be written down by listing the vertices in the B C D
order they are visited: B−E−F−C−B−A.
Note: There are no repeated edges in this trail, but one vertex (B) is
repeated.
E F G
Paths
A path is a walk with no repeated edges and no repeated A start
vertices.
end
The red line in the graph opposite traces out a path. This B C D
path can be written down by listing the vertices in the
order they are visited: A−D−C−F−E−B.
E F G
Circuits
A circuit is a trail (no repeated edges) that starts and A
ends at the same vertex. Circuits are also called closed
trails. start
B C D
The red line in the graph opposite traces out a circuit. end
This circuit can be written down by listing the vertices in
the order they are visited: A−C−F−G−D−C−B−A.
E F G
Note: There are no repeated edges in this circuit, but one vertex,
C, is repeated. The start and end vertices are also repeated because
of the definition of a circuit.
Cycles
A cycle is a path (no repeated edges, no repeated vertices) A
that starts and ends at the same vertex. The start and end
vertex is an exception to repeated vertices. Cycles are also
B C D
called closed paths.
The red line in the graph opposite traces out a cycle. This start
end
cycle can be written down by listing the vertices in the E G
F
order they are visited: F−E−B−C−F.
Note: There are no repeated edges and no repeated vertices in this
cycle, except for the start and end vertices.

Walks, trails, paths, circuits and cycles

A walk is a sequence of edges, linking successive vertices in a graph.
A trail is a walk with no repeated edges.
A path is a walk with no repeated edges and no repeated vertices.
A circuit is a trail (no repeated edges) that starts and ends at the same vertex.
A cycle is a path (no repeated edges and no repeated vertices) that starts and ends at
the same vertex.
Example 7 Identifying types of walks
Identify the walk in each of graphs below as a trail, path, circuit, cycle or walk only.
a B b B
start start
D D
A A
C end C
F F
E E
c B start d B end
D D
A A start
C C
end F
F
E E
Solution
a This walk starts and ends at the same vertex so it is either a circuit or a cycle. The walk
passes through vertex C twice without repeated edges, so it must be a circuit.
b This walk starts and ends at the same vertex so it is either a circuit or a cycle. The walk
has no repeated vertex or edge so it is a cycle.
c This walk starts at one vertex and ends at a different vertex, so it is not a circuit or a
cycle. There is one repeated vertex (B) and no repeated edge, so it must be a trail.
d This walk starts at one vertex and ends at a different vertex so it is not a circuit or a
cycle. There are repeated vertices (C and E) and repeated edges (the edge between C
and E), so it must be a walk only.
Eulerian trails and circuits

Trails and circuits that follow every edge, without duplicating any edge, of a graph are
called Eulerian trails and Eulerian circuits. Eulerian trails and circuits are important for
some real-life applications. If, for example, a graph shows towns as vertices and roads as
edges, then being able to identify a route through the graph that follows every road can be
important for mail delivery, or for© checking the condition of the roads.
Peter Jones et al 2023 Cambridge University Press
ISBN 978-1-009-11041-9
13C Exploring and travelling 627
Eulerian trails and circuits exist under easily identified conditions.

Eulerian trails
An Eulerian trail follows every edge of a graph.
An Eulerian trail will exist if the graph:
is connected
has exactly zero or two vertices that have an odd degree.
If there are no odd vertices, the Eulerian trail can start at any vertex in the graph.
if there are two odd vertices, the Eulerian trail will start at one of the odd vertices and
finish at the other.
Eulerian circuits
An Eulerian circuit is an Eulerian trail (follows every edge) that starts and ends at the
same vertex.
An Eulerian circuit will exist if the graph:
is connected
has vertices that all have an even degree.
An Eulerian circuit can start at any of the vertices.
Note: If a graph has more than two odd-degree vertices, neither an eulerian trail nor an eulerian circuit
exists.
Hamiltonian paths and cycles

Paths and cycles that pass through every vertex of a graph only once are called Hamiltonian
paths and Hamiltonian cycles, named after the mathematician William Rowan Hamilton.
Hamiltonian paths and cycles have real-life applications to situations where every vertex of
a graph needs to be visited, but the route taken is not important. If, for example, the vertices
of a graph represent people and the edges of the graph represent email connections between
those people, a hamiltonian path would ensure that every person in the graph received a
message intended for everyone.
Unlike Eulerian trails and circuits, Hamiltonian paths and cycles do not have a convenient
rule or feature that identifies them. Inspection is the only way to identify them.

Hamiltonian paths
A Hamiltonian path visits every vertex of a graph.
Hamiltonian cycles
A Hamiltonian cycle is a Hamiltonian path (every vertex) that starts and ends at the same
vertex.
Note: Inspection is the only way to identify Hamilton paths and cycles.
Remember: Eulerian trails and circuits do not repeat edges. Hamiltonian paths and cycles do not
repeat vertices.
Hint: To remember the difference between eulerian and hamiltonian travels, remember that eulerian refers to
edges, and both start with ‘e’.
Example 8 Eulerian and Hamiltonian travel
A map showing the towns of St Andrews,

Kinglake, Yarra Glen, Toolangi and
Healesville is shown on the right.
Consider only the yellow routes in your
answer. St Andrews and Yarra Glen are
considered connected, although this route
is not fully shown in the image.
a Draw a graph with a vertex
representing each of these towns and
edges representing the direct road
connections between the towns.
b Explain why an Eulerian trail, but not an Eulerian circuit, is possible through this
graph.
c Write down an Eulerian trail that begins at Toolangi.
d Write down a Hamiltonian cycle that begins at Healesville.
a A road connection exists between: Kinglake
Toolangi
St Andrews and Kinglake
St Andrews and Yarra Glen
St Andrews
Kinglake and Yarra Glen
Kinglake and Toolangi Yarra Glen
Healesville
Yarra Glen and Toolangi
Yarra Glen and Healesville
Healesville and Toolangi.
b The graph has two odd-degree vertices There are exactly two odd-degree vertices
(Toolangi and Kinglake). in this graph. An Eulerian trail will exist,
but an Eulerian circuit does not.
c There are a few different answers to An Eulerian trail, starting at Toolangi
this question. One of these is shown. is: Toolangi−Healesville−Yarra
Glen−Toolangi−Kinglake−Yarra Glen−St
Andrews−Kinglake
d There are two different answers to this A Hamiltonian cycle that begins at
question. One of these is shown. Healesville is: Healesville−Yarra Glen−St
Andrews−Kinglake−Toolangi−Healesville

13C 13C Exploring and travelling 629
Skill-
sheet Exercise 13C
Describing travels
Example 7 1 Identify the walk in each of the graphs below as a trail, path, circuit or walk only.
a start b
end
start
end
c d
start
start
end
end
e B f
E
start
end
end
A
D
start
C
F
2 Using the graph opposite, identify B

the walks below as a trail, path,
F
circuit, cycle or walk only.
A E
a A−B−E−B−F
b B−C−D−E−B
C
c C−D−E−F−B−A
D
d A−B−E−F−B−E−D
e E−F−D−C−B
f C−B−E−F−D−E−B−C−A
g F−E−B−C−A−B−F
h A−C−D−E−B−A

630 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13C

3 i Identify whether each graph below has an Eulerian circuit, an Eulerian trail, both or
neither.
ii Name the Eulerian circuits or trails found.
a A B b B C c A B D F
F
C G
E C E
H
E D A D
d C e E A
B D
B
F
A E D C

4 List a Hamiltonian cycle for each of the following.
a A B C b B C c B
E A
D F C D
D A
F E
G H I E
5 List a Hamiltonian path for this graph, starting at F A B
and finishing at G.
F G
E H
D C
Applications
Example 8 6 Four children each live in a different town. The A
diagram opposite is a map of the roads that link
the four towns, A, B, C and D. B
D
C
a How many different ways can a vehicle travel between town A and town B without
visiting any other town?

13C 13C Exploring and travelling 631
b How many different trails are there from town A to town D?

c Draw this map as a graph by representing towns as vertices and each different route
between two towns as an edge.
d Explain why a vehicle at A could not follow an Eulerian circuit through this graph.
7 The following graph shows the roads linking nine Victorian country towns.
Streatham
Glenthompson
Wickliffe Lake Bolac
Nerrin Nerrin
Woorndoo
Chatsworth
Hexham Mortlake
a Verify Euler’s formula for this graph.

Cycling enthusiasts from the nine towns are planning a race that uses the roads linking
the towns.
b The organisers who live in Lake Bolac want to visit each town to gain support for
the race. They plan to visit each town on the same day but not pass through any
town more than once. They will start and finish at Lake Bolac.
i What name is given for the route they plan to take?
ii Identify the two routes they can follow.
c In planning the race route, the organisers would like the cyclists to travel along each
road linking the towns, but only once, and start and finish at Lake Bolac.
i What name is given for the type of route they would like the cyclists to take?
ii Explain why this cannot be done.
d The race planning in part c above can start and finish at Lake Bolac with only one
road being travelled along twice.
i Which road is this?
ii Identify one possible route the race can follow starting and finishing at Lake
Bolac, with only one road being travelled along twice.

632 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13C
8 A graph has six vertices, A, B, C, D, E  A B C D E F
and F. The adjacency matrix for this A  0 0 1 1 0 0
 
 
graph is shown opposite. B  0 0 0 0 1 1
 
a Is the graph connected? C  1 0 0 2 1 0
 
D  1 0 2 0 0 1
b Is the graph planar?  
E  0 1 1 0 0 0
c Does the graph contain a bridge? F 

0 1 0 1 0 1

d Does the graph contain an Eulerian

trail?
e Does the graph contain an Eulerian circuit?
f Does the graph contain a Hamiltonian path?
g Does the graph contain a Hamiltonian cycle?

What is the minimum number of edges that must be added for an
Eulerian circuit to exist?

10 Which one of the following graphs has an Eulerian circuit?
A B C
D E
11 Consider the graph opposite. Which one of the following A B C

is not a Hamiltonian cycle for this graph?
A ABCFEDA
B BADEFCB
F E D
C CDEFABC
D DEFACBD
E EFCBADE
12 An Eulerian trail for the graph opposite will be possible

if only one edge is removed. In how many different ways
could this be done?
A 1 B 2 C 3 D 4 E 5

13C 13D Weighted graphs and networks 633
Use the following graph to answer questions 13 and 14.
F B
E D C
13 Which one of the following is a Hamiltonian path for the graph above?
A AFEDBA B ABCDEF C AFEDBCBA
D ABDFEDCBA E FDEFABC
14 The graph above will have a Eulerian circuit if an edge could be added between the
vertices
A E and C B A and B C A and F
D A and D E F and B
13D Weighted graphs and networks

Learning intentions
I To be able to analyse a weighted graph.
I To be able to find the shortest path between two vertices for a network.
Weighted graphs
The edges of graphs represent connections between the vertices. Sometimes there is more
information known about that connection. If the edge of a graph represents a road between
two towns, we might also know the length of this road, or the time it takes to travel this road.
Extra numerical information about the edge that connects vertices can be added to a graph
by writing the number next to the edge. Graphs that have a number associated with each
edge are called weighted graphs.
The weighted graph in the diagram Croghon
on the right shows towns, represented
12 6 Melville 13 Kenton
by vertices, and the roads between
those towns, represented by edges. 20 8
Bartow 5
The numbers, or weights, on the 7 11
edges are the distances along the 9
roads. Stratmoore Osburn
Weighted graphs in which the weights are physical quantities, for example distance, time or
cost, are called networks.
634 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13D
Shortest path problems

When we know numerical information about the connections, travelling through a graph
will have extra considerations. If the weights of a network represent time, we can choose a
route that will allow us to travel in the shortest time. If the weights represent distance, we
can determine a route that will allow us to travel the shortest distance.
These types of problems involve finding the shortest path from one vertex to another.
In networks that have only a few vertices, it is often easy to find the shortest path between
two vertices by inspection. All of the possible route options should be listed, but it is
sometimes obvious that certain routes are going to be much longer than others.
Example 9 Finding the shortest path from one vertex to another
Find the shortest path from Bartow to Croghon

Kenton in the network shown on the
12 6 Melville 13
right. Kenton
20 8
Bartow 5
7 11
9
Stratmoore Osburn
1 List options for travelling B−S−M−O−K
from Bartow to Kenton. B−S−M−K
B−S−O−K
2 Add the weights for each B−S −M−O−K 7 + 8 + 5 + 11 = 31 km
route. B−S −M−K 7 + 8 + 13 = 28 km
B−S −O−K 7 + 9 + 11 = 27 km
3 Write your answer. The shortest path from Bartow to Kenton is 27 km
with route B−S −O−K.
Skill-
sheet Exercise 13D
Weighted graphs and networks

1 The network on the right shows towns C 10
15 D
A, B, C, D and E represented by
8
vertices. The edges represent road A 6 9
connections between the towns. The B 12
weights on the edges are the average 11
16
times, in minutes, it takes to travel E
along each road.

13D 13D Weighted graphs and networks 635
a Which two towns are 12 minutes apart by road?

b How long will it take to drive from C to D via B?
c A motorist intends to drive from D to E via B. How much time will they save if they
travel directly from D to E?
d Find the shortest time it would take to start at A, finish at E and visit every town
exactly once.
Shortest path by inspection

Example 9 2 By inspection, find the length of the shortest B F
4 6 7
path from A to E. D
8
A 3
E
5
3
3 The network on the right 8 D

B 10
shows the distance, in 9 H
5 6 8
kilometres, along walkways 8 12 I
that connect landmarks A E
6 8
A, B, C, D, E, F, G, H and I in a 4
10 G
national park. C 4
F
a What distance is travelled
on the path A–B–E–H–I?
b What distance is travelled on the circuit F–E–D–H–E–A–C–F?
c What is the distance travelled on the shortest cycle starting and finishing at E?
d Find the shortest path from A to I.
4 Determine the shortest path from S to F in the following weighted graphs. Write down
the length of the shortest path.
a A 6 C b A 9 D
3 4 1
S F 3 4
C 2
2 3 S
B 7 D 5 F
6
B
c A d A 7 G
3 E 5
7 D 3 2 4
4 6 6 2
5 S B 9 F
S B 1 F
7 4 2 6
8 H
8
8 4
2 E
C C D

636 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13D
5 In the network opposite, the 3

vertices represent small towns and 8 4 12
2 8 3
the edges represent roads. The 2
A
numbers on each edge indicate the 3
9 B
11 3
distances (in kilometres) between
8
15
towns. 14
Determine the length of the
shortest path between the towns
labelled A and B.

C
4 2
1
4
A G
1 D
3 E 8
3 5
4 2
B 4 H
6 2
F
6 What is the shortest path between C and E?

A C−A−E B C−D−A−E C C −D−G−E
D C −G−E E C −G−H −E
7 What is the shortest path between B and G?

A B−E −G B B−E −H −G C B−A−C −G
D B−A−D−G E B−A−D−C −G
A 37 B
20 26
10 9
47 C D
Home School
20 17
14
12 26
30 60
E 8 F

13D 13E Dijkstra’s algorithm 637
8 Victoria rides her bike to school each day. The edges of the network on the previous
page represent the roads that Victoria can use to ride to school from her home. The
numbers on the edges give the time taken, in minutes, to travel along each road.
The fastest Victoria can ride from home to school is
A 80 minutes B 81 minutes C 83 minutes
D 84 minutes E 98 minutes
9 Which of the following represent the fastest route for Victoria’s journey from home to
school.
A Home − A − B − School B Home − A − B − D − School
C Home − C − D − School D Home − C − B − D − School
E Home − E − C − B − D − School
13E Dijkstra’s algorithm

Finding the shortest path from one vertex of a graph to another is easy to determine if
the graph is small and does not have too many vertices and edges. When there are many
vertices and many edges, a systematic method, called an algorithm, can be used to find the
shortest path.
Dutch computer scientist, Edsger Wybe Dijkstra (pronounced ‘Dyke-stra’) developed an
algorithm for determining the shortest path through a graph. This algorithm, and others like
it, have important applications to computerised routing and scheduling programs, such as
GPS navigation devices.
Note: An alternative version is available online. Either method can be used. There is no curriculum
requirement for you to know both methods.
The algorithm
You may choose to read through the example first to see a detailed implementation. Here we
write the algorithm to emphasise its repetitive aspect.
Step 1: Assign the starting vertex a label of value zero and circle the vertex and the zero
together.
Once a vertex and its label have been circled it cannot be changed
Step 2: Consider the vertex which has been most recently circled. Suppose this vertex to be
X and the label of value d assigned to it. Then, in turn, consider each vertex directly joined
to X but not yet permanently circled. For each such vertex, Y say, temporarily assign it with
the value d + (the weight of edge XY) if Y does not have a temporary value or if it does,
assign the lesser of d + (the weight of edge XY) and the existing temporary value.
Step 3: Choose the least of all temporary value labels on the network. Make this value label
permanent by circling it.

Step 4: Repeat Steps 2 and 3 until the destination node has a permanent label.
Step 5: Go backwards through the network, retracing the path of shortest length from the
destination vertex to the starting vertex by
starting at destination and go to the circled vertex with value = destination value − edge
value.
continuing to move back to the start vertex folowing this procedure.
Example 10 Using Dijkstra’s algorithm to find the shortest path in a network:

graphical method
Find the shortest path from A to F in the B 2 E

weighted graph shown on the right. 2
A
3 3
2 2
6 F
5
C 1 D
Step 1 0 B 2 E
2
Assign the starting vertex a zero and A
3 3
circle the vertex and its new value of 2
6 2
zero. F
5
C 1 D
A is the starting vertex, it is assigned
zero and it is circled.
Step 2
Assign a value to each vertex connected
0 B 2 2 E
to the starting vertex. The value 2
A
assigned is the length of the edge 3
2 3
connecting it to the starting vertex. 6 2
F
Circle the vertex with the lowest
5
assigned value.
C6 1 D
The starting vertex A is connected to
vertices B and C.
The vertex B is assigned 2 and the
vertex C is assigned 6.
Vertex B is circled because it has the
lowest value.
13E Dijkstra’s algorithm 639
Step 3 2
B 2 2
4E
A0
From the newly circled vertex, assign
2 3 3
a value to each vertex connected to it 6 2
F
by adding the value of each connecting 5
1 5
C4 D
edge to the newly circled vertex’s
value. The newly circled vertex B is connected
If a connecting vertex already has a to three vertices; C, D and E. Starting
value assigned and the new value is less with vertex B’s value of 2, E is assigned
than it, replace it with the new value. 4 (adding 2 from the connecting edge)
If a vertex is circled, it cannot have its and D is assigned 5 (adding 3 from the
value changed. connecting edge).
Consider all uncircled vertices and The vertex C will be re-assigned 4
circle the one with the lowest value. (adding 2 from the connecting edge)
because it is lower than 6.
Now there are two uncircled vertices
with the lowest assigned value 4,
vertices C and E; it does not matter
which one is circled. C is circled.
Step 4 B2 2 4E
0 2
A
Repeat Step 3 until the destination
3 3
2 2
vertex and its assigned value are 6 F
7
circled. 5
1 5
C4 D
The length of the shortest path will be
the assigned value of the destination Vertex F is the destination vertex,
vertex. assigned a value of 7. Therefore the
Step 5 shortest path from A to F has a length
The shortest path is found by of 7.
backtracking. Starting at the destination To find the shortest path, start at F and
vertex, move to the circled vertex consider the two connecting edges to it.
whose value is equal to the destination The edge of length 3 is correct, because
vertex’s assigned value, subtract the 7 minus 3 equals 4, the value of vertex
connecting edge value. Continue to E.
subtract the connecting edge value from Likewise from E, subtract the
one circled vertex to the next until you connecting edge of 2 to vertex B to
reach the starting vertex. equal 2, then subtract the connecting
Note: Once a vertex is assigned a value, it cannot edge of 2 to A to equal zero.
be assigned a larger value, even if it has not been Therefore the shortest path from A to F is:
circled yet. You do not need to circle all vertices.
A − B − E − F with a length of 7.
Stop when the destination vertex is circled.

640 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13E
Skill-
sheet Exercise 13E
Calculations within Dijkstra’s algorithm

Example 10 1 Using Dijkstra’s algorithm, determine the shortest path and write down the length of
the shortest path from A to H in the following weighted graphs.
a C b B 27 G
36 104
B H 17 18 6 15
E
79 96
A 18 C 15 12 I 11
E 4 H
20 10 17 3
A 76 15 14 F 13
27 13
57 D 29 J
F 56 G
c B 147 F d X 42 D
124 118 12
95 24 34 7 23 G
101 E 75 102 H
A C 38 14
135 61 C 15 F H
74 78 A
125 67 G 35
42 11
D 18 28 11
I
B 47
12 E
2 Renee drives to work each day. 4 9

The edges of the network opposite 7 7 3 5
2 2 3 7
represent the roads that Renee can Home 4 4 6 5
3 9 Work
2 4 4 3
use to drive to work. The numbers 2 13 4 8
3 2
on the edges give the time, in
16
minutes, to travel along each road.
What is the shortest time that
Renee can drive between home and
work?
Note: See Chapter review, written-response question 1 for more problems using Dijkstra’s
algorithm.
13F Trees and minimum connector problems

Learning intentions
I To be able to identify a tree.
I To be able to find a spanning tree for a graph.
I To be able to find the minimum spanning tree for a network using Prim’s algorithm.
In the previous applications of networks, the weights on the edges of the graph were used to
determine a minimum pathway through the graph. In other applications, it is more important
to minimise the number and weights of the edges in order to keep all vertices connected to
the graph. For example, a number of towns might need to be connected to a water supply.
The cost of connecting the towns can be minimised by connecting each town into a network
ISBN 978-1-009-11041-9 © Peter
or water pipes only once, rather than Jones et aleach
connecting 2023 town to every other town.
Cambridge University Press
13F Trees and minimum connector problems 641
Problems of this type are called connector problems. In order to solve connector problems,
you need to learn the language of networks that have as few edges as possible.
Trees
A tree is a connected graph that has no loops, multiple G
B F
edges or cycles.
This tree has seven vertices and six edges.
The number of edges is always one less than the number A E
of vertices. C
D
Spanning trees
Every connected graph will have at least one subgraph that is a tree. A subgraph is a tree,
and if that tree connects all of the vertices in the graph, then it is called a spanning tree.
Trees
A tree has no loops, multiple edges or cycles.
If a tree has n vertices, it will have n − 1 edges.
A spanning tree is a tree that connects all of the vertices of a graph.
There can be more than one spanning tree for any connected graph. The total weight of a
spanning tree is the total of all the weights on the edges that make up the tree.
Example 11 Finding the weight of a spanning tree
a Draw one spanning tree for the graph 3

shown. 4
2 1
b Calculate the weight of this spanning
tree. 3
6
4 2
5 7
2 5 2
a 1 Count the number of vertices and There are 9 vertices and 13 edges.
edges in the graph.
2 Calculate the number of edges in the The spanning tree will have 8 edges.
spanning tree.

3 Calculate how many edges must be Remove 13 − 8 = 5 edges.

removed.
4 Choose edges to remove.
4
2 1
3
4
5
5 2
b Add the weights of the remaining

4
edges. 2 1
3
4
5
2
5
Weight = 5 + 2 + 1 + 4 + 5 + 2 + 3 + 4
= 26
Minimum spanning trees

One of the spanning trees from a particular connected graph will have the smallest total
weight. This tree is called the minimum spanning tree. Minimum spanning trees can be
found using an algorithm called Prim’s algorithm.
Prim’s algorithm for finding a minimum spanning tree

Choose a starting vertex (any will do).
Inspect the edges starting from the starting vertex and choose the one with the smallest
weight. (If there are two edges that have the same weight, it does not matter which
one you choose). The starting vertex, the edge and the vertex it connects to form the
beginning of the minimum spanning tree.
Now inspect all of the edges starting from both of the vertices you have in the tree so
far. Choose the edge with the smallest weight, ignoring edges that would connect the
tree back to itself. The vertices and edges you already have, plus the extra edge and
vertex it connects form the minimum spanning tree so far.
Keep repeating this process until all of the vertices are connected.

Example 12 Finding the minimum spanning tree
Apply Prim’s algorithm to find the C

minimum spanning tree for the graph 8
shown on the right. Write down the total B
2 3
weight of the minimum spanning tree. 6 5
A 5 D
6 E
7
6
2
F
Start with vertex A. C
The smallest weighted edge from vertex A is to B 8
with weight 2. B
2 3
6 5
A 5 D
6 E
7
6
2
F
Look at vertices A and B. The smallest weighted C
edge from either vertex A or vertex B is from A to 8
D with weight 5. B
2 3
6 5
A 5 D
6 E
7
6
2
F
Look at vertices A, B and D. The smallest C
weighted edge from vertex A, B or D is from 8
D to C with weight 3. B
2 3
6 5
A 5 D
6 E
7
6
2
F

Look at vertices A, B, D and C. The smallest C

weighted edge from vertex A, B, D or C is from C 8
to E with weight 5. B
2 3
6 5
A 5 D
6 E
7
6
2
F
Look at vertices A, B, D, C and E. The smallest C
weighted edge from vertex A, B, D, C or E is from 8
E to F with weight 2. B
2 3
All vertices have been included in the graph. This 6 5
is the minimum spanning tree. A 5 D
6 E
7
6
2
F
Add the weights to find the total weight of the The total weight of the
minimum spanning tree. minimum spanning tree is
2 + 5 + 3 + 5 + 2 = 17.
Minimum Connector problems

Minimum spanning trees represent the least weight required to keep all of the vertices
connected in the graph. If the edges of a graph represent the cost of connecting towns
to a gas pipeline, then the total weight of the minimum spanning tree would represent
the minimum cost of connecting the towns to the gas. This is an example of a connector
problem, where the cost of keeping towns or other objects connected together is important to
make as low as possible.
Example 13 Solving a connector problem
Water is to be piped from a water tank to seven outlets on a property. The distances (in
metres) of the outlets from the tank and from each other are shown in the network below.
Starting at the tank, the aim is to find the minimum length of pipe, in metres, which will
be needed to have water piped to all outlets in the property.

Outlet A 13 Outlet B
11 6
12 6 Outlet C
Tank 2
Outlet F Outlet G 11
10 8 5
7
Outlet E 9 Outlet D
a On the diagram, show where the water pipes will be placed in order to minimise the
length required.
b Calculate the total length, in metres, of the water pipe that is required to obtain this
minimum length.
a 1 The water pipes will be a Outlet A 13 Outlet B
minimum length if they are 12 6
6 11
placed on the edges of the Outlet C
Outlet F Outlet G Tank 2
minimum spanning tree for
5 11
the network. 10 8
7
A starting point for Prim’s
Outlet E 9 Outlet D
algorithm is the vertex that
is connected to the tank by
the edge with the smallest
weight. The starting vertex
(Tank), the edge and
the vertex it connects to
form the beginning of the
minimum spanning tree.
2 Follow Prim’s algorithm to Outlet A 13 Outlet B
find the minimum spanning 12 6
6 11
tree. Outlet C
Outlet F Outlet G Tank 2
5 11
10 8
7
Outlet E 9 Outlet D
b Add the weights of the The length of water pipe required is 2 + 6 + 5 + 8 +
minimum spanning tree. 7 + 10 + 6 = 44 metres
Write your answer.

646 Chapter 13 Graphs, networks and trees: travelling and connecting problems 13F
Skill-
sheet Exercise 13F
Spanning trees
Example 11 1 A weighted graph is shown on the right. 5
a How many edges must be removed
in order to leave a spanning tree? 4 2
5 3
b Remove some edges to form three 3 4
different trees.
c For each tree in part b, find the total 2 2 3 6
weight.
3
Minimum spanning trees and Prim’s algorithm
Example 12 2 Find a minimum spanning tree for each of the following graphs and give the total
weight.
a A 2 B 2 E b 24 C
B
3
6 2 2 F 16 16 16
3 17 E 17
5 A D
C 1 D 11
10 12 15
G 17 F
c B 18 C d D
10 H 70 100 140
E 19 E G
18 20 90 C 100 200
10 9 80 90
F
B
A 18 D 90
120
A
Connector problems
Example 13 3 In the network opposite, the vertices represent 300
water tanks on a large property and the edges 70 40 140
80 70
represent pipes used to move water between 90 60 80
40
these tanks. The numbers on each edge 40 40
80 50
indicate the lengths of pipes (in m) connecting 60 110 120
150
different tanks. 90
Determine the shortest length of pipe needed to connect all water storages.

13F 13F Trees and minimum connector problems 647

4 For the graph opposite, the length of the 9 11
minimum spanning tree is

8 5 6
A 44 B 45 C 46 8
10
7
D 47 E 48
7
7
6
5
8
5 The minimum spanning tree for the graph below includes the edge with weight
labelled k.
The total weight of all edges for 6 6
the minimum spanning tree is 58. 9

5 6 5
The value of k is 7
5
A 6 B 7 C 8 8 9 6 9
D 9 E 10
9 8 k
6 The minimum spanning tree for the graph 11

a
opposite includes two edges with weights a 9
13
11
and b. The length of the minimum spanning tree 15
10 10
is 124. 12
14
12 11 b
The values of a and b could be
14 13
A a = 6 and b = 17
12
B a = 12 and b = 12 12
C a = 10 and b = 14
D a = 10 and b = 15
E a = 13 and b = 12
7 A minimum spanning tree is to 4

4 4
be drawn for the weighted graph
3 1
opposite. How many edges with 3 4
3 2
weight 4 will not be included in
3 4 4 4
any particular minimum spanning
tree? 3 4
4 5 3
A 1 B 2 C 3 D 4 E 5
8 Consider the weighted graph opposite. 3 7

How many different minimum spanning trees are possible? 5 5
4 7 5
A 1 B 2 C 3 D 4 E 5 5
5
7 5

Review 648 Chapter 13 Graphs, networks and trees: travelling and connecting problems
Graph A graph is a diagram that consists of a set of points A B

Assign- called vertices and a set of lines called edges. Each
ment
edge joins two vertices. D C
Edge In the graph above, the lines joining A, B, C and D are edges.
Vertex In the graph above, the points A, B, C, D are vertices.
Loop A loop is an edge that connects a vertex to itself. In the graph above
there is a loop at vertex C.
Degree of a The degree of a vertex is the number times edges attach to a vertex.
vertex The degree of vertex A is written as deg(A).
In the graph above, deg(A) = 4, deg(B) = 2 and deg(D) = 2.
A loop has degree of 2. In the graph above, deg(C) = 4.
Multiple edge Sometimes a graph has two or more identical edges. These are called
multiple edges. In the graph above, there are multiple edges between
vertex A and vertex D.
Simple graph Simple graphs are graphs that do not have loops and do not have
multiple edges.
Isolated vertex An isolated vertex is one that is not connected to any other vertex.
Isolated vertices have degree of zero.
Degenerate A degenerate graph has no edges. All of the vertices are isolated.
graph
Connected graph A connected graph has no isolated vertex. There is a path between
each pair of vertices.
Bridge A bridge is a single edge in a connected graph that, if it were removed,

leaves the graph disconnected.
Complete graph A complete graph has every vertex connected to every other vertex by
an edge.
Subgraph A subgraph is a graph that is part of a larger graph and has some of the
same vertices and edges as that larger graph. A subgraph does not have
any extra vertices or edges that do not appear in the larger graph.
Equivalent graph Graphs that contain identical information (connections between

(isomorphic graph) vertices) to each other are equivalent graphs or isomorphic graphs.

Review
Face An area in a graph or network that can only be reached by crossing an
edge. One such area is always the area surrounding a graph.
Planar graph A planar graph can be drawn so that no two edges overlap or intersect,
except at the vertices.
This graph is planar. This graph is not planar.
Euler’s formula Euler’s formula applies to planar graphs.

If v = the number of vertices, e = the number of edges and
f = the number of faces then v + f = e + 2
Adjacency matrix An adjacency matrix is a square matrix that uses a zero or an integer
to record the number of edges connecting each pair of vertices in the
graph.
Travelling Movement through a graph from one B

vertex to another along the edges is A
called travelling through the graph. F
C
E
Walk A walk is a sequence of edges, linking successive vertices, in a graph.

In the graph above, E−A−F−D−C−F−E−A is an example of a walk.
Trail A trail is a walk with no repeated edges.

In the graph above, A−F−D−E−F−C is an example of a trail.
Path A path is a walk with no repeated vertices and no repeated edges.

In the graph above, F−A−B−C−D is an example of a path.
Circuit A circuit is a trail (no repeated edges) that starts and ends at the same
vertex.
In the graph above, A−F−D−E−F−B−A is an example of a circuit.

Cycle A cycle is a path (no repeated edges nor vertices) that starts and ends
at the same vertex. The start and end vertex is an exception to repeated
vertices.
In the graph above, B−F−D−C−B is an example of a cycle.
Eulerian trail An Eulerian trail is a trail (no repeated edges) that includes all of the
edges of a graph. Eulerian trails exist if the graph is connected and has
exactly zero or two vertices of odd degree. The remaining vertices have
even degree.
Eulerian circuit An Eulerian circuit is a trail (no repeated edges) that includes all of the
edges of a graph and that starts and ends at the same vertex. Eulerian
circuits exist if the graph is connected and has all of the vertices with an
even degree.
Hamiltonian path A Hamiltonian path is a path (no repeated edges or vertices) that
includes all of the vertices of a graph.
Hamiltonian cycle A Hamiltonian cycle is a path (no repeated edges or vertices) that
starts and ends at the same vertex. The starting vertex is an exception to
repeated vertices.
Weighted graph A weighted graph has numbers, called weights, associated with the
edges of a graph. The weights often represent physical quantities as
additional information to the edge, such as time, distance or cost.
Network A network is a weighted graph where the weights represent physical

quantities such as time, distance or cost.
Shortest path The shortest path through a network is the path along edges so that the
total of the weights of that path is the minimum for that network.
Shortest path problems involve finding minimum distances, costs
or times through a network. Shortest paths can be determined by
inspection or by using Dijkstra’s algorithm.
Dijkstra’s Dijkstra’s algorithm is an algorithm for determining the shortest path

algorithm through a network from one vertex to another.
Tree A tree is a connected graph that contains no cycles, multiple edges or

loops.
A tree with n vertices has n − 1 edges.
Spanning tree A spanning tree is a tree that connects every vertex of a graph.
A spanning tree is found by counting the number of vertices (n) and
removing enough edges so that there are n − 1 edges left that connect all
vertices.

Review
Minimum A minimum spanning tree is a spanning tree for which the sum of the
spanning tree weights of the edges is as small as possible.
Prim’s algorithm Prim’s algorithm is an algorithm for determining the minimum

spanning tree of a network.
Skills checklist
list
13A 1 I can identify edges, vertices and loops in a graph.
13A 2 I can determine the degree of a vertex in a graph.
13A 3 I can define and identify simple graphs, isolated vertices, degenerate graphs,
connected graphs, bridges and subgraphs.
13A 4 I can recognise isomorphic graphs.
See Exercise 13A Question 3
13A 5 I can define planar graphs.
13A 6 I can redraw graphs in planar form.
13A 7 I can use Euler’s formula.
13B 8 I can write an adjacency matrix from a graph.
See Exercise 13B Question 1
13B 9 I can construct a graph from an adjacency matrix.
13C 10 I can define walks, trails, paths, circuits and cycles through a graph.

13C 11 I can identify Eulerian trails and circuits through graphs.
13C 12 I can determine whether an Eulerian trail or circuit exists in a graph.
13C 13 I can identify Hamiltonian paths and cycles through graphs.
13D 14 I can define a weighted graph.
See Exercise 13D Question 1
13D 15 I can calculate the shortest path from one vertex to another by inspection.
13D 16 I can calculate the shortest path from one vertex to another using Dijkstra’s
algorithm.
13E 17 I can define tree, spanning tree, minimum spanning tree.
13E 18 I can draw a minimum spanning tree using Prim’s algorithm.
1 The minimum number of edges for a graph with seven vertices to be connected is:
A 4 B 5 C 6 D 7 E 21
2 Which of the following graphs is a spanning tree for the A
C
network shown?
B
F
D E
A A B A C A
C C C
B B B
F F F
D E D E D E
G G

Review
D A E A
C C
B B
F F
D E D E
G G
3 For the graph shown, which vertex has degree 5? Q
A Q B T C S D R E U S
P
U
T
R
4 A connected graph with 15 vertices divides the plane into 12 regions. The number of
edges connecting the vertices in this graph will be:
A 15 B 23 C 24 D 25 E 27
5 Which of the following graphs does not have an Eulerian circuit?

A B C
D E
6 A connected planar graph divides the plane into a number of regions. If the graph has
eight vertices and these are linked by 13 edges, then the number of regions is:
A 5 B 6 C 7 D 8 E 10
7 For the graph shown, which of the following paths is a A

Hamiltonian cycle?
E
A A−B−C−D−C−F−D−E−F−A−E−A
F B
B A−E−F−D−C−B−A D
C A−F−C−D−E−A−B−A C
D A−B−C−D−E−A
E A−E−D−C−B−A−F

8 The graph opposite has:

A four faces B five faces
C six faces D seven faces
E eight faces
9 The sum of the degrees of the vertices on the graph

shown here is:
A 20 B 21 C 22
D 23 E 24
10 For the graph shown, which additional edge could be B

added to the network so that the graph formed would C
contain an Eulerian trail? A
E
A A−F B D−E C A−B
F
D C−F E B−F
D
11 For the graph shown here, a minimum spanning tree 7

8
has length:
A 30 B 31 C 33 4 10
5
D 34 E 26
4 6
9
2
12 Of the following graphs, which one has both Eulerian circuit and Hamiltonian cycles?
A B C
D E
13 A graph with six vertices is connected with the minimum number of edges. The
minimum number of extra edges needed to make this a complete graph is
A 5 B 6 C 10 D 14 E 16
14 Four towns, A, B, C and D, are linked by roads as shown. Which B

of the following graphs could be used to represent the network of C
roads? Each edge represents a route between two towns.
A
D

Review
A B B B C B
C C
C
A A A
D D
D
D B E B
C
C
A A
D
D
15 Which one of the following graphs has an Eulerian circuit?

A B C
D E
16 The network below shows the distance, in metres, between points. The shortest path
between S and T has length 36 m.
A 15 F
3 8
22 D
9 10 T
B 2
s 18 G
3 x
E
11
17 6 7
C 15 H
The value of x is.
A 4 B 5 C 6 D 7 E 8


1 Determine the shortest path and write down the length of the shortest path from
vertex A to vertex X in the following weighted graphs.
a B 8 E 2 X
2 4 1 4
1 5
4 2
A 3 G
C F
2
2 4
D
b D 58 G
28 45
16
B 20 30
18 H X
50 E 9
A 56
20 I 16 52
9
12
C 16 F 72 J
c K
A 3 B
4 7
5 26 I 12 L
9 7
12 27 8 20
C H 11
D M 11
E 9 20
15 14 7 X
11 4 6
3 9 6
6
F 9 G 18
J 17 N
d B K
F 9
5 2 5 13
D I 5
A 16 N
12 4
3 G 3
5 10 L 16
6 8 7
6 E 3 8
8 X
C 6 J 21 9
8
H M
2 The map shows six campsites, B
6 9
A, B, C, D, E and F, which are joined by
tracks. The numbers by the paths show C
lake 4
lengths, in kilometres, of that section of A 5 D
3
track. 5
1 4
5
2
F 3 E

Review
a i Complete the graph opposite, B
9
which shows the shortest direct 6
distances between campsites. C
(The campsites are represented A 8 D 4
by vertices and tracks are 5
F 7
represented by edges.) 4
8 5
E
ii A telephone cable is to be laid to enable each campsite to phone each other
campsite. For environmental reasons, cables can only be laid along the tracks
and cables can only connect to one another at the campsites. What is the
minimum length of cable necessary to complete this task?
iii Fill in the missing entries for the
adjacency matrix shown for the A B C D E F
completed graph formed above. A  0 1 0 1 1 1 
 
B  1 0 1 0 0 0 
 
C  0 1 0 1 1 0 
 
D  1 0 − − − − 
 
E  1 0 − − − − 
 
F 1 0 − − − −
b A walker follows the route A−B−A−F−E−D−C−E−F−A.

i How far does this person walk?
ii Why is the route not a Hamiltonian cycle?
iii Write down a route that a walker could follow that is a Hamiltonian cycle.
iv Find the distance walked in following this Hamiltonian cycle.
c It is impossible to start at A and return to A by going along each track exactly once.
An extra track joining two campsites can be constructed so that this is possible.
Which two campsites need to be joined by a track to make this possible? (Two
tracks are only considered the same if they share their entire length. For example,
A–E and F–E are considered different tracks even though they share a common
stretch near E.)

3 The network on the right shows six Croghon

villages represented as vertices of the 6 Kenton
graph. The edges represent the roads 13
12
connecting the villages. The weights Melville
on the edges are the distances, in Bartow
kilometres, alog each of the roads. 5 11
7 8
9
Stratmoore Osburn
20
a What is the degree of the vertex representing Melville?
b Determine the sum of the degrees of the vertices of this graph.
c Verify Euler’s formula for this graph.
A salesperson might need to travel to every village in this network to conduct business.
d If the salesperson follows the path Stratmoore − Melville − Kenton− Osburn
− Melville− Croghon − Bartow − Stratmoore, has the salesperson followed a
Hamiltonian cycle? Give a reason to justify your answer.
e If the salesperson follows the path Croghorn− Bartow − Stratmoore − Melville−
Kenton − Osburn, what is the mathematical term for this path?
It would make sense for the salesperson to avoid visiting a certain village more than
once, and it would also make sense for them to return ‘home’ after travelling the
shortest distance possible.
f If the salesperson starts and ends in Bartow, find the shortest route and state the
shortest distance the salesperson would have to travel.
g If the salesperson can start and end at any village in the network, what is the shortest
route possible?
A road inspector must travel along every road connecting the six villages.
h Explain why the inspector could not follow an Eulerian circuit through this road
network.
i The inspector may start and end their route at different villages, but would like to
travel along each road once only. Which villages can the inspector start their route
from? Write down a path the inspector could take to complete their work.
j The speed limit for each of these roads is 60 km/hr. If the inspector must complete
their work by 5 p.m, what is the latest time that the inspector can begin their work?
New electrical cables connecting the villages are required. They will be installed along
some of the roads listed in the graph above. These cables will form a connected graph
and the shortest total length of cable will be used.
k Give a mathematical term to describe a graph that represents these cables.
l Draw the graph that represents these cables and find the total length of cable
required.
Chapter
14
Flow, matching and
scheduling problems
Chapter objectives
I How do we define a directed graph?
I How do we define flow?
I How do we calculate the maximum flow through a network?
I How do we draw and use a bipartite graph to solve allocation problems?
I How do we find the optimal allocation of multiple groups of objects?
I How do we identify predecessors of an activity?
I How do we draw an activity network and use it to plan for a project?
I How do we account for float times in our project?
I How do we find the earliest starting time and latest finishing time for an
activity in a project?
I How do we identify the critical path of an activity network?
In the previous chapter, undirected graphs were used to define and represent situations.
In this chapter, directed graphs will be used to model networks and solve problems
involving travel, connection, flow, matching, allocation and scheduling.

660 Chapter 14 Flow, matching and scheduling problems
14A Flow problems

Learning intentions
I To be able to define and describe flow.
I To be able to calculate the maximum flow through a flow network by observation.
I To be able to identify cuts and calculate cut capacities.
I To be able to determine the maximum flow through a flow network by finding the
capacity of the minimum cut.
I To be able to solve flow problems by finding minimum cut capacities.
Directed graphs
In the previous chapter, graphs were used to represent connections between people, places
or objects. The vertices of a graph represented objects, such as towns, and edges represented
the connections between them, such as roads. Weighted graphs included extra numerical
information about the connections, such as distance, time or cost. When a graph has this
numerical information we call it a network.
A directed graph, or digraph, records directional information
on networks using arrows on the edges. The network on the right
shows roads around a city. The vertices are the intersections of
the roads and the edges are the possible road connections between
the intersections. The arrows show that some of the roads only
allow traffic in one direction, while others allow traffic in both
directions.
Understanding minimum flow

One of the applications of directed graphs to real-life situations is flow problems. Flow
problems involve the transfer or flow of material from one point, called the source, to
another point called the sink. Examples of this include water flowing through pipes, or
traffic flowing along roads.
source → flow through network → sink
Water flows through pipes in only one direction. In a digraph representing water flow,
the vertices are the origin and destination of the water and the edges represent the pipes
connecting them. The weights on the edges would be the amount of water that can flow
through the pipe in a given time. The weights of flow problem directed graphs are called
capacities.

14A Flow problems 661
The diagram on the right shows two pipes that 58 L/min

are joined together, connecting the source of
25 L/min
water to the sink. There is a small pipe with
capacity 25 litres per minute joined to a large
pipe with capacity 58 litres per minute.
Even though the large pipe has a capacity greater than 25 litres per minute, the small pipe
will only allow 25 litres of water through each minute. The flow through the large pipe
will never be more than 25 litres per minute. The large pipe will experience flow below its
capacity.
If we reverse the connection and direct 58 L/min
water through the large capacity pipe 25 L/min
into the smaller capacity pipe, there
will be a ‘bottleneck’ of flow at the
junction.
The large capacity pipe is capable of delivering 58 litres of water every minute to the small
pipe, but the small pipe will only allow 25 litres per minute to pass.
In both of these situations, the flow through the entire pipe system (both pipes from source to
sink) is restricted to a maximum of 25 litres per minute. This is the capacity of the smallest
pipe in the connection.
If we connect more pipes together, one
after the other, we can calculate the
overall capacity or maximum flow
of the pipe system by looking for the
smallest capacity pipe in that system.
Maximum flow
If pipes of different capacities are connected one after the other, the maximum flow
through the pipes is equal to the minimum capacity of the individual pipes.

Example 1 Calculating the maximum flow
In the digraph shown on the right, the vertices A, B 600 C
B, C, D and E represent towns. The edges of the 800

300
graph represent roads and the weights of those
A E
edges are the maximum number of cars that can 500 D 150
travel on the road each hour. The roads allow only
one-way travel.
a Find the maximum traffic flow from A to E through town C.
b Find the maximum traffic flow from A to E overall.
c A new road is being built to allow traffic from town D to town C. This road can carry
500 cars per hour.
i Add this road to the digraph.
ii Find the maximum traffic flow from A to E overall after this road is built.
a Look at the subgraph that includes 600
B C
town C. The smallest capacity of the
individual roads is 300 cars per hour. 300 800
This will be the maximum flow through A E
town C. (source) (sink)
The maximum flow from A to E

The maximum flow is 300 cars per hour.
through town C is equal to the smallest
capacity road along that route.
b Look at the two subgraphs from A to E. A E
(source) 500 D 150 (sink)
The maximum flow through D will be
150 cars per hour (minimum capacity).
Add the maximum flow through C to The maximum flow from A to E overall is:
the maximum flow through D. 300 + 150 = 450 cars per hour
c i Add the edge to the diagram. B 600 C
500 800
300
A E
500 D 150
ii The maximum flow through

The new maximum flow is now 800 cars
A−B−C−E is 300. But C−E has
per hour.
capacity 800. If another 500 cars per
hour come through A−D−C, they
will also be able to travel from
C to E.

Cuts
It is difficult to determine the maximum flow by inspection for directed networks that
involve many vertices and edges. We can simplify the search for maximum flow by
searching for cuts within the digraph.
A cut divides the network into two parts, B
completely separating the source from the
1000
sink. It is helpful to think of cuts as imaginary 700
breaks within the network that completely
block the flow through that network. For the C 400
S A
network or water pipes shown in this diagram, 1500
the dotted line is a cut. This cut completely
600
blocks the flow of water from the source (S) 1200 600
to the sink (A).
M
source sink source sink
The dotted line on the graph above is a valid The dotted line on the graph above is not
cut because it separates the source and the a valid cut because material can still flow
sink completely. No material can flow from from the source to the sink. Not all of the
the source to the sink. pathways from source to sink have been
blocked by the cut.
Capacity of a cut
The cut capacity is the sum of all the capacities of the edges that the cut passes through,
taking into account the direction of flow. The capacity of an edge is only counted if it
flows from the source side to the sink side of the cut.
In the simple network shown, the cut passes A sink

through three edges. The edge B to A is not
counted in the capacity of the cut because the
flow for that edge is from the sink side to the
source B
source side of the cut.

Example 2 Calculating cut capacity
Calculate the capacity of the four cuts 15 B 14 C

10
shown in the network on the right. The
S 6 15 T
source is vertex S and the sink is vertex T . 12
20 15
F 20 D
C1 C2 C3 C4
All edges in C1 are counted. The capacity of C1 = 15 + 20 = 35
Note that the edge from F to B is not The capacity of C2 = 14 + 20 = 34
counted in C2 . The capacity of C3 = 14 + 15 + 20 = 49
All edges in C3 are counted. The capacity of C4 = 20 + 10 = 30
Note that the edge from D to C is not
counted in C4 .
The capacity of a cut is important to help determine the maximum flow through any digraph.
Look for the smallest, or minimum, cut capacity that exists in the graph. This will be
the same as the maximum flow that is possible through that graph. This is known as the
maximum-flow minimum-cut theorem.
Cut, cut capacity and minimum cut capacity

A cut is an imaginary line across a directed graph that completely separates the source
(start of the flow) from the sink (destination of the flow).
The cut capacity is the sum of the capacities of the edges that are cut. Only edges that
flow from the source side of the cut to the sink side of the cut are included in a cut
capacity calculation.
The minimum cut capacity possible for a graph equals the maximum flow through the
graph.
Example 3 Calculating maximum flow using cuts
Determine the maximum flow from S to T A

for the digraph shown on the right. 8
5 3
B
S T
3 5
1
11
C

1 Mark in all possible cuts on the A
network. 8
5 3
B
S T
3 5
1
C1C 11 C
2 C6
C3 C4 C5
2 Calculate the capacity of all the cuts. The capacity of C1 = 8 + 11 = 19

The capacity of C2 = 3 + 11 = 14
The capacity of C3 = 3 + 5 + 11 = 19
The capacity of C4 = 8 + 3 + 1 = 12
3 Identify the minimum cut capacity and The minimum cut capacity is 7 so the
write your answer. maximum flow from S to T is 7.
Calculating maximum flow by tracking flow through a network (Optional)

There is another method we can use to calculate the maximum flow through a network. This
method relies on tracking the flow through every edge.
Example 4 Calculating maximum flow
The koala sanctuary in Cowes allows visitors to walk through their park. The park is
represented by a network below, where each edge represents one-way tracks for visitors
through the park. The direction of travel on each track is shown by an arrow. The
numbers on the edges indicate the maximum number of people who are permitted to walk
along each track each hour.
B
15 10
C
A 8
D 12
20 9
8 G
13
25
E 10 F
a Starting at A, how many people are permitted to walk to G each hour?

b Given that one group of nine people would like to walk from A to G together as a
group, list all the different routes they could take so that the entire group of nine will
stay together for the duration of their walk.
c What is the largest group of people that could walk through the koala sanctuary if they
must stay together in a group for the entire duration of the walk?

a Firstly, consider the edges The vertex A is the source in this network. The three
coming from the source. edges connected to A, flowing towards B, D and E
When calculating the will be flowing at maximum capacity because they
maximum flow through a are coming from the source. Draw lines along these
network, always assume edges and circle their capacities as they are flowing at
the initial edges from the maximum capacity.
source are flowing at their B
15 10
maximum capacity. C
A 8
D 12
20 9
8 13 G
25
E 10 F
Now, from the vertices At vertex B, although there is a flow of 15 coming in

accepting flow from the from the source, the edges taking flow towards the sink
source, start observing at G are of different capacities. From B to C only a
the flow through the capacity of 10 can flow, therefore the same 10 can also
network and if the only flow from C to G as no other edges are connected
maximum capacity in this route. From B, as 10 flows to C, the leftover 5
of each edge can be can flow to D. Even though the edge from B to D has a
achieved via direct flow maximum capacity of 8, there is only 5 available to flow
from another edge or by through this edge; cross off 8 and write a 5 next to it.
splitting the flow of one B 10
15
edge into two. C
A 5
10
D 9
20
8 13 G
25
E 10 F
At vertex E, a flow of 8 is coming from the source.

Although the edge from E to F has a capacity of 10, only
a maximum of 8 can flow through it. Just as before, cross
out the 10 and write an 8 next to it to signify the flow
passing through the edge.
B 10
15
C
A 5
10
D 9
20
8 13 G
25
E 8 F

At vertex D there is a total of 25 flowing into it,
coming from vertices B and A. This flow of 25 can be
redistributed to the two edges coming from D towards
the sink at G. Of the 25, 9 can flow directly to G and 13
can flow from D to F. The maximum capacities of these
edges can be achieved, so circle these numbers along the
edges.
B 10
15
C
A 5
D 10
9
20
8 13 G
25
E 8 F
Finally, the edge from F to G has a maximum capacity

of 25, however only 21 (8 + 13 = 21) is coming through.
Cross out 25, replace with 21 to indicate the correct flow
through the edge.
B 10
15
C
A 5
D 10
9
20
8 13 G
21
E F
8
The maximum flow through the network is the total
amount incoming to the sink vertex G, which is
10 + 9 + 21 = 40. Therefore a maximum of 40 people
can walk through the koala sanctuary each hour.
b A group of 9 must begin Starting from vertex A there are only two edges the
at vertex A, only pass group of 9 people can walk along; the edges going to
through edges with a vertices B and D. Walking through vertex B there is
capacity greater than one walk possible: A − B − C − G. Walking through
or equal to 9 and end at vertex D there are two walks possible: A − D − G and
vertex G. Take note of A − D − F − G.
the direction of the arrow
heads.

c From the starting vertex A, Starting at vertex A the largest possible group of
consider the largest possible people that can enter the sanctuary is 20, however
group that could start a walk after reaching vertex D the group of 20 cannot stay
through the sanctuary and together as they move towards vertex G because
then analyse how many the capacity of the edges from D can only take a
of that group could then maximum of 13 people. From D, moving towards
walk to vertex G given the G a group of 13 could pass through with no other
capacities of each of the edges restrictions. This is the largest group of people
throughout the network. that can enter the sanctuary at vertex A and pass
through to vertex G, together as one group, given the
restrictions of the edge capacities.
Calculating maximum flow from more than one source
Example 5 Calculating maximum flow from more than one source
Water enters a network of pipes at either Source 1 or Source 2 and flows out at either
Outlet 1 or Outlet 2.
The numbers next to the arrows represent the maximum rate, in kilolitres per minute, at
which water can flow through each pipe.
Source 1 400
300
800 300 100 Outlet 1
400
200
200
100 300 200
Source 2 300 500 400
300 600
100
100 300 Outlet 2
Determine the maximum rate, in kilolitres per minute, at which water can flow from these
pipes into the ocean at Outlet 1 and Outlet 2.
Note that although this method gives us the maximum flow for each outlet, we cannot
always add these values up to find the total maximum flow through the system, because
we might not be able to achieve maximum flow for every outlet at the same time.

The outlets need to be considered
separately.
Outlet 1 The capacity of C1 is: 400 + 800 = 1200
Look for the minimum cut that
The capacity of C2 is: 400 + 300 = 700
prevents water reaching Outlet 1.
Note: The pipe with capacity 200 The capacity of C3 is: 400 + 400 = 800
leading towards Outlet 2 does not need
The capacity of C4 is: 300 + 100 + 300 = 700
to be considered in any cut because
this pipe always prevents water from The capacity of C5 is: 300 + 400 = 700
reaching Outlet 1.
The minimum cut/maximum flow is 700 kilolitres
per minute.
C1 C2 C3 C4 C5
400
Source 1 300
100
800 300 Outlet 1
400
Outlet 2 The capacity of C1 is: 200 + 100 + 300 + 100 = 700

Look for the minimum cut that The capacity of C2 is: 200 + 300 + 300 + 100 = 900
prevents water reaching Outlet 2.
Note: The pipe with capacity 200 The capacity of C3 is: 200 + 300 + 300 + 100 = 900
leading towards Outlet 2 will need to
The capacity of C4 is: 200 + 500 + 100 = 800
be considered in any cut because this
pipe delivers water towards Outlet 2 The capacity of C5 is: 200 + 400 + 300 = 900
and must be ‘cut’ like all the others.
Other cuts are possible, but have not The capacity of C6 is: 600 + 300 = 900
been included in the diagram.
The minimum cut/maximum flow
is 700 kilolitres per minute.
C1
C2 C3
200 C5
200
100 300 200 C6
300 500 400
Source 2
300 600
100
Outlet 2
100 300
C4
Note: The maximum flow through a network with two sources can also be determined by tracking the flow
as outlined in Example 4.

670 Chapter 14 Flow, matching and scheduling problems 14A
Skill-
sheet Exercise 14A
Directed graphs
1 Find the number of vertices that can be reached from vertex A in each of the directed
graphs below.
a b A B
A C
B
C D E F
E
G I
F D H
J L
K
Cuts
Example 2 2 Determine the capacity of each of the C1 C3 C2
cuts in the digraph opposite. The source B C
8
is vertex S and the sink is vertex T . 6 3
3 5 T
S
8 10 3
F 4 E
3 Determine the capacity of each of C1 C2

C3
4
the cuts in the digraph opposite. The
4 4 5
source is vertex S and the sink is 3
3 7
S 8 T
vertex T .
8 2 3
4
Maximum-flow
Example 3 4 Find the maximum flow for each of the following graphs. The source is vertex S and
Example 4 the sink is vertex T .
a A 3 C 5 b A 3 C
4 6 7
S 6 T 2 T
S 7
6 4 8 7
B B 6 D
c A 3 C d A 4 C 8
5 10
T T
5 4 S
4 7
6
5 9 6 10
S 5 B 3 D B 10 D

14A 14A Flow problems 671
Minimum-cut maximum-flow
5 A train journey consists of a cut B cut C
cut D
connected sequence of stages 0
cut A 4
formed by edges on the directed 3 3 9
network opposite from Arlie to 7 4 Bowen
Arlie 7 8
Bowen. The number of available 1
4
seats for each stage is indicated 4 8 1 7 cut E
beside the corresponding edge, 4
as shown on the diagram on the

right.
The five cuts, A, B, C, D and E, shown on the network are attempts to find the
maximum number of available seats that can be booked for a journey from Arlie to
Bowen.
a Write down the capacity of cut A, cut B, cut C and cut D.
b Explain why cut E is not a valid cut when trying to find the minimum cut between
Arlie and Bowen.
c Find the maximum number of available seats for a train journey from Arlie to
Bowen.
Networks with more than one source

Example 5 6 In each of the following, water pipes of different capacities are connected to two water
sources and two sinks. Networks of water pipes are shown in the diagrams below. The
numbers on the edges represent the capacities, in kilolitres per minute, of the pipes.
For each of the following, find the maximum flow, in kilolitres per minute, to each of
the sinks in this network.
a 10
8 4
6 7
Source 1
6 6 Sink 1
5
5
10
9 12
4
12
5
Source 2 Sink 2
3 4 5

672 Chapter 14 Flow, matching and scheduling problems 14A
b
13 6
4
source 1 sink 1
15 2
12
7
9
14 3
source 2
14
1 sink 2
8
6
Analysis of maximum-flow problems

7 The flow of water through a series of pipes, in litres per minute, is shown in the
directed network below.
6 4
5
10 6
7 12
Source Sink
3 7
2 3
8 8
4
9 7 10
a How many different routes from the source to the sink are possible?
b Determine the maximum flow from the source to the sink.
8 The corridors people can walk through to visit Entrance

different exhibits in a museum are given as a 20 30
directed network opposite. To avoid congestion 25 28
around every exhibit, the museum imposes a 15
maximum capacity policy throughout each corridor 30
18 10 32
between exhibits. The numbers on the edges
represent the maximum number of people that can 13 30 25
walk through each corridor of the museum every 30
24 25
30 minutes. 10
a On the network opposite, identify a cut that has 31
a capacity of 80.
Exit
b Determine the maximum flow of people from the entrance to the exit of the
museum.
c One group of primary school students would like to walk through the museum.
The teacher explains that this can happen unsupervised if all students in the group
remain together, not separating to explore different routes. Given that a group of
students must stay together from the entrance to the exit, what is the largest group of
students possible that can pass through the museum every 30 minutes?
14A 14A Flow problems 673

Questions 9 and 10 refer to the 170
200
diagram opposite
120 70
Source 70
The flow of liquid through a series of
170 120
pipes, in litres per minute, is shown in 100
120 120
the directed network opposite. Five 100
170
cuts labelled A to E are shown on the 70 70
100 100
network.
Cut A Sink
Cut B
170 Cut E
Cut C Cut D
9 The capacity of Cut E is

A 70 B 170 C 290 D 390 E 460
10 The number of these cuts with a capacity equal to the maximum flow of liquid from the
source to the sink, in litres per minute, is
A 0 B 1 C 2 D 3 E 4
11 The flow of water through a series of pipes,

in litres per minute, is shown in the network 14
opposite. The weightings of two edges are Source
labelled x. Five cuts labelled A to E are shown 6
x
on the network. The maximum flow of water Cut A 13 11
from the source to the sink, in litres per minute, 17
is given by the capacity of 7
25
x
16
Cut B 10
Sink
Cut C Cut E
Cut D
A Cut A if x = 4 B Cut B if x = 6 C Cut C if x = 8

D Cut D if x = 6 E Cut E if x = 8

14B Matching and allocation problems

Learning intentions
I To be able to define and describe bipartite graphs.
I To be able to solve assignment problems using the Hungarian Algorithm.
Bipartite graphs
In some situations, the vertices of a graph belong Anastasios Trumpet
in two separate sets. Consider a music school
that has four teachers: Anastasios, Panayioti, Panayioti Piano
Olga and Irene. These teachers, between them
can teach four different instruments: trumpet, Olga Violin
piano, violin and clarinet. The teachers and
instruments are represented by vertices, arranged
Irene Clarinet
vertically as shown in the diagram opposite.
The edges of the diagram connect the teachers to the instruments they can teach.
This type of graph is called a bipartite graph. Each edge in a bipartite graph joins a vertex
from one group to a vertex in the other group.
In the situation described above, the school would need to match each teacher to one
instrumental class; this is an example of an allocation problem. The bipartite graph above
graphically shows the instrument(s) that each teacher can teach and can help the school
assign each teacher to an instrument. Anastasios is the only teacher who can teach clarinet,
Irene can only teach violin, therefore Olga must teach piano and Panayioti must teach
trumpet.
Example 6
Nick, Maria, David and Subitha are presenters on a TV travel show. Each presenter will
be assigned a story to film about one country that they have visited before.
Nick has visited Greece, Malaysia and Subitha has visited Malaysia and Brazil
Brazil
Maria has visited Greece and France David has visited Malaysia
Construct a bipartite graph of the information above and use it to decide on the
assignment of each presenter to one country.

14B Matching and allocation problems 675
The two groups of items are: Presenters and Countries. Nick Greece
Draw a vertex for each presenter in one column and each
Maria Malaysia
country in another. Join the vertex of each presenter to the
vertex of each country they have visited. David France
Subitha Brazil
To allocate a presenter to each country, begin by Highlight definite

identifying the vertices with only one edge connected to allocations on the
them. bipartite graph
David is the only presenter who has visited Malaysia, Nick Greece
so he must visit that country.
Maria Malaysia
France has only been visited by one presenter: Maria.
She must be allocated to France, therefore she cannot David France
be allocated to Greece, the other country she has visited Subitha Brazil
before.
Given that each presenter can only visit one country, the
final allocations can be deduced by eliminating impossible
allocations.
Maria is allocated to France, therefore she cannot Nick will be allocated
visit Greece. Nick is the only other presenter who has Greece
visited Greece, therefore he must be allocated that Maria will be allocated
country. France
As David is allocated to Malaysia, Subitha has only David will be allocated
one other country available to visit. Malaysia
Subitha must be allocated to Brazil and this is also Subitha will be
supported by the fact Nick must be allocated to Greece. allocated Brazil
Write the assignments.

The Hungarian algorithm

The graph on the right shows four employee 1 machine 1
employees in a factory. There are four
employee 2 machine 2
different machines that are used in the
production of an item. Every employee
can use every machine and so this is a
complete bipartite graph. Employees
and machines can be matched in many
different ways.
Rather than just assign an employee to a machine randomly, the factory could use
information about how well each employee uses each machine, perhaps in terms of how
quickly each performs the task. The times taken would be the weights on the edges of the
bipartite graph. Rather than writing all of the weights on a complete bipartite graph (which
would be a very complicated diagram), we can summarise the time information in a table
and then use an algorithm, called the Hungarian algorithm, to allocate employees to
machines in order to minimise the time taken to finish the tasks.
The table on the right shows the four Employee A B C D
employees: Wendy, Xenefon, Yolanda
Wendy 30 40 50 60
and Zelda. The machines in a factory are
Xenefon 70 30 40 70
represented by the letters A, B, C and D.
Yolanda 60 50 60 30
The numbers in the table are the times, in Zelda 20 80 50 70
minutes, it takes each employee to finish the
task on each machine.
The table is called a cost matrix. Even though the numbers do not represent money value,
this table contains information about the cost, in terms of time, of employees using each
machine. The cost matrix can be used to determine the best way to allocate an employee to a
machine so that the overall cost, in terms of the time taken to finish the work, is minimised.
The Hungarian algorithm is used to do this.
Performing the Hungarian algorithm
Step 1: Subtract the lowest value in each row, from every value in that row.
30 has been subtracted from every value in

Employee A B C D
the row for Wendy.
Wendy 0 10 20 30
the row for Xenefon. Xenefon 40 0 10 40
30 has been subtracted from every value in Yolanda 30 20 30 0

the row for Yolanda. Zelda 0 60 30 50
the row for Zelda.

14B Matching and allocation problems 677
Step 2: If the minimum number of lines required to cover all the zeros in the
table is equal to the number of allocations to be made, jump to step 6. Otherwise,
continue to step 3.
The zeros can be covered with three lines.

This is less than the number of allocations Employee A B C D
to be made (4). Wendy 0 10 20 30
Continue to step 3. Xenefon 40 0 10 40
Yolanda 30 20 30 0
Zelda 0 60 30 50
Step 3: If a column does not contain a zero, subtract the lowest value in that
column from every value in that column.
Column C does not have a zero.

Employee A B C D
Wendy 0 10 10 30
column C.
Xenefon 40 0 0 40
Yolanda 30 20 20 0
Zelda 0 60 20 50
Step 4: If the minimum number of lines required to cover all the zeros in the
table is equal to the number of allocations to be made, jump to step 6. Otherwise,
continue to step 5a.
The zeros can be covered with three lines.

This is less than the number of allocations Employee A B C D
to be made (4). Wendy 0 10 10 30
Continue to step 5a. Xenefon 40 0 0 40
Yolanda 30 20 20 0
Zelda 0 60 20 50
Step 5a: Add the smallest uncovered value to any value that is covered by two
lines. Subtract the smallest uncovered value from all the uncovered values.
The smallest uncovered element is 10.

Employee A B C D
10 has been added to Xenefon–A and
Wendy 0 0 0 30
Xenefon–D because these values are
covered by two lines. Xenefon 50 0 0 50
10 has been subtracted from all the Yolanda 30 10 10 0

uncovered values. Zelda 0 50 10 50

678 Chapter 14 Flow, matching and scheduling problems 14B
Step 5b: Repeat from step 4.
The zeros can be covered with a minimum

of four lines. This is the same as the Employee A B C D
number or allocations to make. Wendy 0 0 0 30
Continue to step 6. Xenefon 50 0 0 50
Yolanda 30 10 10 0
Zelda 0 50 10 50
Step 6: Draw a bipartite graph with an edge for every zero value in the table.
In the bipartite graph: Wendy A
Wendy will be connected to A, B and C

Xenefon B
Xenefon will be connected to B and C
Yolanda will be connected to D
Zelda will be connected to A. Yolanda C
Zelda D
Step 7: Make the allocation and calculate minimum cost
Zelda must operate machine A (20 minutes).

Yolanda must operate machine D (30 minutes).
Wendy can operate either machine B (40 minutes) or C (50 minutes).
Xenefon can operate either machine B (30 minutes) or C (40 minutes).
Note: Because Wendy and Xenefon can operate either B or C, there are two possible allocations. Both
allocations will have the same minimum cost.
The minimum time taken to finish the work = 20 + 30 + 50 + 30 = 130 minutes.
Skill-
sheet Exercise 14B
Bipartite graphs
Example 6 1 a On Monday, three workers are each to be allocated one task at work. The bipartite
graph below shows which task(s) each person is able to complete.
Worker 1 Task 1
Worker 2 Task 2
Worker 3 Task 3
If each person completes a different task, write down the task each worker must be
allocated to on Monday.

14B 14B Matching and allocation problems 679
b On Tuesday, the same three workers will be allocated to a new set of tasks. The
bipartite graph below shows which task(s) each person is able to complete.
Worker 1 Task 4
Worker 2 Task 5
Worker 3 Task 6
Given that Worker 2 must complete Task 6, write down the new task each worker
must be allocated to on Tuesday.
2 It is Miko’s birthday and his sister Aria has asked some of his friends to assist with the
celebrations by purchasing some items for a party. The bipartite graph below shows
which item(s) each person is able to purchase on their way to the party.
Niranjan Cake
Nishara Serviettes
Dhinesh Balloons
Dhishani Candles
Each friend must purchase an item. Write down which item each friend must purchase.
3 The sport of ice hockey has six player positions: goalie, left defence, right defence,
right wing, left wing and centre. A group of six have decided to play. Only one person
is happy to play goalie. The other five people must be allocated to the other five
positions.
The bipartite graph below shows which positions each of the five players can play.
Player 1 Centre
Player 2 Right wing
Player 3 Right defence
Player 4 Left wing
Player 5 Left defence
Each player plays a different position. Write down two possible allocations, describing
which position each player must play.

4 Gloria, Minh, Carlos and Trevor are buying ice-cream. They have a choice of five
flavours: chocolate, vanilla, peppermint, butterscotch and strawberry. Gloria likes
vanilla and butterscotch, but not the others. Minh only likes strawberry. Carlos likes
chocolate, peppermint and butterscotch. Trevor likes all flavours.
a Explain why a bipartite graph can be used to display this information.
b Draw a bipartite graph with the people on the left and flavours on the right.
c What is the degree of the vertex representing Trevor?
The Hungarian algorithm

5 a A cost matrix is shown. Find the A B C D
allocation(s) by the Hungarian algorithm
W 110 95 140 80
that will give the minimum cost.
X 105 82 145 80
Y 125 78 140 75
Z 115 90 135 85
b Find the minimum cost for the given cost matrix A B C D
and give a possible allocation.
W 2 4 3 5
X 3 5 3 4
Y 2 3 4 2
Z 2 4 2 3
6 A school is to enter four students Student 100 m 400 m 800 m 1500 m

in four track events: 100 m, 400 m,
Dimitri 11 62 144 379
800 m and 1500 m. The four students’
John 13 60 146 359
times (in seconds) are given in the
Carol 12 61 149 369
table. The rules permit each student
Elizabeth 13 63 142 349
to enter only one event. The aim is to
obtain the minimum total time.
Use the Hungarian algorithm to select the ‘best’ student for each event.
7 Three volunteer workers, Joe, Meg and Ali, are Job

available to help with three jobs. The time (in
Student A B C
minutes) in which each worker is able to complete
each task is given in the table opposite. Which Joe 20 20 36
Meg 16 20 44
allocation of workers to jobs will enable the jobs
Ali 26 26 44
to be completed in the minimum time?

14B 14B Matching and allocation problems 681
8 A company has four machine operators Machine

and four different machines that they can
Operator W X Y Z
operate. The table shows the hourly cost
in dollars of running each machine for A 38 35 26 54
B 32 29 32 26
each operator. How should the machinists
C 44 26 23 35
be allocated to the machines to minimise
D 20 26 32 29
the hourly cost from each of the machines
with the staff available?
9 A football association is scheduling football games to be played by three teams (the

Champs, the Stars and the Wests) on a public holiday. On this day, one team must play
at their Home ground, one will play Away and one will play at a Neutral ground.
The costs (in $’000s) for each team to play at each of the grounds are given in the table
below.
Determine a schedule that will minimise the Team Home Away Neutral
total cost of playing the three games and
Champs 10 9 8
determine this cost.
Stars 7 4 5
Note: There are two different ways of scheduling the
games to achieve the same minimum cost. Identify both Wests 8 7 6
of these.
10 A roadside vehicle assistance Motorist

organisation has four service vehicles
Service vehicle Jess Mark Raj Karla
located in four different places. The
table below shows the distance (in A 18 15 15 16
kilometres) of each of these service B 7 17 11 13
vehicles from four motorists in need of C 25 19 18 21
roadside assistance. D 9 22 19 23
Determine a service vehicle assignment that will ensure that the total distance travelled
by the service vehicles is minimised. Determine this distance.

11 The sport of futsal has five player positions: Aaliyah Forward
goalkeeper, forward, defender, right wing and
left wing. In a group of five friends, Ezekiel will Brock Defender
always play goalkeeper, but the other four friends
Aaliyah, Brock, Corazon and Daniel will rotate Corazon Right wing
their responsibilities and are able to play a number
of positions each. Daniel Left wing
The bipartite graph below shows which positions
each of the four friends can play.

Based on the bipartite graph, which one of the following allocations is not possible?
A Friend Position
B Friend Position
Aaliyah Right wing Aaliyah Right wing
Brock Forward Brock Defender
Corazon Defender Corazon Forward
Daniel Left wing Daniel Left wing
C Friend Position
D Friend Position
Aaliyah Forward Aaliyah Forward
Brock Right wing Brock Defender
Corazon Left wing Corazon Left wing
Daniel Defender Daniel Right wing
E Friend Position
Aaliyah Forward
Brock Right wing
Corazon Defender
Daniel Left wing

Five people work at a bank. Each person will perform one task. The time taken for each
person to complete tasks 1, 2, 3, 4 and 5, in hours, is shown in the table below.
Anita Brad Carmen Dexter Electra

Task 1 1 2 2 5 4
Task 2 4 9 7 11 6
Task 3 5 3 3 9 4
Task 4 8 5 6 6 7
Task 5 5 8 4 6 9
12 The manager of the bank wants to allocate the tasks so as to minimise the total time
taken to complete the five tasks. If each person starts their allocated task at the same
time, then the first person to finish could be either
A Anita or Brad B Anita or Elektra C Brad or Carmen
D Brad or Dexter E Brad or Elektra
13 Before the tasks are performed, it is found that Elektra will only require 4 hours to
complete Task 5 rather than 9 hours. If the tasks are allocated based on this new
information, the minimum total time for all tasks will
A increase by 4 days. B decrease by 4 days. C decrease by 3 days.
D decrease by 2 days. E decrease by 1 day.

14B 14C Precedence tables and activity networks 683
14 Four people, Xena, Wilson, Yasmine, Zachary, are each assigned a different job by
their manager. The table below shows the time, in hours, that each person would take
to complete each of the four jobs.
Job 1 Job 2 Job 3 Job 4

Xena 5 3 7 p
Wilson 1 2 5 6
Yasmine 1 7 1 5
Zachary 4 7 6 p
Wilson takes 6 minutes to complete Job 4, while Yasmine only takes 5 minutes to
complete Job 4. Both Xena and Zachary take p minutes to complete Job 4.
The manager will allocate the jobs as follows:
Job 1 to Wilson Job 2 to Xena
Job 3 to Yasmine Job 4 to Zachary
This allocation will achieve the minimum total completion time if the value of p is not
greater than
A 6 B 7 C 8 D 9 E 10
14C Precedence tables and activity networks

Learning intentions
I To be able to identify activities in a project.
I To be able to understand the precedence that some activities have over others in a
project.
I To be able to identify immediate predecessors of activities from an activity network.
I To be able to draw an activity network from a precedence table.
I To be able to understand and explain the need for dummy activities in projects.
I To be able to include dummy activities in activity networks as required.
Drawing activity networks from precedence tables

Building a house, manufacturing a product, organising a wedding and other similar projects
all require many individual activities to be completed before the project is finished. The
individual activities often rely upon each other and some can’t be performed until other
activities are complete.
In the organisation of a wedding, invitations would be sent out to guests, but a plan for
seating people at the tables during the reception can’t be completed until the invitations are
accepted. When building a house, the plastering of the walls can’t begin until the house is
sealed from the weather.

For any project, if activity A must be completed before activity B can begin then activity A
is said to be an immediate predecessor of activity B. The activities within a project can
have multiple immediate predecessors and these are usually recorded in a table called a
precedence table.
This precedence table shows some of the activities
Immediate
involved in a project and their immediate predecessors. Activity predecessors
The information in the precedence table can be used to A −
draw a network diagram called an activity network. B −
Activity networks do not have labelled vertices, other than C A
the start and finish of the project. The activities in the D B
project are represented by the edges of the diagram and so E B
it is the edges that must be labelled, not the vertices. F C, D
G E, F
Activities A and B have no immediate A

predecessors. start
These activities can start immediately
B
and can be completed at the same time.
C
Activity A is an immediate predecessor A F
of activity C, so activity C must start
follow immediately after activity A.
B
Activity C is an immediate predecessor
of activity F, so activity F must follow
immediately after activity C.
C
Activity D has immediate predecessor A F
activity B so it follows immediately start
D
after activity B.
B
Activity D is also an immediate
predecessor of activity F so activity F
must follow immediately after activity D.
C
Activity E has immediate predecessor F
A
activity B so it will follow immediately start
D finish
after activity B. G
B E
Activity G has immediate predecessor
activity F and activity E and so it must
follow immediately after both of these
activities.
Activity G is not an immediate predecessor for any activity and so the project is finished
after this activity is complete.

14C Precedence tables and activity networks 685
Activity networks
When activity A must be completed before activity B can begin, activity A is called an
immediate predecessor of activity B.
A table containing the activities of a project, and their immediate predecessors, is called a
precedence table.
An activity network can be drawn from a precedence table. Activity networks have
edges representing activities. The vertices are not labelled, other than the start and finish
vertices.
Example 7 Drawing an activity network from a precedence table.
Draw an activity network from the

Activity Immediate predecessors
precedence table shown on the right.
A −
In this solution, the activity network will B A
be drawn from the finish back to the start. C A
D A
E B
F C
G D
H E, F, G
H is not an immediate predecessor for any finish

other activity so it will lead to the finish of H
the project.
H has immediate predecessors E, F and G E

and so these three activities will lead into finish
F
activity H. H
Activity D is an immediate predecessor of E

activity G and has immediate predecessor start F
finish
activity A. There will be a path through A H
G
activity A, activity D and then activity G.
D
Activity C is an immediate predecessor of E

activity F and has immediate predecessor start F
finish
C G
activity A, activity C and then activity F.
ISBN 978-1-009-11041-9 © Peter Jones et al 2023 D Cambridge University Press
Activity B is an immediate predecessor of E

activity E and has immediate predecessor start F
B finish
C G
activity A, activity B and then activity E.
D
Activity A has no immediate predecessors, E

so it is the start of the project. start F
B finish
A H
C G
Sketching activity networks

Activities that have no immediate predecessors follow from the start vertex.
Activities that are not immediate predecessors for other activities lead to the finish vertex.
For every other activity, look for:
which activities it has as immediate predecessors
which activities it is an immediate predecessor for.
Construct the activity network from this information.
Dummy activities
Sometimes two activities will have some of the same
Immediate
immediate predecessors, but not all of them. In this
Activity predecessors
very simple precedence table, activity D and activity
A −
E share the immediate predecessor activity B, but they
both have an immediate predecessor activity that the B −
other does not. C −
This overlap of predecessors presents some difficulty D A, B

when constructing the activity network, but this E B, C
difficulty is easily overcome.
Activity D and activity E are not immediate
A D
predecessors for any other activity, so they will
B
lead directly to the finish vertex of the project. start finish
Activites A, B and C have no immediate E

C
predecessors, so they will follow directly from
the start vertex of the project.

14C Precedence tables and activity networks 687
The start and finish of the activity network are shown in the diagram above. We need to use
the precedence information for activity D and activity E to join these two parts together.
Activity D needs to follow directly from activity A and activity B, but we can only draw one
edge for activity D. Activity E needs to follow directly from both activity B and activity C,
but again we only have one edge for activity E, not two.
The solution is to draw the diagram with
A D
activity D starting after one of its immediate dummy
B
predecessors, and using a dummy activity for start finish
the other. The dummy activities are represented
E
by dotted edges and are, in effect, imaginary. C
They are not real activities, but they allow all of
the predecessors from the table to be correctly represented.
The dummy activity for D allows activity D to directly follow both activity A and B.
A dummy activity is also needed for activity E
A D
because it, too, has to start after two different dummy
B
activities, activity B and C. start finish
dummy
E
C
Dummy activities
A dummy activity is required if two activities share some, but not all, of their immediate
predecessors.
A dummy activity will be required from the end of each shared immediate predecessor to
the start of the activity that has additional immediate predecessors.
Dummy activities are represented in the activity network using dotted lines.
Example 8 Using a dummy activity in an activity network
Draw an activity network from the

Immediate
precedence table shown on the right. Activity predecessors
A −
B −
C A
D B
E C, D
F C
G E, F

A and B will lead from the C

start vertex. A F
G will lead to the end vertex. start G
dummy finish
A dummy will be required
B E
from the end of activity
C (shared immediate D
predecessor) to the start of
activity E (the activity with
an additional immediate
predecessor.)
Example 9 Creating a precedence table from an activity network
Write down a precedence D

table for the activity
B G
network shown on the dummy
A I
right. start F
dummy
C J
E H
1 Create a table with a row for each activity.
Immediate
2 Look at the start of an activity. Write down all of Activity predecessors
the activities that lead directly to this activity in
A −
the immediate predecessor column.
B A
3 Activity C is a predecessor of activity E, and the
dummy activity makes it also a predecessor of F C A
and G. D B
4 Activity G is a predecessor of activity I, and the E C
dummy activity makes it also a predecessor of J. F D, C
G D, C
H E, F
I G
J G, H

14C 14C Precedence tables and activity networks 689
Skill-
sheet Exercise 14C
Constructing activity networks from precedence tables

Example 7 1 Draw an activity network for each of the precedence tables below.
a b
Immediate Immediate
Activity predecessors Activity predecessors
A − P −
B A Q −
C A R P
D B S Q
E C T R, S
c d
Immediate Immediate
T − F −
U − G −
V T H −
W U I F
X V, W J G, I
Y X K H, J
Z Y L K
e f
Immediate Immediate
K − A −
L − B −
M K C −
N M D B
O N, L E A, D
P O F E, C
Q P G F
R M H G
S R, Q I E, C
J G
K H, I

690 Chapter 14 Flow, matching and scheduling problems 14C
Constructing activity networks requiring dummy activities from precedence tables

Example 8 2 Draw an activity network for the following precedence tables. Dummy activities
will need to be used.
a b
Immediate Immediate
F − A −
G − B A
H F C A
I H, G D B
J G E B, C
c d
Immediate Immediate
P − A −
Q − B A
R P C A
S Q D B, C
T Q E C
U R, S F E
V R, S , T G D
H F, G
Constructing precedence tables from activity networks

Example 9 3 Write down a precedence table for the activity networks shown below.
a D
A F
start finish
C
B G
E
b T
Q
P X
start S U finish
V
R W
c L
Q
J
T
M
start P finish
K N R
O S

14C 14C Precedence tables and activity networks 691
d C
A
D
start E
F
dummy
B G
finish
H
e R U
P
S X
V Z
start finish
Q
dummy
T Y
W
f E H
B K
A C
start F finish
dummy
J
dummy
D I
G
4 The following activity network shows the

Activity Description
activities in a project to repair a dent in a car
A Remove panel
panel. The activities are listed in the table on the
right. B Order component
A C D
E C Remove broken
start dummy G
finish component
B F D Pound out dent
a Which activity or activities are the immediate E Repaint
predecessors of the event ‘remove broken F Install new
component’? component
b Which activities are the immediate G Replace panel
predecessors of the activity ‘install new
component’?
Analysis of activity networks and precedence tables

5 Consider the following activity network for a project.
D H M
Finish
A I
B E
Start O
J
C F G K
N
L
692 Chapter 14 Flow, matching and scheduling problems 14C
a Write down a precedence table for the network above.

b Write down the two paths from start to finish that begin with activity A.
c Write down the four paths from start to finish that begin with activity B.
d Write down the four paths from start to finish that begin with activity C.
6 Consider the following activity network.

C
A F
I
Start K
D
H Finish
B
L
E G J
a Which activities are immediate predecessors of activity I?

b Which activities must be completed before activity I can commence?
7 The information in the table opposite can be used Immediate

to complete a directed network. This network will Activity predecessors
require a dummy activity. A −
a Which activity will the dummy activity be drawn B −
from the end of? C A
b Which activity will the dummy activity be drawn D B
E A
to the start of?
F C, D
c Why is it necessary to include a dummy activity in G B
this network? H F
I G
J I
K G, H, J
L E

8 The activity network below shows the sequence of activities required to complete a
project.
G
A C M
H
E L
Start Finish
D
B F J
K
I
Beginning with activity B, the number of paths from start to finish is

A 1 B 2 C 3 D 4 E 5

14C 14D Scheduling problems 693

A project involves eight activities, A to H. Immediate
The immediate predecessor(s) of each activity is shown in Activity predecessors
the table opposite. A directed network for this project will A −
require a dummy activity. B −
C −
9 The dummy activity will be drawn from the end of D A
A activity B to the start of activity F. E C
F B, E
B activity B to the start of activity E.
G B
C activity G to the start of activity F. H G
D activity G to the start of activity E.
E activity E to the start of activity F.
10 The number of paths from start to finish is

A 4 B 5 C 6 D 7 E 8
14D Scheduling problems

Learning intentions
I To be able to understand activity networks that include weights (durations) of each
activity.
I To be able to determine the EST for activities using forward scanning.
I To be able to determine the LST for activities using backward scanning.
I To be able to calculate and understand the existence of float times for some activities.
I To be able to identify the critical path and completion time of a project.
Scheduling
Projects that involve multiple activities are usually completed against a time schedule.
Knowing how long individual activities within a project are likely to take allows managers
of such projects to hire staff, book equipment and also to estimate overall costs of the
project. Allocating time to the completion of activities in a project is called scheduling.
Scheduling problems involve analysis to determine the minimum overall time it would take
to complete a project.

Weighted precedence tables

The estimated time to complete activities within a project can be recorded in a precedence
table, alongside the immediate predecessor information.
A precedence table that contains the Estimated
estimated duration, in days, of each completion Immediate
activity is shown on the right. Activity time (days) predecessors
These estimated times are the weights A 8 −
for the edges of the activity network and B 6 −
need to be recorded alongside the name C 1 A
of the activity on the graph. D 2 B
A dummy activity is required from E 3 C
the end of activity C (C is repeated in F 1 C, D
immediate predecessors) to the start of G 2 E, F
activity F (F is the activity that has an H 1 G
extra immediate predecessor).
The weight (duration) of dummy activities is always zero.
The activity network for this project is shown below.

C, 1
A, 8 E, 3
G, 2 H, 1
start dummy, 0
B, 6 D, 2
F, 1
Float times
The diagram below shows a small section of a different activity network. There are three
activities shown, with their individual durations, in hours.
Activity B and activity C are both immediate A, 5 B, 3
predecessors to the next activity, so the project
cannot continue until both of these tasks are finished. C, 6
Activity B cannot begin until activity A is finished.
Activity C can be completed at the same time as activity A and activity B.
Activity A and B will take a total of 5 + 3 = 8 hours, while activity C only requires 6 hours.
There is some flexibility around when activity C needs to start. There are 8 − 6 = 2 hours
spare for the completion of activity C. This value is called the float time for activity C.

14D Scheduling problems 695
The flexibility around the starting time for activity C can be demonstrated with the following
diagram.
A A A A A B B B
Start at same time C C C C C C Slack Slack
Delay C by 1 hour Slack C C C C C C Slack
Delay C by 2 hour Slack Slack C C C C C C
The five red squares represent the 5 hours it takes to complete activity A. The three green
squares represent the 3 hours it takes to complete activity B.
The six yellow squares represent the 6 hours it takes to complete activity C. Activity C does
not have to start at the same time as activity A because it has some slack time available
(2 hours).
Activity C should not be delayed by more than 2 hours because this would cause delays to
the project. The next activity requires B and C to be complete before it can begin.
Calculating and recording earliest starting times (EST)

In order for a project to be completed in the shortest time possible, it is important that certain
key activities start at the earliest possible time. The earliest starting time, or EST, for each
activity is the earliest time after the start of the entire project that the individual activity can
start. An EST of 8 means an activity can start 8 hours (or whatever time period is given)
after the start of the project.
The EST for each activity is found by a process called forward scanning.
Forward scanning
Forward scanning will be demonstrated using the activity network below. The durations of
each are in days.
1 Draw a box, split into two C, 1
cells, next to each vertex of A, 8 E, 3
(for A)
the activity network, as shown G, 2 H, 1
start dummy, 0 finish
in the diagram opposite. If (for B)
B, 6 D, 2 F, 1
more than one activity begins
at a vertex, draw a box for
each of these activities.
2 Activities that begin at the C, 1
start of the project have an A, 8 E, 3
(for A) 0
EST of zero (0). Write this dummy, 0 G, 2 H, 1
start finish
in the left box, shown shaded (for B) 0 D, 2
B, 6 F, 1
yellow in the diagram.

3 Calculate the EST of each 8 C, 1 9

activity of the project by A, 8 E, 3
(for A) 0
adding the EST of the G, 2 H, 1
immediate predecessor to the (for B) 0 D, 2
B, 6 F, 1
duration of the immediate
6
predecessor.
Notes: EST of C = EST of A + duration of A (EST of C = 0 + 8 = 8)
EST of D = EST of B + duration of B (EST of D = 0 + 6 = 6)
EST of E = EST of C + duration of C (EST of C = 8 + 1 = 9)
4 If an activity has more than 8 C, 1 9

one predecessor, calculate A, 8 E, 3
(for A) 0
the EST using each of the G, 2 H, 1
predecessors and choose the (for B) 0 D, 2 12
B, 6 F, 1
largest value.
6 9
Notes: EST of F = 6 + 2 = 8 or EST of F = 9 + 0 = 9. Use 9.
EST of G = 9 + 1 = 10 or EST of G = 9 + 3 = 12. Use 12.
5 The EST value at the finish 8 C, 1 9

of the project is the minimum A, 8 E, 3
(for A) 0
time it takes to complete the G, 2 H, 1
project. (for B) 0 D, 2 12 14 15
B, 6 F, 1
6 9
Notes: The minimum time to complete this project is 15 days.
Calculating and recording latest starting times (LST)

Some activities, as we saw earlier, have some flexibility around the time that they can start.
The latest start time, or LST, for each activity is the latest time after the start of the entire
project that the individual activity can start. LSTs for each activity are calculated using
the reverse of the process used to calculate the ESTs. This process is called backward
scanning.
Backward scanning
Backward scanning will be demonstrated using the activity network with completed forward
scanning from above.
1 Copy the minimum time to 8 C, 1 9
complete the project into the A, 8 E, 3
(for A) 0
right cell shown shaded blue dummy, 0 G, 2 H, 1
start finish
B, 6
in the diagram. (for B) 0 12 14 1515
D, 2 F, 1
6 9

2 Calculate the LST for each 8 C, 1 9

activity by subtracting the A, 8 E, 3
(for A) 0
duration of the activity from dummy, 0 G, 2 H, 1
start finish
B, 6
the LST of the following (for B) 0 1212 1414 1515
D, 2 F, 1
activity.
6 9 11
Notes: 1 LST of H = LST of finish – duration of H (LST of H = 15 − 1 = 14).
2 LST of G = LST of H – duration of G (LST of G = 14 − 2 = 12).
3 LST of F = LST of G – duration of F (LST of F = 12 − 1 = 11).
4 LST of E = LST of G – duration of E (LST of E = 12 − 3 = 9).
3 If more than one activity 8 C, 1 9

have the same predecessor, A, 8 E, 3
(for A) 0
calculate the LST using each dummy, 0 G, 2 H, 1
start finish
B, 6
of the activities that follow (for B) 0 1212 1414 1515
D, 2 F, 1
and choose the smallest
6 9 11
value.
Note: LST of E = 12 − 3 = 9 (from duration of E) or LST of E = 11 − 0 = 11 (from duration of
dummy). Use 9.
The completed activity network with all EST and LST is shown below.
8 8 9 9
C, 1
A, 8 E, 3
(for A) 0 0
G, 2 H, 1
B, 6
(for B) 0 3 12 12 14 14 15 15
D, 2
F, 1
6 9 9 11
Identifying float times and the critical path

The boxes at the vertices in the activity network above give the EST and LST for the activity
that begins at that vertex.
2 The EST for activity D is 6 and the LST

D for activity D is 9. This means activity D
6 9
has a float time of 9 − 6 = 3 hours.
Activity D can be delayed by 3 hours without delaying the rest of the project.
8 8 The EST for activity C is 8 and the LST

for activity C is 8. This means activity C
C, 1 has a float time of 8 − 8 = 0 hours.
Activity C has no flexibility around its starting time at all. Any delay to the start of this
activity will delay the whole project and extend the minimum time for completion.

Activities that have no float time are critical ones for completion of the project. Tracking
through the activity network along the edges of critical activities gives the critical path for
the project. The critical path for this project is highlighted in red on the diagram below.
8 8 9 9
C, 1
A, 8 E, 3
(for A) 0 0
G, 2 H, 1
B, 6
(for B) 0 3 12 12 14 14 15 15
D, 2
F, 1
6 9 9 11
Critical path
A critical path is the longest or equal longest path in an activity network.
There can be more than one critical path in an activity network.
The critical path is the sequence of activities that cannot be delayed without affecting
the overall completion time of the project.
The process for determining the critical path is called critical path analysis.
Critical path analysis

Draw a box with two cells next to each vertex of the activity network.
Calculate the EST for each activity by forward scanning:
EST = EST of predecessor + duration of predecessor
If an activity has more than one predecessor, the EST is the largest of the alternatives.
The minimum overall completion time of the project is the EST value at the end
vertex.
Calculate the LST for each activity by backward scanning:
LST = LST of following activity − duration of activity
If an activity has more than one following activity, the LST is the smallest of the
alternatives.
Float = LST − EST
If float time = 0, the activity is on the critical path.

Example 10 Finding the critical path from a precedence table
A project has six activities as shown in the

Duration Immediate
precedence table opposite and the associated
Activity (days) predecessors
activity network is shown below.
A 6 −
a Complete the critical path analysis to calculate
B 9 −
the EST and LST for each activity.
C 8 A, B
b Write down the critical path of this project.
c What is the minimum time required to D 7 A, B
complete the project? E 10 D
F 12 C, E
C, 8 F, 12
Finish
A, 6
D, 7 E, 10
Start
dummy, 0
B, 9
a C: 9 18
F: 26 26 38 38
D: 9 9 F, 12
C, 8 Finish
A: 0 3 A, 6
Start D, 7
B, 9 E, 10
B: 0 0 dummy, 0
E: 16 16
b The critical path is highlighted in red. The critical path of this project is
Note: The dummy is not included in the critical
B → D → E → F.
path.
c The minimum completion time is in EST of The minimum completion time of
the end box. this project is 38 days.

Example 11 Finding the critical path
A project has eight activities as shown in the

Duration Immediate
precedence table opposite.
Activity (weeks) predecessors
a Draw an activity network for this project.
A 3 −
b Complete the critical path analysis to calculate
B 5 −
the EST and LST for each activity.
C 12 −
c What is the earliest starting time for activity H?
D 4 A
d What is the latest starting time for activity H?
e What is the float time of activity H?
E 7 A
f Write down the critical path of this project. F 5 B, D

g What is the minimum time required to G 6 E, F
complete the project? H 2 C
h The person responsible for completing activity
E falls sick three weeks into the project. If he
will be away from work for two weeks, will
this cause the entire project to be delayed?
a A, B and C have no predecessors and so A,3

E,7
can begin at the same time. Continue D,4 G,6
Start Finish
drawing the network as outlined by B,5 F,5
the table above, including arrowheads,
C,12 H, 2
activity labels, duration labels and
correct immediate predecessors.
b Begin by drawing boxes, split into two D:
E:
cells, at the beginning of each activity. A,3
E,7
Label them with the name of each A: D,4 G,6
Start Finish
activity. You must also include a box, B: B,5 F: F,5 G:
C:
split into two cells, at the final vertex H, 2
C,12
where the project finishes.
H:

Use forward scanning to identify the D: 3
E: 3
EST for each activity. Activities with no
A,3
E,7
immediate predecessors always have an A: 0 D,4 G,6
EST of zero. Add the left cell value at Start Finish
B,5 F: 7 F,5 G: 12 18
B: 0
the start of the activity to the duration C: 0
and write the result in the left cell at the C,12 H, 2
end of the activity. Use the largest of the

H: 12
possibilities if there is more than one
activity ending at the same vertex. Identify
the minimum project completion time as
the left cell value at the finish vertex
Use backward scanning to identify
D: 3 3
the LST for each activity. Subtract the E: 3 5
duration from the right cell value at the A,3
E,7
A: 0 0 D,4
end of the activity and write the result in G,6
Start Finish
the right cell at the start of the activity. B: 0 2
B,5 F: 7 7 F,5 G: 12 12 18 18
Use the smallest of the possibilities if C: 0 4
C,12 H, 2
there is more than one activity beginning
at the same vertex.
H: 12 16
c EST values are in the left cell at the The EST for activity H is 12 weeks
start of each activity.
d LST values are in the right cell at the The EST for activity H is 16 weeks
end of each activity.
e Float = LS T − ES T . Float H = LS T − ES T
= 16 − 12
= 4 weeks
f The critical path joins all of the
D: 3 3
activities that have the same EST and
E: 3 5
LST, and therefore which have zero A,3
E,7
float time. A: 0 0 D,4 G,6
Start Finish
B,5 F: 7 7 F,5 G: 12 12 18 18
B: 0 2
C: 0 4
C,12 H, 2
H: 12 16
The critical path for this project is

A−D−F −G

702 Chapter 14 Flow, matching and scheduling problems 14D
g The minimum time required to 18 weeks

complete the project is the EST (also,
always equal to the LST) at the finish
vertex.
h If the float time is more or equal to The person will be away for two weeks,
the delay in the start of activity E, the starting 3 weeks into the project. This is
project will not be affected. equal to the float time for activity E, and
so delaying the start of activity E until the
person comes back to work will not affect
the overall completion time of the project.
Skill-
sheet Exercise 14D
Calculations from elements of an activity network

1 Write down the value of each pronumeral in the sections of activity networks below.
a 8 8 p 12 b 4 8 w 14
P, 4 W, 6
c m n 12 12 d 6 a
M, 4 A, 5
c 15
B, 3
14 b
e 3 f f 4 4 12 12
F, 6 Q, q
R, r
dummy P, p
n 9
5 G, 7 g

14D 14D Scheduling problems 703
2 Consider the section of an activity network shown in the diagram opposite.

a What is the duration of 9 9
activity A?
A C
b What is the critical path 6 6
through this section of the
15 15
activity network?
c What is the float time of B D
activity B?
8 13
d What is the latest time that
activity D can start?
e What is the duration of activity D?
3 Consider the section of an activity network shown in the diagram below.

a What is the duration of 3 6
activity B?
A
b What is the latest start time C, 4
for activity E? E, 2 12 12 F
c What is the earliest time D 9 10

that activity E can start? 0 0 B dummy
d What is the float time for
12 12
activity E?
e What is the duration of activity A?
f What is the duration of activity D?
4 An activity network is 3 4 12 12
B, 8
shown in the diagram
opposite. A, 3 E, 10
0 0 D, 12 42 42
a Write down the critical
path for this project. C, 7 F, 20
22 22
b Calculate the float
times for non-critical
activities.
5 An activity network is shown in the diagram below.

14 14
F, 1
10 11 13 13
D, 2
A, 10
13 14 H, 3
B, 9 E, 4 G, 1
0 0 18 18
9 9
14 15
C, 3 I, 2 J, 1
17 17
a Write down the critical path.

b Write down the float times for all non-critical activities.
6 A precedence table and activity network for a project are shown below. The precedence
table is incomplete.
6 6 12 12 19 19 22 22
Duration Immediate C, 6 E, 7 G, 3
A, 3
Activity (weeks) predecessors dummy, 0 H, 3
dummy, 0
A 3 − 0 0 B, 6
D, 5 F, 1
B 6 − 6 6 11 12 12 19
I, 2
C 6
D B
E 7
F 1 D
G E
H 3
I 2 B
a Complete the table above.

b Write down the critical path for this project.
Critical path analysis from a given activity network

Example 10 7 Consider the following activity network for a project. The duration of each activity is
given in the network, in days.
A, 3 D, 5
E, 2
Start Finish
B, 7
G, 4
C, 5
F, 2
a Determine the earliest start time for activity E.

b Find the minimum completion time for this project.
c Write down the critical path for this project.
d Which activity has a float time of two days?
A, 4 C, 2
E, 3
Start G, 4
F, 5
B, 7 Finish
D, 5 H, 2

a Write down the three activities that are immediate predecessors of activity H.
b Determine the earliest start time of activity H.
c For activity H the earliest start time and the latest start time are the same. What does
this tell us about activity H?
d Determine the minimum completion time, in hours, for this project.
e Which activity could be delayed for the longest time without affecting the minimum
completion time of the project?

A, 10 H, 4
Start C, 8
E, 7 I, 7
F, 9 Finish
B, 13
K, 1
D, 4
G, 12 J, p
a Determine the earliest start time for the following activities:

i H ii I iii J
b Determine the value of p, in weeks, that would create more than one critical path.
c If the value of p is 3 weeks, what will be the float time, in weeks, of activity G.

C, 1 F, 8
A, 5 H, 7
dummy Finish
Start G, 10
B, 7 J, 2
D, 3 I, 3
E, 5
a Write down the two immediate predecessors of activity G.

b Which of the ten activities must be completed before activity I can begin?
c Write down the critical path of for this project.
d Determine the float time of activity E.
e Which three activities could have their completion times increased by two days
without altering the minimum completion time?

H, 9 K, 5
A, 5 F, 7
D, 6 L, 3 Finish
B, 2 I, 11
Start
M, 3
E, 3 dummy
C, 4
J, 10
G, 11
a Complete a precedence table for this network, using two columns, one column for
the activities and a second column for the Immediate predecessors.
b How many activities have an earliest start time of 16 hours?
c Find the latest start time of activity F.
d There are two critical paths. Write down both critical paths.
e How many activities can be delayed by 1 hour without increasing the minimum
Critical path analysis from precedence table only

Example 11 12 The precedence table for a project is
Duration Immediate
shown opposite. Activity (weeks) predecessors
a Draw an activity network for this P 4 −
project. Q 5 −
b Complete the critical path analysis R 12 −
to calculate the EST and LST for S 3 P
each activity. T 6 Q
c Write down the critical path of this U 3 S
project. V 4 R
d What is the minimum time required W 8 R, T , U
to complete the project? X 13 V
Y 6 W, X

13 The precedence table for a project is

Duration Immediate
shown opposite. Activity (weeks) predecessors
a Draw an activity network for this I 2 −
project. J 3 −
b Complete the critical path analysis K 5 −
to calculate the EST and LST for L 4 I
each activity. M 8 J, N
c Write down the critical path of this N 1 K
project O 6 L, M
d What is the minimum time required P 6 J, N
to complete the project? Q 7 J, N
R 5 K
S 1 O
T 9 Q, R

Use the following information to answer questions 14, 15 and 16
The directed network opposite shows A, 4
D, 6 F, 5
the sequence of eleven activities that
are needed to complete a project. The Start
B, 7 H, 7
time, in days, that it takes to complete
C, 7 G, 6
each activity is also shown. Finish
I, 9
E, 9 K, 3
J, 4
14 The earliest starting time, in days, for activity J is

A 12 B 13 C 14 D 15 E 16
15 The number of activities that have exactly two immediate predecessors is

A 1 B 2 C 3 D 4 E 5
16 How many of these activities could be delayed without affecting the minimum
A 3 B 4 C 5 D 6 E 7
17 The directed graph opposite D, 3 G, 5

shows the sequence of
A, 7 H, 2 K, 6
activities required to complete
E, 6
a project. The time taken to Start
B, 4 L, 8
complete each activity, in N, 5
F, 3 M, 5
weeks, is also shown. C, 6 I, 7
Finish
ISBN 978-1-009-11041-9 © Peter Jones et al 2023 O, 9 University Press
Cambridge
J, x
The minimum completion time for this project is 28 weeks. The time taken to complete
activity J is labelled x. The maximum value of x is
A 12 B 10 C 8 D 4 E 2
18 A project consists of ten activities, A to J. The table below shows the immediate
predecessor(s) and earliest start time, in days, of each activity.
Immediate Earliest
Activity
predecessors starting time
A − 0
B − 0
C − 0
D A 6
E B 5
F B 5
G C 4
H D, E 13
I F, G 14
J H, I 25
It is known that activity H has a completion time of ten days. The project can still be
completed in minimum time if activity D is delayed. The maximum length of the delay
for activity D is
A one day B two days C seven days D eight days E nine days
14E Crashing
Learning intentions
I To be able to use crashing to reduce the completion time of a project.
I To be able to minimise the cost of crashing activities to achieve the maximum
reduction in completion time of a project.
Altering completion times

The minimum time it takes to complete a project depends upon the time it takes to complete
the individual activities of the project, and upon the predecessors each of the activities have.
Critical path analysis can be completed to find the overall minimum completion time.
Sometimes, the managers of a project might arrange for one or more activities within the
project to be completed in a shorter time than originally planned. Changing the conditions of
an activity within a project, and recalculating the minimum completion time for the project,
is called crashing.

14E Crashing 709
An individual activity could be crashed by employing more staff, sourcing alternate

materials or simply because weather or other factors allow the activity to be completed in a
shorter time than usual.
A simple crashing example

A simple activity network is shown in the 2 6 B, 4 6 10
diagram on the right. The forwards and A, 2 C, 3
A 0 4
backwards scanning processes have been
start finish
completed and the critical path has been
determined. The critical path is shown in D 0 0 13 13
D, 6 E, 7
red on the diagram.
6 6
The minimum time for completion is currently 13 hours. In order to reduce this overall
time, the manager of the project should try to complete one, or more, of the activities in a
shorter time than normal. Reducing the time taken to complete activity A, B or C would not
achieve this goal however. These activities are not on the critical path and so they already
have slack time. Reducing their completion time will not shorten the overall time taken to
complete the project.
Activity D and E, on the other hand, lie on the critical path. Reducing the duration of these
activities will reduce the overall time for the project. If activity D was reduced in time to
4 hours instead, the project will be completed in 11, not 13, hours.
Crashing with cost

Shortening the completion time for any individual activity could result in an extra cost
for the project. In the simple example above, the cost of reducing the completion time of
activity D by 1 hour is $150, while the cost of reducing the completion time of activity E by
1 hour is $18.
Clearly it is best to reduce the completion time, or crash, the activity that will cost the least.
Example 12 Crashing one activity with cost
The directed network below shows the sequence of 8 activities that are needed to
complete a project. The time, in days, that it takes to complete each activity is also
shown.
D, 9
A, 4 H, 1
B, 5 E, 5
Start Finish
G, 3
C, 6 F, 7
a Write down the critical path for this project.

b What is the minimum completion time for the project?
Activity F can be reduced by a maximum of 3 days at a cost of $100 per day.

c What is the new minimum completion time for the project?
d What is the minimum cost that will achieve the greatest reduction in time taken to
complete the project?
a In crashing problems, we first need Path Duration(days)
to identify the critical path, or paths.
A−D−H 14
We will do this by remembering that
B−E −G 13
a critical path is the longest or equal C −F −G 16
longest path in the activity network.
Using this method, set up a table, list The critical path is C − F − G
all possible paths from Start to Finish
for the directed network and calculate
the length of each path. The critical
path is the path with the longest time
from Start to Finish.
b Write the duration of the 16 days
critical path identified in the
previous part. Path Duration New duration with
(days) maximum reduction
c Crash all possible activities by
(F by 3)
the maximum reduction. Add
A−D−H 14 14
a new column to the summary
B−E −G 13 13
table to get an overview of the
C −F −G 16 13
new duration of each path.
The new minimum completion time for the project
This may result in a new
is 14 days.
critical path. Consider the cost
Activity F originally took 7 days to complete. It can
of crashing and whether it is
be crashed, which means activity F may be reduced
worth applying the maximum
by a maximum of 3 days, to result in a completion
reduction.
time of 4 days. It is possible to choose to reduce
acitvity F by 0, 1, 2 or 3 days. Reducing activity F
by the maximum 3 days would result in the original
critical path to be reduced from a total of 16 days,
down to 13 days. Considering there is a cost of $100
per day, this is not a desirable outcome; crashing
activity F by 3 days results in a new critical path
A − D − H with a total completion time of 14 days.
If we crash activity F by 2 days only, we create 2
equal critical paths requiring 14 days to complete
the project.
14E Crashing 711
d Reducing activity F by 2 days $200.

allows us to reduce the overall
completion time of the project
at minimum cost.
Example 13 Crashing multiple activities with cost
The directed network below shows the sequence of 9 activities that are needed to
shown.
C, 8
A, 9 E, 10
dummy
Start Finish
D, 8 F, 7 H, 3
B, 10 I, 6
G, 12
The minimum completion time for the project is 28 days. It is possible to reduce the
completion time for activities B, E, G, H and I. The completion time for each of these five
activities can be reduced by a maximum of two days.
a What is the new minimum completion time, in days, that the project could take?
The reduction in completion time for each of these five Activity Daily cost($)
activities will incur an additional cost. The table opposite
B 1500
shows the five activities that can have their completion
E 2000
times reduced and the associated daily cost, in dollars. G 700
b What is the minimum cost that will achieve the H 900
greatest reduction in time taken to complete the I 800
project?
a List all possible paths from Path Duration New duration after
Start to Finish, including the (days) maximum reduction
completion time of each. Crash (B,E,G,H,I by 2)
all activities by their maximum A−C −E 27 25
reduction. Identify the new A−F−H 19 17
critical path (path with the B−D−F −H 28 24
longest completion time) after B−G−I 28 22
the maximum reductions are
applied. A − C − E is the new critical path with a duration
of 25 days.

712 Chapter 14 Flow, matching and scheduling problems 14E
Explanation
b 1 Begin with the new critical path A − C − E. The reduction of activity E must occur
to achieve the new minimal completion time, therefore reducing activity E by 2
days is essential.
2 Ignore the path A − F − H because its completion time of 19 days is already lower
than the critical path.
3 Consider the path B − D − F − H. It has a completion time of 28 days and must be
reduced to 25 days to equal the critical path of A − C − E. There are two options;
reduce B by 2 and H by 1 or reduce B by 1 and H by 2. From the table, it is more
expensive per day to reduce B than H, however by choosing to reduce B this will
also reduce the completion time of the final path B − G − I, which is more cost
effective; reducing activity B reduces the completion time of two different paths.
So reduce activity B by 2 and H by 1 day to reduce the overall completion time of
B − D − F − H down to 25 days.
4 The final path B − G − I has already been reduced by 2 days due to the reduction
of activity B previously chosen. One more activity must be reduced for this path to
equal the critical path. Activity G has a lower cost of reduction than activity I per
day, so include this in your calculation.
5 Calculate your total cost of crashing.
Solution
b Cost of crashing = E by 2 days + B by 2 days + H by 1 day + G by 1 day
= 2000 × 2 + 1500 × 2 + 900 + 700
= $8600
Exercise 14E
1 The activity network for a project is shown in the diagram below. The duration for each
activity is in hours.
A, 7
Start D, 10
B, 3 E, 4
F, 13
Finish
C, 5
G, 6 I, 8
H, 3

14E 14E Crashing 713
a List all four paths from the Start to the Finish of the project, with their respective
completion times.
b Identify the critical path and the minimum completion time for the project.
c If Activity E is reduced by 3 hours, identify the new minimum completion time for
this project.
Example 11 2 The directed network below shows the sequence of 8 activities that are needed to
complete a project. The time, in days, that is takes to complete each activity is also
shown.
D, 4
B, 7 dummy G, 3
Start Finish
A, 5 C, 6 E, 8
F, 5 H, 4
a Write down the critical path for this project.

b What is the minimum completion time for the project?
Activity B can be reduced by a maximum of 3 days at a cost of $100 per day.
c What is the new minimum completion time for the project?
d What is the minimum cost that will achieve the greatest reduction in time taken to
3 The activity network D, 4

for a project is shown A, 3
E, 5 H, 8
in the diagram on the start
right. The duration B, 6 G, 3 finish
J, 3
for each activity is in F, 4 I, 7
C, 2
hours.
a Identify the critical path for this project.
b What is the maximum number of hours that the completion time for activity E can
be reduced by without changing the minimum completion time of the project?
c What is the maximum number of hours that the completion time for activity H can
be reduced without changing the minimum completion time of the project?
d Every activity can be reduced in duration by a maximum of 2 hours. If every
activity was reduced by the maximum amount possible, what is the new minimum
completion time for the project?

4 The activity network for a project is shown in the diagram below. The duration for each
activity is in hours.
D, 2
A, 4 E, 9 K, 11
B, 3 F, 13 L, 1
Start Finish
C, 6 I, 2
J, 5
G, 8
H, 4
a How many activities could be delayed by 4 hours without altering the minimum
completion time for the project?
b If the project is to be crashed by reducing the completion time of one activity only,
what is the minimum time, in hours, that the project can be completed in?
c Activity G can be reduced in time at a cost of $200 per hour. Activity J can
be reduced in time at a cost of $150 per hour. What is the cost of reducing the
completion time of this project as much as possible?
Example 13 5 The directed network below shows the sequence of 8 activities that are needed to
shown.
A, 7 D, 5
G, 5
B, 11 F, 6
Start Finish
H, 10
C, 5 E, 8
The minimum completion time for the project is 24 days. It is possible to reduce the
completion time for activities D, E and H. The completion time for each of these three
activities can be reduced by a maximum of two days.
a What is the new minimum completion time, in days, of the project?
The reduction in completion time for each of these Activity Daily cost($)
three activities will incur an additional cost. The table
D 170
opposite shows the three activities that can have their
E 350
completion times reduced and the associated daily H 200
cost, in dollars.
b What is the minimum cost that will achieve the greatest reduction in time taken to

6 The directed network below shows the sequence of 11 activities that are needed to
shown.
C, 5 F, 3
A, 7 dummy
I, 2 J, 6
Start D, 4
Finish
G, 3 K, 7
B, 2 H, 5
E, 6
a Which activities are immediate predecessors to activity G?

b Which activities, if crashed, would create more than one critical path?
c The project could finish earlier if some activities were crashed. Five activities,
B, E, G, H and I, can all be reduced by one hour. The cost of this crashing is $150
per hour.
i What is the minimum number of hours in which the project could now be
completed?
ii What is the minimum cost of completing the project in this time?
7 The directed network below shows the sequence of 12 activities that are needed to
complete a project. The time, in weeks, that it takes to complete each activity is also
shown.
B, 4 D, 5
A, 6
dummy I, 4
Start G, 9
C, 2 E, 6 Finish
J, 11
L, 1
F, 4 H, 3
K, 13
a Determine the shortest time in which this project can be completed.

b Determine the earliest start time for activity D.
c Determine the latest start time for activity H.
d Which activity has a float time of more than two weeks?

e The completion times for activities D, E, G, H Activity Weekly cost($)

and J can each be reduced by a maximum of two
D 2000
weeks. The table opposite shows the five activities
E 1000
than can have their completion time reduced and G 500
the associated weekly cost, in dollars. H 1500
What is the minimum cost to complete the project J 3000
in the shortest time possible?
8 The directed network below shows the sequence of 15 activities that are needed
to complete a maintenance project at the MCG. The time, in days, that it takes to
complete each activity is also shown.
I, 6
H, 7 N, 6
A, 5 C, 3 J, 3
G, 4 M, 4
Start Finish
K, 9 O, 5
D, 7 F, 2
B, 2 L, 11
E, 8
a What is the minimum completion time?

b How many activities are on the critical path?
c How many paths have a completion time of 28 days?
d The completion times for activities H, J, K, L and M can each be reduced by a
maximum of two days. The cost of reducing the time of each activity is $500 per
day. The MCG requires the overall completion time for the maintenance project to
be reduced by three days at minimum cost. Complete the table below, showing the
reductions in individual activity completion times that would achieve this.
Activity Reduction in
completion time
(0, 1 or 2 days)
H
J
K
L
M


Questions 9 and 10 refer to the diagram E, 10
opposite. A, 3 I, 4
C, 9 G, 7
The directed graph opposite shows the sequence Start Finish
of activities required to complete a project. All F, 3 J, 4
B, 6
times are in hours.
D, 2 H, 5
9 There is one critical path for this project.

The critical path is
A A−E−I B A−C −G
C B−D−F −G D B−D−H−J
E A−C −F −H − J
10 Four critical paths would exist if the duration of activities

A A and B were reduced by one hour. B C and G were reduced by one hour.
C A and C were reduced by two hours. D B and G were reduced by two hours.
E D and H were reduced by three hours.
11 The directed graph below shows the sequence of activities required to complete a
project. All times are in weeks. There is one critical path for this project.
C, 6 H, 7
A, 4
L, 4
Start G, 9
I, 8
D, 6 Finish
F, 7 M, 2
B, 5
J, 6 K, 10
E, 9
The total completion time of the project can be reduced by four weeks by reducing
A activity B by four weeks
B activity F by four weeks.
C activity J by four weeks.
D activity I by three weeks and activity J by one week.
E activity D by three weeks and activity E by one week.

Review 718 Chapter 14 Flow, matching and scheduling problems
Weighted graph A weighted graph is a graph in which a number representing the size of
Assign- same quantity is associated with each edge. These numbers are called
ment
weights.
Network A network is a weighted graph in which the weights are physical

quantities, for example distance, time or cost.
Directed graph A directed graph is a graph where direction is indicated for every edge.
(digraph) This is often abbreviated to digraph.
Flow The transfer of material through a directed network. Flow can refer to
the movement of water or traffic.
Capacity The maximum flow of substance that an edge of a directed graph can
allow during a particular time interval. The capacity of water pipes is
the amount of water (usually in litres) that the pipe will allow through
per time period (minutes, hours, etc.). Other examples of capacity are
number of cars per minute or number of people per hour.
Source The source is the origin of the material flowing through a network.
Sink The sink is the final destination of the material flowing through a
network.
Cut A cut is a line dividing a directed graph into A

two parts (shown as a broken line dividing
3 6
the graph below into two sections, labelled X X 2 Y
and Y). C D
Cut capacity The sum of the capacities (weights) of the 5 2
edges directed from X to Y B
that the cut passes through. For the weighted digraph shown, the
capacity of the cut is 7.
Minimum cut The minimum cut is the cut with the minimum capacity. The cut must
separate the source from the sink.
Maximum flow The maximum flow through a directed graph is equal to the capacity of
the minimum cut.
Bipartite graph A bipartite graph has two distinct groups or categories for the vertices.
Connections exist between a vertex or vertices from one group with
a vertex or vertices from the other group. There are no connections
between the vertices within a group.

Review
Allocation An allocation is made when each of the vertices in one group from a
bipartite graph are matched with one of the vertices in the other group
from that graph. An allocation is possible when both groups have
exactly the same number of vertices. The vertices in each group are
matched to only one vertex from the other group.
Cost matrix A table that contains the costs of allocating objects from one group
(such as people) to another (such as tasks). The ‘cost’ can be money, or
other factors such as the time taken.
Hungarian The Hungarian algorithm is an algorithm that is used to determine the

algorithm best allocation to minimise the overall cost.
Activity network An activity network is a directed graph that shows the required order of
completing individual activities that make up a project.
Immediate If activity A is an immediate predecessor to activity B, activity A must

predecessor be completed before activity B can begin.
Precedence table A precedence table is a table that records the activities of a project and
their immediate predecessors. Precedence tables can also contain the
duration of each activity.
Dummy activity A dummy activity has zero cost. It is required if two activities share
some, but not all, of the same immediate predecessors. It allows the
network to show all precedence relationships in a project correctly.
Earliest starting EST is the earliest time an activity in a project can begin.
time (EST)
Latest starting LST is the latest time an activity in a project can begin, without
time (LST) affecting the overall completion time for the project.
Float (slack) time Float (slack) time is the difference between the latest starting time and
the earliest starting time.
Float = LST − EST
The float time is sometimes called the slack time. It is the largest
amount of time that an activity can be delayed without affecting the
overall completion time for the project.
Forward Forward scanning is a process of determining the EST for each activity
scanning in an activity network. The EST of an activity is added to the duration
of that activity to determine the EST of the next activity. The EST of
any activity is equal to the largest forward scanning value determined
from all immediate predecessors.

Backward Backward scanning is a process of determining the LST for each

scanning activity in an activity network. The LST of an activity is equal to the
LST of the activity that follows, minus the duration of the activity.
Critical path The critical path is the series of activities that cannot be delayed
without affecting the overall completion time of the project. Activities
on the critical path have no slack time. Their EST and LST are equal.
Critical path Critical path analysis is a project planning method in which activity
analysis durations are known with certainty.
Crashing Crashing is the process of shortening the length of time in which a

project can be completed by reducing the time required to complete
individual activities. Reducing the individual activity completion times
often costs money; this increases the overall cost of a project.
Skills checklist
list
14A 1 I can define and describe a directed graph.
14A 2 I can determine the maximum flow for any section of sequential edges of a
directed graph.
14A 3 I can determine cut capacities.
14A 4 I can determine the maximum flow as equal to the minimum cut capacity.
14B 5 I can draw directed and weighted bipartite graphs.
14B 6 I can use the Hungarian algorithm to determine an optimum allocation in order
to minimise cost.

Review
14C 7 I can create an activity network from a precedence table.
14C 8 I can write down a precedence table from an activity network.
See Exercise 14C Question 1
14C 9 I can decide when to use dummy activities in an activity network.
14D 10 I can use forward scanning to determine the earliest starting time of activities
in an activity network.
14D 11 I can use backward scanning to determine the latest starting time of activities
in an activity network.
14D 12 I can determine the float time for activities in an activity network.
14D 13 I can determine the overall minimum completion time for a project using
critical path analysis.
14D 14 I can determine the critical path for an activity network.
14E 15 I can use crashing to reduce the completion time of a project.

1 The shortest path from A to Z in the 9 E
B
network on the right has length:
5 7
10 F
A 10 B 15 7 11
A D 4 Z
C 22 D 26 7
12 6 3 10
E 28 8
G 8 H
C 10
2 For the network shown on the right, the capacity of 7

9
the cut is: 2 3 cut
5 4
A 3 B 6 C 9 3
4
D 10 E 14 5 6
3 In the communications network shown, the 10

14 5
numbers represent transmission capacities for
P 4 4 7
information (data) in scaled units. What is the
Q
maximum flow of information from station P 10 3
to station Q? 5 12
A 20 B 22 8
C 23 D 24
E 30
4 The maximum flow in the network 8

4
opposite, from source to sink, is: 7
6
A 10 B 11 6
C 12 D 13
6
E 14 5
7
5 A group of five students represent their school Travis basketball

in five different sports. The information is
displayed in a bipartite graph. From this Fulvia swimming
graph we can conclude that: Miriam athletics

A Travis and Miriam played all the sports
Kieren volleyball
between them.
B In total, Miriam and Fulvia played fewer Andrew tennis
sports than Andrew and Travis.
C Kieren and Miriam each played the same number of sports.
D In total, Kieren and Travis played fewer different sports than Miriam and Fulvia.
E Andrew played fewer sports than any of the others.

Review
6 This activity network is for a project E, 3
B, 4 I, 3
where the component times in days A, 5 D, 6 H, 3
are shown. The critical path for the G, 2 K, 6
C, 3 J, 1
network of this project is given by: F, 6
A A–B–E–I–K
B A–D–H–I–K
C A–C–G–H–I–K
D A–C–F–J–K
E A–D–G–F–J–K
7 The activity network shown represents a C, 5

F, 5
A, 9 finish
project development with activities and their D, 7 G, 6
start dummy
durations (in days) listed on the edges of the
B, 7 E, 8
graph. Note that the dummy activity takes
zero time.
The earliest time (in days) that activity F can begin is:
A 0 B 12 C 14 D 22 E 24
8 The table opposite lists the seven activities in a

Immediate
project and the earliest start time, in hours, and
Task predecessor EST
the predecessor(s) of each task. The time taken
A – 0
for activity F is five hours. Without affecting
the time taken for the entire project, the time B – 0
taken for activity D could be increased by: C A 24
A 0 hours B 2 hours D B 29
C 3 hours D 4 hours E C 39
E 12 hours F D 41
G E, F 50
9 The edges in this activity network B, 6

correspond to the tasks involved in A, 3 E, 2
the preparation of an examination.
D, 5
The numbers indicate the time, in H, 6
weeks, needed for each task. The C, 2 G, 4
total number of weeks needed for the F, 3
preparation of the examination is:
A 14 B 15 C 16 D 17 E 27

10 The activity network represents a manufactur- A, 2 E, 1

ing process with activities and their duration start G, 1
C, 4 F, 2
(in hours) listed on the edges of the graph. finish
B, 3
The earliest time (in hours) after the start that H, 4
D, 2
activity G can begin is:
A 3 B 5 C 6 D 7 E 8
Written-response questions
1 An English class recently performed poorly in their essay writing assessment. To help
them improve, the teacher separated the class into groups of five and assigned one of
the following tasks to each person: Introduction, Body paragraph 1, Body paragraph
2, Body paragraph 3 and Conclusion. Each task will be completed by one person. The
table below shows the time, in minutes, that each person would take to complete each
of the five tasks.
Intro Par 1 Par 2 Par 3 Con

Alvin 16 14 19 9 9
Billy 17 18 10 9 9
Chloe 9 8 6 15 8
Danielle 11 12 11 16 6
Elena 10 10 8 15 8
The tasks will be allocated so that the total time of completing the five tasks is a
minimum
a Complete the sentences below by clearly stating which task each student should
write in order for the essay to be completed in the minimum time possible.
Alvin should write the...
Billy should write the...
Chloe should write the...
Danielle should write the...
Elena should write the...
b What is the minimum total time the group will dedicate to completing the essay?

Review
2 WestAir Company flies routes in Mildura 2 Echuca
western Victoria. The network 7 1
3 4
shows the layout of connecting Ballarat Melbourne
8
flight paths for WestAir, which Horsham 2 7 10
2
2 1
originate in Mildura and terminate 3 Geelong
Hamilton
in either Melbourne or on the way to 2
Melbourne. On this network, Warmambool
the available spaces for passengers flying out of various locations on one morning
are shown.
The network has one cut shown.
a What is the capacity of this cut?
b What is the maximum number of passengers who could travel from Mildura to
Melbourne for the morning?
3 A school swimming team wants to select a 4 × 200 metre relay team. The fastest times
of its four best swimmers in each of the strokes are shown in the table below. Which
swimmer should swim which stroke to give the team the best chance of winning, and
what would be their time to swim the relay?
Swimmer Backstroke Breaststroke Butterfly Freestyle

Rob 76 78 70 62
Joel 74 80 66 62
Henk 72 76 68 58
Sav 78 80 66 60
4 In the network opposite, the values 3 cut

on the edges give the maximum 10 8
7
flow possible between each pair source 6 9 2 2
5 8 sink
of vertices. The arrows show the 5 9
11
direction of flow in the network. 13
15
Also shown is a cut that separates the 14
source from the sink.

a Determine the capacity of the cut shown.
b Determine the maximum flow through this network.

5 A project requires eight activities (A − H)

Duration Immediate
to be completed. The duration, in hours,
Activity (hours) predecessor(s)
and the immediate predecessor(s) of each
A 4 −
activity are shown in the table below.
B 5 −
C 6 A
D 7 A, B
E 10 C, D
F 5 D
G 4 F
H 5 E, G
a The directed network that shows these activities is shown below. Add the three
missing features to the network.
C, 6
A, 4
H, 5
Start dummy Finish
B, 5 G, 4
F, 5
b Determine the earliest start time for activity E.

c What is the float time of activity G?
d How many of these activities have a non-zero float time?
e Write down the critical path for this project.
f What is the minimum completion time for this project?
g The project could finish earlier if some activities were crashed.
i Activity E can be crashed by two hours. If this occurs, what will be the new
critical path for the project?
ii In addition to activity E crashing by two hours, activities, A, B and D can also be
crashed by one hour each. What is the minimum number of days in which the
project can now be completed?
h After careful deliberation, it is decided that crashing the original directed network is
not possible. Alternatively, an upgrade will be made to this project, where one extra
activity will be added. This activity has a duration of five hours, an earliest starting
time of twelve hours and a latest starting time of seventeen hours. Complete the
following sentence by filling in the boxes provided.
The extra activity could be represented on the network above by a directed edge
from the end of activity to the start of activity

Chapter
15
Revision
Revision: Networks and
decision mathematics

Revision 728 Chapter 15 Revision: Networks and decision mathematics
15A Exam 1 style questions

1 The diagram opposite shows a map of the roads G
between four towns: F, G, H and I.
A graph that represents the connections between the towns on the map is:
A B C
G G G
F F F
I H I H I H
D F G E F G
I H I H
2 Five people are to be each allocated Name A B C D E

one of five tasks (A, B, C, D, E).
Francis 12 15 99 10 14
The table shows the time, in
hours, that each person takes to David 10 9 10 7 12
complete the tasks. The tasks must Herman 99 10 11 6 12
be completed in the least possible Indira 8 8 12 9 99
total amount of time. If no person Natalie 8 99 9 8 11
can help another, Francis should be
allocated task:
A A B B C C D D E E
3 The sum of the degrees of all the vertices in

the graph opposite is:
A 6 B 7 C 8
D 15 E 16
4 Adding which one of the following edges to the T U

graph opposite makes an Eulerian trail possible?
Z
A ST B SU C SX S V
D XW E ZY
Y X W

Revision
5 A connected planar graph has an equal number of vertices and faces. If there are
20 edges in this graph, the number of vertices must be:
A 9 B 10 C 11 D 20 E 22
6 Underground water pipes are needed to water a new golf course. Water will be pumped
from the dam in the back corner of the course. To find the smallest total length of water
pipe needed, we must find:
A a critical path B a minimal spanning tree
C the shortest Eulerian circuit D the shortest Hamiltonian cycle
E the perimeter of the golf course
7 Which one of the following is a true statement about a critical path in a project?
A Knowledge of the critical path can be used to decide if any tasks in a project can be
delayed without extending the length of time of the project.
B All tasks on the critical path must be completed before any other task in the same
project can be started.
C Decreasing the times of tasks not on the critical path will decrease the length of time
of the project.
D The critical path must always include at least two tasks in a project.
E There is only one critical path in any project.
8 The length of the shortest path between the P
5 9
origin, O, and destination, D, in the weighted graph 5 U
3 S 4
shown here is: O 2
Q 6 5
A 11 B 12 C 13 1
3 6 T
2 5
D 14 E 15 7
4
R
10
D
9 Four students, talking about five ski Ann Falls Creek
resorts they have visited, represented their
information on the bipartite graph shown here. Matt Val D’Isere
Which one of the following statements is

Tom Zermatt
implied by this bipartite graph?
A Ann and Maria between them have visited Maria Aspen
fewer ski resorts than Matt and Tom
between them. Mt Hutt
B Matt and Tom have been to four ski resorts between them.
C Maria has visited fewer ski resorts than any of the others.
D Ann and Maria between them have visited all five ski resorts discussed.
E Ann and Tom between them have visited fewer resorts than Matt and Maria between
them.

10 A gas pipeline is to be constructed to link several towns in the country. Assuming the
pipeline construction costs are the same everywhere in the region, the cheapest network
formed by the pipelines and the towns as vertices would form:
A a Hamiltonian cycle B an Eulerian circuit C a minimum spanning tree
D a critical path E a complete graph
11 Which one of the following Sally Kate Jon Greg
statements is not implied by this
bipartite graph?
A There are more translators of
French than Greek. Spanish Italian Greek Turkish French
B Sally and Kate can translate five
languages between them.
C Jon and Greg can translate four languages between them.
D Kate and Jon can translate more languages between them than can Sally and Greg.
E Sally and Jon can translate more languages between them than can Kate and Greg.
12 There are four different human blood types: O, A, B and AB. The relationships between
donor and recipients for these blood types are as follows:
Type O can donate blood to any type.
Type AB can receive blood from any type.
Each type can donate blood to its own type.
Each type can receive blood from its own type.
Which one of the following donor–recipient bipartite graphs correctly represents this
information?
A O A B AB B O A B AB
donor donor
recipient recipient
O A B AB O A B AB
C O A B AB D O A B AB
donor donor
recipient recipient
O A B AB O A B AB
E O A B AB
donor
recipient
O A B AB

Revision
13 For the weighted graph shown, the length (total 8
8
weight) of the minimum spanning tree is: 7
4
A 28 B 29 C 30 3 6
D 31 E 32 5
9
4
14 A connected graph with 12 edges divides a plane into four faces. The number of
vertices in this graph will be:
A 6 B 10 C 12 D 13 E 14
15 The number of edges for a complete graph with twenty vertices is:
A 10 B 20 C 21 D 180 E 190
16 The number of vertices for a complete graph with twenty-one edges is:
A 7 B 8 C 14 D 42 E 43
17 The number of edges for a tree with four vertices must be:
A 1 B 2 C 3 D 4 E 5
18 The number of vertices for a tree with 13 edges must be:

A 6 B 7 C 14 D 15 E 26
19 What additional edge could be added to the graph B

opposite to ensure that the resulting graph would
contain an Eulerian circuit?
A AB B AC C AD A C
D AE E BC D
E
20 The graph opposite represents a project with C D
activities listed on the edges of the graph. A I
E J
Which of the following statements must be
true? F
H K
B
A A must be completed before B can start.
B A must be completed before F can start. G
C E and F must start at the same time.
D E and F must finish at the same time.
E E cannot start until A is finished.

21 An adjacency matrix for the graph opposite could be:
     
0 0 1 0 0 0 1 0 0 0 1 0
     
0 0 1 0 0 0 1 0 0 0 1 0
A  B  C 
    
1 1 2 0 1 1 1 0 1 1 3 0
     
0 0 0 0 0 0 0 0 0 0 0 0
  
   
0 0 1 0 0 0 1 0
   
0 0 1 0 0 1 1 0
D  E 
  
1 1 0 0 1 1 2 0
   
0 0 1 0 0 0 0 0
 
22 A vehicle is travelling from town P to town 90

160
Q. The journey requires the vehicle to travel 80 60
along a network linking suitable fuel stops. P 135 110
70 100 85
The cost, in dollars, of travel between these 120 Q
205 120
is shown on the network opposite, where 95
the vertices represent fuel stops.
What is the minimum cost, in dollars, for the trip?

A 400 B 405 C 410 D 420 E 440
23 The capacity of the cut in the network flow 9 cut

3
diagram shown is: 4
2
A 0 B 2 C 10 8 6
D 13 E 16 3
5 3
24 The sum of the degrees of the vertices on the graph

shown is:
A 12 B 13 C 14
D 15 E 16
25 A connected planar graph divides the plane into a number of faces. If the graph has
nine vertices and these are linked by 20 edges, then the number of faces is:
A 11 B 13 C 21 D 27 E 31

Revision
26 The sum of the weights of the minimum spanning tree of
the weighted graph is: 6 4
7
A 2 B 30 C 32 8
D 33 E 35 5
7
3
The following graph relates to Questions 27 and 28.
B
1
7 D
A 5
2
3
F
4
C 3
6 E
27 The maximum flow in the network linking vertex A to vertex F is:

A 5 B 6 C 7 D 8 E 9
28 The number of ways that vertex F can be reached from vertex A is:
A 1 B 2 C 3 D 4 E 5
15B Exam 2 style questions

a What is the sum of the degrees of all the vertices
in the graph?
b What is the maximum number of edges that can
be removed so that the graph remains connected?
c What is the maximum number of edges that can
be removed for an Eulerian trail to exist?
d Circle the bridge in the graph.
e Verify Euler’s formula for the graph opposite.
2 The adjacency matrix and graph below represent the same information. Some elements
are missing from the adjacency matrix and some edges from the graph. Write down the
missing elements in the matrix and add the missing edges to the graph.
Q
 P Q R S
P P 
 1 0 1 
R
 
Q  1 0 1 
 
R  0 2 0 1 
 
S  1 1 1 0 
S

3 Consider the diagram Source

A
opposite, which represents 8
the number of cars per 15 D
9
hour over a suburban road 4
5 Sink
system. T
S B
a Give a reason why Cut
1 is not a valid cut.
14
b What is the capacity of 8
10
Cut 2?
c What is the maximum Cut 1
Cut 2 C
flow for this network?
4 A project requires nine activities (A − J) to be completed. The duration, in hours, and

the immediate predecessor(s) of each activity are shown in the table below.

A −
B −
C A
D A
E B
F B
G C
H C
I D, E, G
J H, I, F
The directed network that shows these activities is shown below. Three features are
missing. Using the table above, complete the network diagram below.
A C J
Start Finish
D I
F
E

Revision
5 Consider the activity network below.
A, 6
C, 2
Start
E, 3
G, 6
B, 4
Finish
D, 2
H, 4
F, 4
a Determine the earliest start time of activity G.

b Determine the latest start time of activity F.
c Determine the float time for activity C.
d Write down the critical path.
e Determine the minimum completion time for this project.
6 All the activities and their durations (in hours) in a project at a quarry are shown in the
network diagram below. The least time required for completing this entire project is 30
hours.
G, 4
E, 4 T, 0 J, 3
A, 6 F, 6 I, 2 K,
start finish
B, 5 C, 2 H, 3
D,
For each activity in this project, the table on the next page shows the completion time,
the earliest starting time and the latest starting time.

Completion Earliest starting Latest starting

Activity time (hours) time (hours) time (hours)
A 6 0
B 5 0 0
C 2 5 5
D 5 9
E 4 7 7
F 6 7
G 4 11 11
H 3 9 13
I 2 13 16
J 3 15 15
K 18 18
a Complete the missing times in the table above.

b Write down the critical path for this project.
7 A rural town, built on hills, contains a Q

0.8 0.9
set of roads represented by edges in the S
network shown here. The numbers on P 0.7
0.7 0.8 0.6
the network refer to distances along the
roads (in kilometres) and the letters refer R 0.8 T 1.0
to intersections of the roads. The edges
0.6
without endpoints refer to the two roads in
U
and out of town.
a i What is the length of the shortest route through the town from P to U?
ii A safety officer who enters the town at P needs to examine all intersections in
the town before leaving from U to travel to the next town. To save time, she
wants to pass through each intersection only once. State a path through the
network of roads that would enable her to do this.
b A technician from the electricity company is checking the overhead cables along
each street. The technician elects to follow an Eulerian path through the network
streets (ignoring the roads in and out of town) starting at R and finishing at S .
i Complete the following Eulerian trail:
R−Q−P−R− − − −T −U−S
ii How would the technician benefit from choosing an Eulerian path?

Revision
8 Seven towns on an island have been surveyed 19 B
20
for transport and communications needs. D 56
A
The towns (labelled A, B, C, D, E, F, G) form E
32
the network shown here. The road distances 29 21
18
between the towns are marked in kilometres. 33
C 28
(All towns may be treated as points being of
F
no size compared to the network lengths.) 25 16
G
a Explain what is meant by the description of the graph as ‘planar’.
b Verify Euler’s formula for the graph above.
An inspector of roads is stationed at B. Starting from B, she must travel the complete
network of roads to examine them.
c If she wishes to travel the least distance where will she end up in the network?
d What will that distance be?
e Is the route unique? Briefly justify your answer.
f Determine the shortest distance that a fire truck stationed at E must travel to assist at
an emergency at A.
g To establish a cable network for 19 B
20
telecommunications on the island, it is D 56
A
proposed to put the cable underground E
32
beside the existing roads. What is the 29 21
18
minimal length of cable required here 33
C 28
if back-up links are not considered
F
necessary; that is, there are no loops in the 25 16
cable network? G
The Island Bank has outlets in each of the towns. The regional assistant manager
stationed at C must visit each outlet every second Friday and then return to the
office at C.
h Treating the towns as vertices and roads as edges in a graph, what is the distance of
a journey that forms a Hamiltonian cycle in the graph?
9 The assembly of machined parts in a manufacturing process can be represented by the

following network. The activities are represented by the letters on the edges and the
numbers represent the time taken (in hours) for the activities scheduled.
C, 2 G, 4
A, 2 D, 2 J, 4
F, 1 H, 8
Start Finish
B, 1 E, 6 I, 5 K, 2

Activity A B C D E F G H I J K
EST 0 0 2 2 4 4 10 10 18 22
a The earliest start times (EST) for each activity except G are given in the table.
Complete the table by finding the EST for G.
b What is the shortest time required to assemble the product?
c What is the float (slack time) for activity I?
10 A reservoir at E pumps water through pipes B

along the network routes shown. The capacities 5 5
A D 8
of the flow are given in the digraph shown here E
2
in megalitres per day. 5
6 4
Occasionally, there are fire emergencies in the 4
forest beside A and additional flow of water C F

6
is used. What is the maximum flow that can
2 3
reach A from E?
G
11 A number of towns need to be linked by M
38 N
pipelines to a natural gas supply. In the 51 31 40
network shown, the existing road links 43
L S 72
between towns L, M, N, O, P, Q and R and 24
50 65
to the supply point, S , are shown as edges. 35 R
O
The towns and the gas supply are shown 47
57 63
as vertices. The distances along roads are P
given in kilometres. 55
Q
a What is the shortest distance along roads from the gas supply point S to the
town O?
b The gas company decides to run the gas lines along the existing roads. To ensure
that all nodes on the network are linked, the company does not need to place pipes
along all the roads in the network.
i What is the usual name given to the network within a graph (here, the road
system) which links all nodes (towns and supply) and which gives the shortest
total length?
ii Sketch this network.
iii What is the minimum length of gas pipeline the company can use to supply all
the towns by running the pipes along the existing roads?
c The gas company decides it wants to run the pipeline directly to any town which is
linked by road to its supply at S . Towns not directly connected to S by road will be
linked via other towns in the network.
What is the minimum length of pipeline that will enable all towns to be connected to
the gas supply under these circumstances?
Revision
12 In laying a pipeline, the various jobs involved have been grouped into a set of specific
tasks A−K, which are performed in the precedence described in the network below.
D J
F I
A
Start Finish
E H K
B C
a List all the task(s) that must be completed before task E is started. The durations of
the tasks are given in Table 1.
b Use the information in Table 1 to complete Table 2.
Table 1 Task durations Table 2 Starting times for tasks
Normal completion Task EST LST
Task time (months) A 0 0
A 10 B 0
B 6 C 6 7
C 3 D 10 10
D 4 E 11
E 7 F 14 14
F 4 G 14 18
G 5 H 18 20
H 4 I 18
I 5 J 23 23
J 4 K 22 24
K 3
c For this project:

i write down the critical path
ii determine the length of the critical path (that is, the earliest time the project can
be completed).
d If the project managers are prepared to pay more for additional labour and
machinery, the time taken to complete task A can be reduced to 8 months, task E
can be reduced to 5 months and task I can be reduced to 4 months.
Under these circumstances:
i what would be the critical path(s)?
ii how long would it take to complete the project?

13 Camp sites A, B, C and D are to be Camp site W X Y Z

supplied with food. Four local residents,
A 30 70 60 20
W, X, Y and Z, offer to supply one
campsite each. The cost in dollars of B 40 30 50 80
supplying one load of food from each C 50 40 60 50
resident to each campsite is tabulated. D 60 70 30 70
a Find the two possible matchings
between campsites and residents so
that the total cost is a minimum.
b State this minimum cost.
14 Consider the activity network below.

D, 3
B, 3
dummy G, 3
E, 4
A, 1 I, 4
Start Finish
H, 4
dummy J, 3
C, 5
F, 5
By employing more workers it is possible to reduce the time of some activities,

however this will incur extra costs. The activities which can be reduced in time, the
associated costs and maximum reduction in time are shown in the table below.
Activity Cost Maximum reduction

(dollars per week) (weeks)
E 1000 2
F 1500 3
H 2000 3
J 200 2
a What is the new minimum completion time now possible for the project?
b What is the minimum cost of completing the project in this time?
c How many activities will be reduced in time to achieve the new minimum
completion time at minimal cost?

Chapter
16
Revision
Revision of Chapters
1–15

Revision 742 Chapter 16 Revision of Chapters 1–15
16A Exam 1 questions

Data analysis, probability and statistics
1 In a study of cats, data relating to the following five variables were collected:
name
breed
age (1 = less than 1 year, 2 = 1–5 years, 3 = more than 5 years)
weight in kg
length in cm
The number of these variables that are discrete numerical variables is:
A 0 B 1 C 2 D 3 E 4

The invoices from Janelle’s 80
company request payment 70
60
from her customers within 14
Frequency
40
days. She records the number 40
of days between the customers 30
receiving the invoice, and the 20
10
invoice being paid (day paid),
0
for 105 invoices. Her data is 0 5 10 15 20 25 30
shown in the histogram. Day paid
2 Janelle decides to consider invoices which were paid in 15 days or more as late
payments. The percentage of the invoices classified as late payments is closest to:
A 9.5% B 10.0% C 19.0% D 20.0% E 85.7%
3 The third quartile (Q3 ) for this distribution could be:

A 8 days B 14 days C 15 days D 20 days E 22 days
4 A class of 28 students in Year 11 (16 girls and 12 boys) sat for a French test. The mean
score on the test for the 16 girls in the class was 42. The mean score on the test for the
12 boys in the class was 28. The mean score on the test for the whole class was
A 30 B 32 C 35 D 36 E 37
5 The amount of flour in a box, in grams, is approximately normally distributed with a

mean of 1002 g and a standard deviation of 4 g. The percentage of these boxes which
contain from 998 g to 1010 g of flour is closest to
A 32% B 68% C 81.5% D 95% E 97.35%

Revision
Use the following information to answer questions 6–8
The kilometres driven by secondhand cars were mean 36 947

recorded to the nearest 100 km for a sample of 300
minimum 500
cars. The mean and the five number summary for
these data are shown in the table opposite. first quartile (Q1 ) 15 000
median (M) 32 000
third quartile (Q3 ) 48 500
maximum 213 000
6 The difference between the mean kilometres driven and the median kilometres driven,
to the nearest 100 km, is closest to:
A 500 km B 4900 km C 5000 km
D 16 500 km E 33 500 km
7 Of these 300 cars, the number that have been driven less than 15 000 km is closest to:
A 50 B 75 C 100 D 125 E 150
8 The shape of the distribution of the the kilometres driven is best described as
A approximately symmetric
B positively skewed
C positively skewed with one or more outliers
D negatively skewed
E negatively skewed with one or more outliers
9 Suppose that the weights of adults wombats are approximately normally distributed. If
16% of wombats weigh less than 3.5 kg, and 0.15% of wombats weigh more than 5.5
kg, the mean and standard deviation of the weight of wombats, in kg, is closest to
A mean = 4.0, standard deviation = 0.5 B mean = 4.3, standard deviation = 0.4
C mean = 5.5, standard deviation = 2.0 D mean = 4.5, standard deviation = 1.0
E mean = 4.5, standard deviation = 0.5
The data in the following table Year level

were collected when a group of 250
Preferred camp Year 9 Year 10 Total
students from Year 9 and Year 10
were asked where they would like Beach 78 43 121
to go for their school camp. Snow 40 66 106
Other 12 11 23
Total 130 120 250

10 The percentage of students surveyed who chose to go to the beach is closest to:
A 48.4% B 60.0% C 35.8% D 65.0% E 35.5%
11 The data in the table supports the contention that there is an association between
preferred camp and year level because:
A more students preferred to go the beach than go to the snow.
B 35.8% of students in Year 10 preferred to go to the beach, compared to only 55.0%
of Year 10 students who preferred to go the snow.
C 60.0% of students in Year 9 preferred to go to the beach, compared to only 30.8% of
Year 9 students who preferred to go the snow.
D 48.2% of students preferred the beach compared to 42.4% who preferred the snow.
E 60.0% of students in Year 9 preferred to go to the beach, compared to only 35.8% of
Year 10 students who preferred to go the beach.
Use following information to answer Questions 12 and 13

In a study of the effect of sunlight on the growth of seedlings, a number of seedlings were
planted in various locations in a garden bed. The average number of hours of sunlight each
plant received each day for fourteen days (sunlight hours), as well as the amount they grew,
in cm, over that period of time (growth) were recorded. Growth can be predicted from
sunlight hours from the regression line:
growth = 4.28 + 0.586 × sunlight hours, with r = 0.690
12 The percentage of variation in growth NOT explained by the variation in the sunlight
hours is closest to:
A 16.9% B 31.0% C 34.3% D 47.6% E 52.4%
13 This regression line predicts that, on average, growth:

A increases by 4.28 cm for each additional hour of sunlight
B increases by 0.586 cm for each additional hour of sunlight
C decreases by 0.586 cm for each additional hour of sunlight
D decreases by 4.28 cm for each additional hour of sunlight
E increases by 0.690 cm for each additional hour of sunlight

Revision
14 The scatterplot shows the reaction 30
time, measured in hundredths of 28
seconds, for a group of 20 people, 26
together with their age. A least 24
reaction time
22
squares line had been fitted to
20
the scatterplot with age as the
18
explanatory variable, and reaction
16
time as the response variable. The 14
equation of the least squares line is 12
closest to: 10
16 18 20 22 24 26 28 30 32 34 36 38 40
age
A reaction time = 26 + 2.0 × age B reaction time = 4.0 + 0.5 × age

C reaction time = 12.0 + 0.5 × age D reaction time = 12.0 + 0.3 × age
E reaction time = 4.0 − 0.5 × age
15 The table below shows the life expectancy in years and the percentage of government
expenditure which is spent on health (health) in 8 countries.
Health (%) 17.3 10.3 4.7 6.0 20.1 6.4 13.2 7.7
Life expectancy (years) 82 76 68 69 83 75 76 76
A least squares line which enables a country’s life expectancy to be predicted from
their expenditure on health is fitted to the data. The value of the residual (to the nearest
year) when the actual percentage of government expenditure which is spent on health is
6% is closest to:
A -3 B -2 C 1 D 2 E 3
16 The price of shares in a company has increased non-linearly in the last 12 months. A
log transformation was applied to the maximum share price each month (share price),
and a least squares line fitted to the transformed data, with month as the explanatory
variable. The equation of the least squares line is:
log(shareprice) = 0.855 + 0.154 × month
A $0.50 B $1.04 C $3.16 D $1462.18 E $94.42
17 The table below records the monthly electricity cost (in dollars) for an apartment over
one calendar year.
123 90 153 136 101 129 153 143 95 61 85 107
The six-mean smoothed with centring cost of electricity in July is closest to:
A $129 B $120 C $128 D $131 E $143

18 The time series plot below shows shows earnings per quarter ($000) for a certain
salesperson over a 3 year period.
21
20
19
18
17
Earnings ($000)
16
15
14
13
12
11
10
9
8
1 2 3 4 5 6 7 8 9 10 11 12
Quarter
The nine-median smoothed earnings for the salesperson in Quarter 8, in $000s, is

closest to:
A 14 B 16 C 17.5 D 18 E 20
19 The table below shows the long term mean monthly sales figures (in $’000s) for a
company, and the associated seasonal indices for the sales. The long-term mean sales
figure for January is missing.
Sales 80.0 70.3 62.6 54.6 55.0 52.1 54.2 56.5 52.8 61.8 99.7
SI 0.727 1.289 1.132 1.008 0.880 0.886 0.840 0.874 0.911 0.850 0.996 1.607
The long-term mean sales figure for January is closest to:
A 45.1 B 58.9 C 62.1 D 73.4 E 85.4
20 The number of job applications received by a large supermarket chain is seasonal. Data
has been collected, and a least squares regression line fitted to the deseasonalised data.
The equation of the line is
deseasonalised job applications = 457.8 + 12.27 × month number
where month number 1 is January 2022.
The monthly seasonal indices for job applications are shown in the following table:
1.14 1.06 1.22 1.03 0.95 0.95 0.83 0.70 0.78 0.88 1.23 1.23
The actual number of job applications predicted for January 2023 was closest to
A 605 B 617 C 542 D 690 E 704

16A Recursion and financial modelling 747
Revision
21 Consider the recurrence relation below.

A0 = 20, An+1 = 2An − 26
Which term of the sequence generated by this recurrence relation is the first to be
negative?
A A1 B A2 C A3 D A4 E A5
22 The following recurrence relation can generate a sequence of numbers

M0 = 48 Mn+1 = Mn + B
The value of M3 is 30. The value of B is
A −18 B −6 C −3 D 3 E 6
23 An amount of money is deposited into an account that earns compound interest.

Which combination of interest rate and compounding period has the largest effective
interest rate?
A 4.0% per annum, compounding quarterly
B 4.1% per annum, compounding weekly
C 4.1% per annum, compounding monthly
D 4.2% per annum, compounding fortnightly
E 4.2% per annum, compounding quarterly
24 A tractor was purchased for $189 000.

Using the reducing balance method, the value of the tractor depreciates by 8% each
year.
Which one of the following recurrence relations could be used to determine the value
of the tractor after n years, Vn ?
A V0 = 189 000, Vn+1 = 0.92Vn
B V0 = 189 000, Vn+1 = 1.08Vn
C V0 = 189 000, Vn+1 = Vn − 15120
D V0 = 189 000, Vn+1 = 0.92Vn − 15120
E V0 = 189 000, Vn+1 = 1.08Vn − 15120
25 Nicos borrowed $15 000 to pay for a holiday.

He was charged interest at the rate of 5.9% per annum, compounding monthly.
The loan was repaid with monthly repayments of $600.
After four months, the total interest that Nicos had paid was closest to
A $74 B $280 C $2100 D $2400 E $13 000

26 Reuben has purchased a new oven for his restaurant for $37 000.
He depreciates the value of the oven using the unit cost method at the rate of $3 per
hour of use.
The recurrence relation below can be used to model the value of the oven, Vn , after n
years.
V0 = 37 000, Vn+1 = Vn − 5460
Reuben uses the oven all 52 weeks of the year for the same number of hours each
week.
The number of hours each week that the oven is used for is closest to
A 15 B 21 C 35 D 105 E 5460
27 The first three lines of an amortisation table for a reducing balance loan are shown
below.
0 0.00 0.00 0.00 140 000.00
1 755.00 560.00 195.00 139 805.00
2 755.00 559.22
What is the principal reduction from repayment number 2?

A $195.00 B $195.78 C $560.00 D $559.22 E $755.00
28 An annuity investment earns interest at the rate of 5.3% per annum, compounding
monthly.
Tim initially invested $60 000 and will add monthly payments of $1600.
The value of this investment will first exceed $74 000 after
A five months
B six months
C seven months
D eight months
E nine months
29 Raj borrowed $54 000 to buy a car and was charged interest at the rate of 9.6% per
For the first year of the loan, Raj made monthly repayments of $1080.
For the second year of the loan, Raj made monthly repayments of $1200.
The total amount of interest that Raj paid over this two-year period is closest to
A $8785 B $12 960 C $14 400 D $18 575 E $27 360

Revision
30 Victoria has invested an amount of money in a perpetuity.
The perpetuity earns interest at the rate of 4.8% per annum.
Interest is calculated and paid quarterly.
If Victoria receives $1020 per quarter from the perpetuity, then the amount that she has
invested is
A $4896 B $21 250 C $85 000 D $255 000 E $489 600
Matrices
 
3 4 2
31 If matrix M =   then its transpose M T is
5 6 8

     
3 4 3 5     4 3
    5 6 8 5 4 8  
A 5 6
  B 4 6
  C   D   E 6 5
3 4 2 3 6 2
      
7 8 2 8 8 7
 
 
3 4  
  1 9 11
32 Matrix A = 5 6 and matrix B =   . Matrix R = A × B. Element r32 is
0 12 14
  
7 0
calculated by
A 3 × 9 + 4 × 12 B 7 × 9 + 0 × 12 C 1 × 11 + 6 × 14
D 1 × 11 + 6 × 14 E 7 × 11 + 0 × 12
33 A gymnasium has the same number of customers every day. They can do either
pilates (P) or yoga (Y). The customers may change activities every day. Their change
in involvement is shown in the transition matrix below.
Today
P Y
 
P 65% 70%
Y 35% 30%
There must be 20 customers in the pilates class. Each day the number of customers in
the yoga class is
A 10 B 12 C 15 D 20 E 25
34 Consider the following communication matrix

receiver
 A B C D E
 
A  0 1 1 0 0
 
B  0 0 0 1 0
 
sender C  0 1 0 0 1
 
D  1 0 1 0 0
 
E  1 1 0 1 0

How many 4-step communications are there from C to B?

A 1 B 2 C 3 D 4 E 5
35 Matrices A, B and C have orders 2 × 5, 2 × 4 and 5 × 4 respectively. What is the order

of the matrix C T × (BT × A)T ?
A 5×5 B 4×5 C 2×4 D 4×4 E 2×5
36 Matrix A is a 5 × 5 matrix.
Matrix B is a row matrix. Matrix C is a column matrix.
Which one of the matrix products below could result in a 1 × 1 matrix?
A ACB B ABC C CAB D BAC E BCA
   
0 1 0 0 0 a
   
0 0 0 1 0 b
 

Matrix P = 1 0 0 0 0 and matrix Z = c. The smallest value of n such that
  
37
   
0 0 0 0 1 a
   
0 0 1 0 0 b
  
Pn Z = Z is
A 1 B 2 C 3 D 4 E 5
Questions 38 and 39 are based on the following information.

A survey company has selected 1000 people to rate a news program. The rating given
by a survey participant can be low (L), medium (M) or good (G). By observing for a
long while it is observed that the participants change their rating from week to week as
shown in the transition matrix T below.
This week
L M G
 
0.8 0.2 0.3  L
 
T = 0.1 0.7 0.1  M Next week
 
0.1 0.1 0.6 G
The expected number of each rating received after n weeks can be determined by the
recurrence relation
300
 
S 0 = 600 , S n+1 = T S n
 
100
where S 0 is the state matrix for the first week of the survey with 300 low ratings, 600
medium ratings and 100 high ratings.
38 What percentage of these 1000 participants are not expected to change their rating in
the second week from the first?
A 60% B 43% C 56% D 72% E 82 %
Revision
39 In the long term (say after more than 20 weeks of the survey), how many of the
participants will give a low rating?
A 150 B 250 C 350 D 450 E 550
40 Consider the matrix recurrence relation below.
   
25  x 0.4 0.6
   
S 0 = 16 , S n+1 = T S n where T = y 0.2 0.4
   
56 z w v
(Matrix T is a regular transition matrix, The columns of the matrix T sum to one).
Which one of the following statements is not necessarily true?
A z+w+v=1 B x+y+z=1 C w = 0.4
D v=0 E z + w + v ≥ 0.4
41 Four teams, X, Y, Z and W, competed in a round-robin competition where each team

played each of the other teams once. There were no draws. The results are shown in the
matrix below.
loser
X Y Z W
X  0 1 1 0 
 
 
winner  a
Y  0 0 1 

Z  0 1 0 c 
 
W b 0 1 0
A ‘1’ in the matrix shows that the team named in that row defeated the team named in
that column. In this matrix, the values of a, b and c are
A a = 1, b = 0, c = 0 B a = 0, b = 1, c = 0 C a = 1, b = 0, c = 1
D a = 0, b = 1, c = 1 E a = 1, b = 1, c = 0
42 Consider the graph below.
The number of vertices with a degree greater than or equal to 5 is

A 0 B 1 C 2 D 3 E 4

43 What is the maximum number of edges a bipartite graph with 30 vertices can have?
A 60 B 125 C 200 D 225 E 250
44 Consider the graph opposite. Euler’s formula can be verifted for this graph. What
values of e, v and f can be used in this verification?
A e = 5, v = 5, f = 3
B e = 7, v = 5, f = 4
C e = 6, v = 5, f = 3
D e = 6, v = 5, f = 4
E e = 8, v = 4, f = 4
45 The graph opposite has five vertices and ten edges.

How many of the vertices in this graph have an
even degree?
A 0 B 1 C 2 D 3 E 4
46 For a connected graph with five vertices and five edges, the sum of the degrees of the
vertices is
A 4 B 6 C 8 D 9 E 10
47 If an extra edge is added to this graph it will have C

an Eulerian circuit. The edge is B I
A AB B BG C QG
D IB E AG
A
F H
48 The directed graph below shows the sequence of activities required to complete
a project. The time taken to complete each activity, in hours, is also shown. The
minimum completion time for this project is 21 hours. The time taken to complete
activity G is labelled x. The maximum value of x is
C, 10
L, 2
A, 4 G, x
D, 4 K, 2 Finish
Start M, 5
F, 3 H, 2
B, 3 J, 4
E, 6
A 1 B 2 C 3 D 5 E 7

Revision
49 The flow of water through a series of pipes is shown in the network below. The
numbers on the edges show the maximum flow through each pipe in litres per minute.
A 3 C
5
5 4 Sink
4 6
5 Cut
Source 5 B 3 D
The capacity of the cut in litres per minute is
A 10 B 12 C 14 D 18 E 20
50 The network opposite shows the cabling between five locations B

A, B, C, D and E. C
An adjacency matrix for this network is formed. The number of A
zeros in this matrix is
A 8 B 9 C 10 D 11 E 12 D
51 The minimum spanning tree for the graph opposite has 18

a weight of 16 12 26
16
A 52 B 72 C 76
8 28 10
D 80 E 86 24 6
14
16B Exam 2 questions

1 The weights (in kg) carried by the horses in a handicap race are given below.
60 57 57 55 54 53 53 53 52 52 51.5 51
a Calculate the mean and the standard deviation of the weights carried by the horses
in this race, rounding your answers to three decimal places.
b One horse in the race carried a weight of 51 kg. Use your answer to part a. to
calculate the standardised score (z) for the weight carried by that horse, rounding
your answer to one decimal place.
c Suppose that the weights carried by horses in handicap races are approximately
normally distributed. If the weight carried by one horse, Silver, has a standardised
score of z = −2:
i Using the mean and standard deviation determined in part a., how much weight
was carried by Silver? Round your answer to one decimal place.
ii What percentage of horses would be expected to carry weight less than Silver?

2 In a study on the effect of distractions on concentration, three groups of subjects were

timed completing a complex task. One group completed the task with no distractions, a
second group with a few distractions, and the third group with many distractions. The
two variables in this study are number of distractions and time in seconds.
Five number summaries describing the distribution of time for each group are displayed
below, along with the size of each group.
The associated boxplots are shown following the table.
Number of distractions Group size Min Q1 Median Q3 Max

none 32 15.2 22.0 25.0 28.2 38.0
a few 36 21.0 23.0 26.2 29.4 36.0
many 36 23.4 26.0 29.2 36.0 42.1
44
42
40
38
36
34
time(secs)
32
30
28
26
24
22
20
18
16
14
None A few Many
number of distractions
a Which variable is the explanatory variable and which is the response variable in this
study?
b Which variables in this study are numerical?
c For the group who completed the task with no distractions
i What is the interquartile range?
ii Show why the value of 38.0 seconds is an outlier for this group.
d People who took more than 36 seconds to complete the task were classified as slow.
How many people in the study would be classified as slow?
e Do the boxplots support the contention that there is an association between number
of distractions and time? Refer to the values of an appropriate statistic in your
response.
3 The equation of the least squares line that relates the fuel consumption of a certain car,
in litres/100 km, to the speed at which the car is travelling, in km/hr is:
fuel consumption = 6.827 + 0.0218 × speed

16B Exam 2 questions 755
Revision
a Use the summary statistics fuel consumption speed
shown to determine the
mean 8.7556 88.444
coefficient of determination as
a percentage, rounded to one standard deviation 0.52941 22.367
decimal place.
b Interpret the value of the coefficient of determination in terms of fuel consumption
and speed.
c Use the equation to predict the fuel consumption of the car if it is travelling at 100
km/hr. Round the answer to one decimal place.
d Write down the slope of the regression line and interpret in terms of fuel consumtion
and the speed.
e When the speed was 72 km/hr, the actual fuel consumption was 8.3 litre/100 km.
Show that, when the least squares line is used to predict the fuel consumption at 72
km/hr, the residual is −0.10 rounded to two decimal places.
4 In a study of the association between a person’s enthusiasm for their job (a numerical
variable measured on a scale from 0 to 15), and their efficiency when performing their
job (a numerical variable measured on a scale from 0 to 25), the following data was
collected from a group of 12 employees.
enthusiasm 14.2 13.2 11.6 12.6 11.6 10.0 8.8 8.3 5.1 4.5 2.8 2.4
efficiency 22.5 16.7 12.0 11.3 10.0 5.2 6.1 4.0 3.4 2.2 2.1 1.9
The following scatterplot was constructed, with enthusiasm as the explanatory variable,
and efficiency as the response variable.
25
20
15
efficiency
10
0
2 4 6 8 10 12 14 16
enthusiasm
a Describe the association between efficiency and enthusiasm in terms of form and
strength.
b Which transformations could be used in order to linearise the association?
c Apply a log transformation to the variable efficiency to linearise the association. Fit
a least squares line to the transformed data, and write down its equation. Round the
values of the intercept and slope to three significant figures.
d Use the equation from part c. to predict the efficiency score for a person who scores
ISBN 978-1-009-11041-9 9.3 on enthusiasm. © Peter Jones et al 2023 Cambridge University Press
5 The table below shows the quarterly house sales achieved by a real estate company in
the years 2020-2021.
Year Q1 Q2 Q3 Q4
2020 52 59 68 27
2021 57 65 75 29
a Use the data in the table to find seasonal indices. Give your answers rounded to two
decimal places.
b The number of houses sold in each of the four quarters in 2022 is shown in the table
below.
Year Q1 Q2 Q3 Q4
2022 63 69 80 33
Use the seasonal indices from part a to deseasonalise the data. Round your answers
6 The following table gives the value of average price of unleaded petrol in Victoria each
year from 2015–2021.
Year 2015 2016 2017 2018 2019 2020 2021
Price (cents/litre) 126.3 116.4 128.7 143.4 141.1 123.9 147.6
a Find the centred four-mean smoothed value of petrol in Victoria for the year 2017 in
cents/litre, rounding your answer to one decimal place.
b Find the five-median smoothed value of petrol in Victoria for the year 2019 in
cents/litre, rounding your answer to one decimal place
The following time series plot shows the price of petrol in Victoria in cents/litre,
and the price of petrol in the Northern Territory (NT) in cents/litre, over the years
2002–2021.
180
Price of petrol (cents/litre)
160
140
120
100
80
60
2000 2005 2010 2015 2020 2025
Year
VIC NT

Revision
Least squares regression lines (shown on the plot) have been fitted to both sets of time
series data, and the following equations determined:
Victoria: petrol price = −4071.53 + 2.08699 × year
NT: petrol price = −4290.07 + 2.20128 × year
c i Write down the slope of the least squares line for Victoria rounded to two
decimal places, and interpret.
ii Write down the slope of the least squares line for the NT rounded to two decimal
places, and interpret.
d Use the least squares regression lines to predict the price of petrol in 2026:
i in Victoria ii in the NT
e Do the equations predict that the difference in petrol prices between Victoria and
the NT will decrease, stay the same, or increase? Explain your answer, quoting
appropriate statistics.
7 Joslyn originally paid $9500 for office furniture that she now wishes to sell. Joslyn will
sell the furniture at a depreciated value.
a Joslyn could use a reducing balance method, with an annual depreciation rate of
7.4%.
Using this depreciation method, what is the value of the furniture five years after it
was purchased? Round your answer to the nearest cent.
b If Joslyn used a reducing balance depreciation method and the furniture was sold
for $6890 after five years, what annual percentage rate of depreciation did this
represent? Give your answer correct to one decimal place.
c Joslyn could use a flat rate depreciation method.
Let Jn be the value, in dollars, of Joslyn’s furniture n years after it was purchased.
The value of the furniture, Jn can be modelled by the recurrence relation below.
J0 = 9500, Jn+1 = Jn − 855
i Using this depreciation method, what is the value of the furniture five years after
it was purchased?
ii What annual flat rate of depreciation is represented?
8 Marina would like to buy a new bicycle and has saved $2600.
a Marina could invest this money in an account that pays interest which compounds
monthly.
The balance of this investment after n months, Mn , could be determined using the
recurrence relation below.
M0 = 2600, Mn+1 = 1.003 × Mn

Calculate the total interest that would be earned by Marina’s investment in the first
five months. Round your answer to the nearest cent.
Marina could invest the $2600 in a different account that allows her to make an
additional payment of $140 each month.
b Marina would like to have a balance of $5500, to the nearest dollar, after 18 months.
What annual interest rate would Marina require? Round your answer to two decimal
places.
c The interest rate is 3% per annum, compounding monthly. Let Vn be the value of
Marina’s investment after n months.
Write down a recurrence relation, in terms of V0 , Vn+1 and Vn , that would model the
change in the value of this investment.
9 Alessandro borrows $12 000 with interest on the loan charged at the rate of 7.9% per
Immediately after the interest has been calculated and charged each month, Alessandro
will make a repayment.
a Alessandro considers making interest only repayments. What would be the value of
each interest only repayment?
b Alessandro makes equal monthly repayments for four years. After these four
years, the balance of his loan will be $2946.24 correct to two decimal places. What
amount, in dollars, will Alessandro repay each month during the four years?
c If Alessandro instead decides to fully repay the loan in three years with 35 equal
monthly payments followed by a final payment that is as close to the regular
payment as possible. Find both the regular payment and the final payment. Round
your answers to the nearest cent.
Matrices
10 A regional city has three supermarkets HSL (H), Radcliffs (R) and Cottonworths (C)
The total number of shoppers at each of the stores on a weekday is shown in matrix W.
H R C

W = 1500 2500 3200
a Write down the order of matrix W.
Each of the supermarkets has a meat counter and a delicatessen The proportion of daily
shoppers who only purchase from the meat counter (M), only from the delicatessan (D)
and those that do shopping from several sections (G) is the same for each supermarket
and is described in the matrix shown here.
 
0.1 M
P = 0.2 D
 
 
0.7 G

Revision
b Find the matrix Q = P × W and describe what the value element q32 is.
c On a particular day at Cottonworths the amount spent, in dollars, is described by the
following matrix. The amounts indicate the typical amount spent by a customer.
 
20 M
 
A = 20 D
 
60 G
The numbers of each type of of customer that spent exactly these amounts is
descibed by the matrix
M D G
T = 150 250 600
Show a matrix multiplication that will give the total amount spent on that day.
d There can be a lot of change in the daily shopping numbers of customers between
the three supermarkets. We use a 3 × 3 transition matrix and a recurrence relation to
describe the changed shopping locations of customers in the city month by month.
This month
H R C
 
1500 H  
  0.13 0.8 0.2  H
Let S 0 = 2500 R and T = 
  


  
   0.7 0.1 0.2 R Next month
3200 C 
 
0.17 0.1 0.6 C
The recurrence relation is S n+1 = T × S n where S n is the state matrix n months after
our starting time.
i Find S 1 ii Find S 50
iii Describe the long term situation for the number of customers at each of the
supermarkets.
11 When a new infectious disease was first noticed in a particular country, there were
100 people already infected. As the disease spread, the country’s health authority
collected the following information:
The duration of the disease is at most 3 weeks. People who contract the disease
either die during these 3 weeks or else recover during the third week.
The survival rate of people who have the disease is 90% in the first week and 80% in
the second week.
People who contract the disease are not infectious during the first week. In their
second week of the disease they have a 90% probability of infecting one other
person, and in their third week a 70% probability of infecting one other person.
a Construct a Leslie matrix L for the disease based on three one-week stages.
b Assume that, when the disease is first noticed, the 100 infected people are all in the
second week of the disease. Write down the initial population matrix S 0 .

For each of the following, give answers to the nearest whole number.
c Determine how the disease is spreading by using the Leslie matrix L to find S 1 , S 2 ,
S 3 and S 4 . Comment.
d Find S 40 , and then use S 40 to find S 41 . Verify that, in these two population matrices,
the sizes of the three groups are in nearly the same ratio. Hence estimate the growth
rate of the disease at this stage.
e Suppose that the health authority had taken immediate action to reduce the rate of
infection in the third week to 35%. Repeat parts a–d. Would this action have been
sufficient to eradicate the disease? Give evidence for your answer.
f Now suppose that the health authority had taken more drastic action and reduced the
rate of infection in the third week to 10%. Repeat parts a–d.
12 A certain population of female marsupials is divided into six age groups, each spanning
3 months. The population can then be modelled by the following Leslie matrix, L, and
initial population matrix, S 0 :
   
0 0.3 0.8 0.7 0.4 0  24
   
0.6 0 0 0 0 0  16
   
0 0.9 0 0 0 0  24
L =  and S =
   
 0  
0 0 0.9 0 0 0  8
   
0 0 0 0.8 0 0   0
 
  
0 0 0 0 0.6 0 0
   
a Determine the population matrices S 1 , S 40 and S 41 .

b Using S 40 and S 41 , show that the long-term growth rate of the population is
approximately 1.03.
Now assume that the initial population is 2202 and the sizes of the six age groups
(from youngest to oldest) are 715, 416, 363, 317, 247 and 144.
c Write down S 0 and determine S 1 .
d Using S 0 and S 1 , show that the growth rate of the population over the first 3 months
is approximately 1.03.
13 The choice of seats and their costs at the Leslie theatre are
Stalls (S) $34.00 Balcony (B) $42.00 Dress circle (D) $60.00
The number of seats available in each class are
Stalls (S) 200 Balcony (B) 150 Dress circle (D) 80
a The column matrix A contains the number of seats in each class.
 
200 S
 
A = 150 B
 
80 D
State the order of A.

Revision
b Matrix C gives the cost of each type of seat
S B D

C = 34 42 60
Determine the matrix C × Aand explain

what it represents.
c All seats are sold. Let M = 1 0 1 .
What does the matrix product
 
200 0 0 
 
M  0 150 0  C T
 
0 0 80
represent?
14 a A very large country home garden is divided into five

regions labelled 1 to 5 on the diagram opposite. The red lines 1
represent the boundary stonewalls between two regions.
5
i Draw a graph, where the five regions of the garden 2
4
are represented as vertices and the edges of the graph
represent the boundary stonewalls between areas. 3
ii What is the sum of the degrees of the vertices of this
graph?
b Region 1 of of the garden contains 6 circular B
30
40
beds that are labelled A to F, as shown in the
A C
graph opposite. The owner wants to have a
walk around region 1 visiting each bed on the 10 F
55 35 35
way. The numbers on the edges joining the
vertices give the shortest distance, in metres, E
40 D
between beds.
i Explain why the owner could not follow an Eulerian circuit through this
network.
ii If the owner follows the shortest Hamiltonian path, name a garden bed at which
the owner could start and a garden bed at which the owner could finish.
iii List an Eulerian trail for the graph.
c All areas of the garden require a 15
constant supply of water. The directed 30 sink
20
graph opposite shows an irrigation 15
25 35
system for the garden with capacity 15
source
of each section shown in litres per 15
minute. The beginning of the system
30 40
is labelled source and the end of the
system labelled sink.
i What is the capacity of the marked cut in litres per minute?

ii Determine the maximum flow of water, in litres per minute, from the source to
the sink.
15 a The Penvale swimming club has six new C

members A, B, C, D, E and F. The graph
E
below shows the members who have competed B
together before joining the club. For example, A
the edge between A and B shows that they D
have previously competed together.
F
i How many of these swimmers had E competed with before joining the club?
ii Who had competed with both A and B before joining the club?
b The swimming club has a medley relay team. Three of the new club members, A,
B and C can can complete the following sectors of the medley race: backstroke,
butterfly and breastroke. The table below shows the average times in seconds for
100 m for these sectors for each of the three swimmers. The freestyle swimmer has
been chosen and has much better freestyle times than the three new members.
How should the swimmers be allocated to minimise the team’s time?
Swimmer Backstroke Breaststroke Butterfly

A 72 74 66
B 68 72 62
C 70 76 62
c The Penvale swimming club rooms

Activity EST Duration Immediate
are to undergo renovation. This
predecessor(s)
project involves eight activities:
A 0 3 −
A to I. The table opposite shows
the earliest start time (EST) B 0 6 −
and duration, in months, for C 6 2 A, B
each activity. The immediate D 8 ... C
predecessor(s) is also shown. The E 8 9 C
duration for activity D is missing. F 6 3 B
The information in the table
G 15 4 D
opposite can be used to complete an
H 9 9 F
activity network. This network will
I 19 2 E, G, H
require a dummy activity.
i What is the duration, in months, of activity D?
ii Draw the associated activity network for this renovation.
iii Name the four activities that have a float time.

Revision
iv The project is to be crashed by reducing the completion time of one activity
only. What is the minimum time, in months, that the project can be completed
in?
16 A family is visiting a theme park and will B

visit five rides. A map of the theme park is
shown opposite with the vertices representing D
the rides and the edges representing the paths

A
connecting the rides.
E
C
a Determine a Hamiltonian cycle beginning and ending at A that the family can
follow.
b Determine an Eulerian trail that the family can follow. Explain how you determined
this trail.
The network opposite shows the lengths B 1.6 km
of the paths that join the rides. The
family decides to visit the rides by 2.4 km D
4.2 km
following an Eulerian trail. Assume
A 3 km 1.8 km 2.1 km
that the family can walk at a speed of 3
km/hr. The theme park will close at 5 3.2 km
E
p.m. C
c What is the latest time that the family
3 km
can enter the theme park?
17 The diagram below shows the buildings of a new university. The lines on the diagram
show the location of the pathways between the buildings.
20 m 20 m Office
A
10 m 25 m
30 m 10 m
30 m 20 m 10 m
B
10 m 10 m
25 m
15 m
C 20 m 10 m D
a i How many different ways can a student walk directly from building A to
building B?
ii Represent this diagram as a weighted graph in planar form.

iii Which buildings are immediately adjacent to building C?
Some of the pathways will be covered to protect students from the rain as they move
between buildings. The covering structure will cost $240 per metre to make and install.
b i Modify your planar weighted graph from question aii above to show only the
shortest direct pathway between adjacent buildings.
ii How much will the covered walkways on these pathways cost to build?
c It has been decided that covering all of these walkways is too expensive. Only the
minimum number of pathways that are necessary to allow students to walk from one
building to any other while remaining under cover will be built.
i Draw a graph that shows the pathways that should be covered so that the overall
cost of making and installing the covering structure is a minimum.
ii Calculate the cost of the covering structure in part i.
18 Anthony is creating a robot for a university project. The activities required to design
and build the robot are shown in the table below, along with their duration in weeks and
the immediate predecessor for each activity.
Activity Description Duration Immediate

(weeks) predecessors
A Research robot design and control 5 -
B Design the internal electronics 8 A
C Design the remote control 3 A
D Construct and assemble the robot 15 B
E Write the code to control the robot 10 B
F Construct and program the remote 6 C
control
G Debug the code to control the robot 4 E

H Install the software 1 D, F
I Test the robot 3 G, H
a Construct an activity network for this project.

b Determine the shortest time in which Anthony can expect to create his robot.
c i Construct a table that shows the EST, LST and float for each activity.
ii Write down the critical path of this project.
d Use the information in the table from ci to describe and explain what would happen
if Anthony took:
i 3 weeks to research robot design and control instead of 5
ii 10 weeks to construct and program the remote control instead of 6
iii 20 weeks to construct and assemble the robot instead of 15

Glossary
Glossary
A → B
68–95–99.7% rule: [p. 77] A rule for reducing balance loan or annuity on a step-by-step
determining the percentage of values that lie (payment-by payment) basis or the payment of
within one, two and three deviations of the mean a compound interest investment with additional
in a normal distribution. payments.
Annuity: [p. 402] An annuity is a compound
A interest investment from which regular payments
are made.
Activity: [p. 683] A task to be completed as
part of a project. Activities are represented by the
edges in the project diagram.
Activity network: [p. 684] An activity B
network is a weighted directed graph that shows Backward scanning: [p. 696] Backward
the required order of completion of the activities scanning is the process of determining the LST
that make up a project. The weights indicate the for each activity in a project activity network.
durations of the activities they represent.
Balance: [p. 342] The balance of a loan or
Adding to the principal: see annuity investment is the amount owed or accrued after a
investment. period of time.
Adjacency matrix: [p. 619] A square matrix Bar chart: [p. 8] A statistical graph used to display
showing the number of edges joining each pair of the frequency distribution of categorical data.
vertices in a graph.
Binary matrix: [p. 523] A matrix whose
Algorithm: [p. 637] A step-by-step procedure elements are either zero or one.
for solving a particular problem that involves
Bipartite graph (bigraph): [p. 674] A graph
applying the same process repeatedly. Examples
whose set of vertices can be split into two subsets,
include Prim’s algorithm and the Hungarian
X and Y, in such a way that each edge of the graph
algorithm.
joins a vertex in X and a vertex in Y.
Allocation: [p. 74] Allocation is the process of
Bivariate data: [p. 104] Data in which each
assigning a series of tasks to different members
observation involves recording information about
of a group in a way that enables the tasks to be
two variables for the same person or thing. An
completed for the minimum time or cost.
example would be data recording the height and
Amortisation: [p. 407] Amortisation is the weight of the children in a preschool.
repayment of a loan or an investment with regular
Boxplot: [p. 61] A graphical display of the
payments made over a period of time.
five-number summary of a data set showing
Amortisation table: [p. 408] An amortisa- outliers if present. See outliers.
tion table charts the amortisation (repayment) of a

C 766 Glossary
Bridge: [p. 611] A single edge in a connected Compound interest: [p. 360] Where the
graph that, if removed, leaves the graph interest paid on a loan or investment is added to
Glossary
disconnected. the principal and subsequent interest is calculated

on the total.
C Compound interest investments with

additional payments: [p. 393] A compound
Capacities (flow network) [p. 660] The interest investments with periodic payments
weights of the directed edges in a flow network is an investment to which additions are made to
are called capacities. They give the maximum the principal on a regular basis. Also known as
amount that can move between the two points in ‘adding to the principal’.
the flow network represented by these vertices in a
Compounding period: [p. 393] The
particular time interval. This could be, for example,
compounding period is the time period for
the maximum amount of water in litres per minute
the calculation of interest for an investment or
or the maximum number of cars per hour.
loan. Typical compounding periods are yearly,
Categorical variable: [p. 2] Categorical quarterly, monthly or daily.
variables are used to represent characteristics of
Connected graph: [p. 611] A graph in which
individuals, for example place of birth, house
there is a path between every pair of vertices.
number. Categorical variables come in types,
nominal and ordinal. Continuous variable: [p. 3] A numerical
variable that represents a quantity that is measured
Centre of a distribution: [p. 22] A measure
rather than counted, for example the weights of
of location of a distribution. Measures of centre
people in kilograms.
include the median and the mean.
Correlation coefficient r: [p. 141] A
Centring: [p. 270] If smoothing takes place
statistical measure of the strength of the linear
over an even number of data values, the smoothed
association between two numerical variables.
values do not align with an original data value. A
Cost matrix: [p. 676] A cost matrix is a table
second stage of smoothing is carried out to centre
that contains the cost of allocating objects from
the smoothed values at an original data value.
one group, such as people, to objects from another
Circle of transformations: [p. 239] group, such as tasks. The cost can be money, or
Provides guidance in choosing the transformations other factors such as the time taken to complete
that can be used to linearise various forms of the project.
scatterplots. Crashing: [p. 708] Crashing is the process of
Circuit: [p. 625] A walk with no repeated edges shortening the length of time taken to complete a
that starts and ends at the same vertex. See also project by reducing the time required to complete
cycle (graphs). individual activities.
Coefficient of determination (r2 ): [p. 146] Critical path: [p. 698] The project path that
A coefficient which gives a measure of the has the longest completion time.
predictive power of a regression line. It gives Critical path analysis: [p. 698] A project
the percentage of variation in the RV that can be planning method in which activity durations are
explained by the variation in the EV. known with certainty.
Column matrix: [p. 474] A matrix with only Cut: [p. 663] A line dividing a directed (flow)
one column. graph into two parts in a way that separates all
Column vector: [p. 474] see column matrix. ‘sinks’ from their ‘sources’.
Communication matrix [p. 525] A square Cut capacity: [p. 663] The capacity of a cut
binary matrix in which the 1s represent direct is the sum of the capacities of the cuts passing
(one-step) communication links. through the cut that represents flow from the
source to the sink. Edges that represent flow from
Complete graph: [p. 611] A graph with edges
the sink to the source do not contribute to the
connecting all pairs of vertices.
capacity of the cut.

Glossary 767
Glossary
Cycle (graphs): [p. 625] A walk with no to ensure that all predecessor activities are
repeated vertices that starts and ends at the same properly accounted for.
vertex. See also circuit.
Cycle (time series): [p. 258] Periodic
movement in a time series but over a period
greater than a year. E
Earliest starting time (EST): [p. 695] The
D → E
D earliest time an activity in a project can be started.
Edge: [p. 609] A line joining one vertex in
Data transformation: [p. 211] Using a
a graph or network to another vertex or itself
mathematical rule to change the scale on either
(a loop).
the x- or y-axis in order to linearise a non-linear
scatterplot. Effective interest rate: [p. 375] Used to
compare the interest paid on loans (or invest-
Degenerate graph: [p. 610] A graph in which
ments) with the same annual nominal interest rate
no vertex is connected to any other vertex. All the
r but with different compounding periods (daily,
vertices are isolated.
monthly, quarterly, annually, other).
Degree of a vertex (deg(A)): [p. 609] The
number of edges attached to the vertex. The Elements: [p. 474] The numbers or symbols
degree of vertex A is written as deg(A). displayed in a matrix.
Depreciation: [p. 343] The reduction in value Equal matrices: [p. 491] Matrices that have
of an item over time. the same order and identical elements in identical
Deseasonalise: [p. 285] The process of positions.
removing seasonality in time series data. Equivalent graph: [p. 613] see isomorphic
Determinant: [p. 514] A number associated graphs.
with square matrices. The determinant of a matrix
Eulerian circuit: [p. 627] An Eulerian walk
A, written det(A), is used to decide if the matrix
that starts and finishes at the same vertex. To have
has an inverse. If det(A) = 0, the matrix has no
an Eulerian circuit, a network must be connected
inverse; it is singular.
and all vertices must be of even degree.
Dijkstra’s algorithm: [p. 637] An algorithm
Eulerian trail: [p. 627] A walk in a graph or
for finding the shortest path between two vertices
network that includes every edge just once (but
in a weighted graph. Pronounced ‘Di-stra’: ‘Di’ as
does not start and finish at the same vertex).
in ‘die’ and ‘stra’ as in ‘car’.
To have an Eulerian walk (but not an eulerian
Directed graph (digraph): [p. 660] A graph
circuit), a network must be connected and have
or network in which directions are associated with
exactly two vertices of odd degree, with the
each of the edges.
remaining vertices having even degree.
Discrete variable: [p. 3] A numerical
variable that represents a quantity that is Euler’s formula: [p. 614] The formula
determined by counting; for example, the number v − e + f = 2, which relates the number of
of people waiting in a queue is a discrete variable. vertices, edges and faces in a connected graph.
Dominance matrix: [p. 531] A square Explanatory variable: [p. 105] When
binary matrix in which the 1s represent one-step investigating associations in bivariate data, the
dominances between the members of a graph. explanatory variable (EV) is the variable used
to explain or predict the value of the response
Dot plot: [p. 28] A statistical graph that uses
variable (RV).
dots to display individual data values on a number
line; suitable for small sets of data only. Extrapolation: [p. 182] Using a mathematical
Dummy activity: [p. 687] An artificial activity model to make a prediction outside the range of
of zero time duration added to a project diagram data used to construct the model.

768 Glossary
F H
F → I
Face: [p. 614] An area in a graph or network Hamiltonian cycle: [p. 627] A Hamiltonian
that can only be reached by crossing an edge. One path that starts and finishes at the same vertex.
such area is always the area surrounding a graph. Hamiltonian path: [p. 627] A path through a
Glossary
Finance Solver: [p. 422] A finance solver is graph or network that passes through each vertex
a computer/calculator application that automates exactly once. It may or may not start and finish at
the computations associated with analysing a the same vertex.
reducing balance loan, an annuity or an annuity Histogram: [p. 15] A statistical graph used to
investment. display the frequency distribution of a numerical
Five-number summary [p. 61] A list of the variable; most suitable for medium to large sized
five key points in a data distribution: the minimum data sets.
value (min), the first quartile (Q1 ), the median Hungarian algorithm: [p. 676] An algorithm
(M), the third quartile (Q3 ) and the maximum for solving allocation (assignment) problems.
value (max).
Flat-rate depreciation: [p. 344] Deprecia-
tion where the value of an item is reduced by the I
same amount each year. Flat-rate depreciation is Identity matrix (I): [p. 477] A matrix that
equivalent, but opposite, to simple interest. behaves like the number one in arithmetic. Any
Float (slack) time: [p. 694] The amount of matrix multiplied by an identity matrix remains
time available to complete a particular activity that unchanged. An identity matrix is represented by
does not increase the total time taken to complete the symbol I.
the project. Immediate predecessor: [p. 683] An
Flow: [p. 660] Flow is the movement of activity that must be completed immediately
something from a source to a sink. before another one can start.
Forward scanning: [p. 695] Forward Initial state matrix: [p. 560] A column
scanning is the process of determining the EST matrix used to represent the starting state of a
for each activity in a project activity network. dynamic system.
Frequency table: [p. 7] A listing of the values Interest: [p. 342] The amount of money paid
a variable takes in a data set along with how often (earned) for borrowing (lending) money over a
(frequently) each value occurs. Frequency can be period of time.
recorded as a count or as a percentage. Interest-only loans: [p. 446] A loan on
which only the interest is paid. At the end of the
loan, the principal must be repaid in full.
G Interest rate: [p. 342] The rate at which

interest is charged or paid. Usually expressed as a
Geometric decay [p. 359] When a recurrence percentage of the money owed or lent.
rule involves multiplying by a factor less than
Interpolation: [p. 182] Using a regression line
one, the terms in the resulting sequence are said to
to make a prediction within the range of values of
decay geometrically.
the explanatory variable.
Geometric growth [p. 359] When a recur-
Interquartile range (IQR): [p. 47] The
rence rule involves multiplying by a factor greater
interquartile range is defined as IQR = Q3 − Q1 .
than one, the terms in the resulting sequence are
IQR gives the spread of the middle 50% of data
said to grow geometrically.
values.
Graph: [p. 609] A collection of points called
Inverse matrix: [p. 513] A matrix which,
vertices and a set of connecting lines called edges.
when multiplied by the original matrix, gives

Glossary 769
Glossary
the identity matrix (I). For a matrix A, the
inverse is written as A−1 and has the property that
A−1 A = AA−1 = I. Matrix: [p. 473] A rectangular array of numbers
or symbols set out in rows and columns within
Irregular (random) fluctuations [p. 262]
square brackets (plural: matrices).
Unpredictable fluctuations in a time series.
Always present in any real world time series plot. Matrix multiplication: [p. 499] The process
of multiplying a matrix by a matrix.
Isolated vertex: [p. 610] A vertex that is not
L → M
connected to any other vertex. Its degree is zero. Maximum flow (graph): [p. 661] The
capacity of the ‘minimum’ cut.
Isomorphic graphs: [p. 613] Equivalent
graphs. Graphs that have the same number of Maximum or minimum value of the
edges and vertices that are identically connected. objective function: [p. 759] The value found
by evaluating the objective function’s value at the
Iteration [p. 336] Each application of a
vertices or along the boundaries of the feasible
recurrence rule to calculate a new term in a
region.
sequence is called an iteration.
Mean (x): [p. 50] The balance point of a data
Σx
distribution. The mean is given by x̄ = , where
n
Σx is the sum of the data values and n is the
L number of data values. Best used for symmetric
distributions.
Latest start time (LST): [p. 696] The latest
time an activity in a project can begin, without Median: [p. 44] The median (M) is the middle
affecting the overall completion time for the value in a data distribution. It is the midpoint of a
project. distribution dividing an ordered data set into two
equal parts. Can be used for skewed or symmetric
Least squares method: [p. 170] One way
distributions.
of finding the equation of a regression line. It
minimises the sum of the squares of the residuals. Minimum cut (graph): [p. 664] The cut
It works best when there are no outliers. through a graph or network with the minimum
capacity.
Linear decay [p. 340] When a recurrence rule
involves subtracting a fixed amount, the terms in Minimum spanning tree: [p. 642] The
the resulting sequence are said to decay linearly. spanning tree of minimum length. For a given
connected graph, there may be more than one
Linear growth [p. 340] When a recurrence
minimum spanning tree.
rule involves adding a fixed amount, the terms in
the resulting sequence are said to grow linearly. Modal category or modal interval: [p. 9]
The category or data interval that occurs most
Linear regression: [p. 169] The process of
frequently in a data set.
fitting a straight line to bivariate data.
Mode: [p. 9] The most frequently occurring
Log scale [p. 34] A scale used to transform
value in a data set. There may be more than one.
a strongly skewed histogram to symmetry or
linearise a scatterplot. Modelling: [ch. 7] Mathematical modelling
is the use of a mathematical rule or formula to
Logarithmic transformations (log x or
represent real-life situations.
log y): [p. 221] Transformations that linearise a
scatterplot by compressing the upper end of the Moving mean smoothing: [p. 268] In
scale on an axis. three-moving mean smoothing, each original
data value is replaced by the mean of itself and
Loop: [p. 610] An edge in a graph or network
the value on either side. In five-moving mean
that joins a vertex to itself.
smoothing, each original data value is replaced
Lower fence: [p. 63] See outliers. by the mean of itself and the two values on either
side.

N → Q 770 Glossary
Moving median smoothing: [p. 277] as data values greater than the upper fence
Moving median smoothing is a graphical (Q3 + 1.5 × IQR) or less than the lower fence
technique for smoothing a time series plot using (Q1 − 1.5 × IQR).
moving medians rather than moving means.
Multiple edges: [p. 610] Where more than one
Glossary
edge connects the same two vertices in a graph.

P
N Parallel box plots: [p. 125] A statistical
graph in which two or more box plots are drawn
Negatively skewed distribution: [p. 21] A side-by-side so that the distributions can be
data distribution with a long tail to the left. compared.
Network: [p. 633] A weighted graph in which Path [p. 625] A walk with no repeated vertices.
the weights are physical quantities, for example See also trail.
distance, time or cost. Percentage frequency: [p. 7] Frequency
Nominal interest rate: [p. 373] The expressed as a percentage.
annual interest rate for a loan or investment that Permutation matrix: [p. 523] A square
assumes the compounding period is 1 year. If binary matrix in which there is only a single one
the compounding period is less than a year, for in each row and column.
example monthly, the actual or effective interest
Perpetuity: [p. 452] An investment where
rate will be greater than r.
an equal amount is paid out on a regular basis
Nominal variable: [p. 3] A categorical forever.
variable that generates data values that can be
Planar graph: [p. 613] A graph that can be
used by name only – for example, eye colour:
drawn in such a way that no two edges intersect,
blue, green, brown.
except at the vertices.
Normal distribution: [p. 76] A data
Positively skewed distribution: [p. 21] A
distribution that has a bell shape. For normal
data distribution that has a long tail to the right.
distributions, the 68–95–99.7% rule can be used
to relate the mean and standard deviation to Power of a matrix: [p. 507] Defined
percentages in the distribution. in the same way as the powers of numbers:
A2 = A × AA3 = A × A × A, etc. Only square
Numerical variable: [p. 3] A variable used to
matrices can be raised to a power. A0 is defined to
represent quantities that are counted or measured.
be I, the identity matrix.
For example, the number of people in a queue, the
heights of these people in cm. Numerical variables Precedence table: [p. 684] A table that
come in types: discrete and continuous. records the activities of a project, their immediate
predecessors and often the duration of each
activity.
O Prim’s algorithm: [p. 642] An algorithm
Order: [p. 474] Used to indicate the size and for determining a minimum spanning tree in a
shape of a matrix. For a matrix with n rows and connected graph.
m columns, the order of a matrix is written as Principal (P): [pp. 342] The initial amount
(n × m). borrowed, lent or invested.
Ordinal variable: [p. 3] A categorical
variable that generates data values that can be
used to both name and order, for example house
number. Q
Outliers: [p. 23] Data values that appear to Quartiles (Q1 , Q2 , Q3 ): [p. 47] Summary
stand out from the main body of a data set. In a statistics that divide an ordered data set into four
boxplot possible outliers are defined equal sized groups.

Glossary 771
Glossary
Row matrix: [p. 474] A matrix with only one
row. A row matrix is also called a row vector.
Range (R): [p. 46] The difference between the
Row vector: [p. 474] See row matrix.
smallest and the largest observations in a data set;
a measure of spread.
Reciprocal transformations (1/x or
1/y): [p. 230] Transformations that linearise a
S
scatterplot by compressing the upper end of the Scalar multiplication: [p. 492] The
R → S
scale on an axis to a greater extent than the log multiplication of a matrix by a number.
transformation. Scatterplot: [p. 130] A statistical graph used
Recurrence relation: [p. 336] A relation that for displaying bivariate data. Data pairs are
enables the value of the next term in a sequence represented by points on a coordinate plane, the
to be obtained by one or more current terms. EV is plotted on the horizontal axis and the RV is
Examples include ‘to find the next term, add two plotted on the vertical axis.
to the current term’ and ‘to find the next term, Scrap value: [p. 344] The value at which an
multiply the current term by three and subtract item is no longer of use to a business.
five’.
Seasonal indices: [p. 284] Indices calculated
Reducing-balance depreciation: [p. 361] when the data shows seasonal variation. Seasonal
When the value of an item is reduced by the indices quantify seasonal variation. A seasonal
same percentage each year. Reducing-balance index is defined by the formula:
depreciation is equivalent to, but opposite to,
compound interest. value for season
seasonal index =
seasonal average
Reducing-balance loan: [p. 400] A loan
that attracts compound interest, but where regular For seasonal indices, the average is 1 (or 100%).
repayments are also made. In most instances the
Seasonality: [p. 259] The tendency for values
repayments are calculated so that the amount of
in the time series to follow a seasonal pattern,
the loan and the interest are eventually repaid in
increasing or decreasing predictably according to
full.
time periods such as time of day, day of the week,
Redundant communication link: [p. 527] A month, or quarter.
communication link is said to be redundant if the
Segmented bar chart: [p. 9] A statistical
sender and the receiver are the same people.
graph used to display the information contained in
Reseasonalise [p. 285] The process of a two-way frequency table. It is a useful tool for
converting seasonal data back into its original identifying associations between two categorical
form. variables.
Residual: [p. 170] The vertical distance from a
Sequence: [p. 334] A list of numbers or
data point to a straight line fitted to a scatterplot is
symbols written down in succession, for example
called a residual:
5, 15, 25, . . .
residual = actual value − predicted value
Residuals are sometimes called errors of Shape of a distribution: [p. 21] The
prediction. general form of a data distribution described
Residual plot: [p. 184] A plot of the residuals as symmetric, positively skewed or negatively
against the explanatory variable. Residual skewed.
plots can be used to investigate the linearity Shortest path: [p. 634] The path through a
assumption. graph or network with minimum length.
Response variable [p. 105] The variable of Simple graph: [p. 610] A graph with no loops
primary interest in a statistical investigation. or multiple edges.
Round-robin tournament: [p. 531] A Simple interest: [p. 342] Interest that is
tournament in which each participant plays each calculated for an agreed period and paid only on
other participant once. the original amount invested or borrowed.

T 772 Glossary
Singular matrix: [p. 517] A matrix that does Steady-state matrix: [p. 565] A column
not have an inverse; its determinant is zero. matrix that represents the final state of a dynamic
Glossary
Sink: See sink and source. system. Also called the equilibrium state.
Sink and source: [p. 660] In a flow network, Stem plot (stem-and-leaf plot): [p. 29]
a source generates flow while a sink absorbs the A method for displaying data in which each
flow. observation is split into two parts, a ‘stem’ and a
‘leaf’. A stem plot is an alternative display to a
Slope (of a straight line): [p. 169] The slope histogram, suitable for small to medium sized data
rise
of a straight line is defined to be: slope = . sets. When data values are tightly clustered, stems
run
The slope is also known as the gradient. can be split to give finer detail.
Smoothing: [p. 268] A technique used to Strength of a linear relationship:
eliminate some of the variation in a time series [p. 143] Classified as weak, moderate or strong.
plot so that features such as seasonality or trend Determined by observing the degree of scatter in a
are more easily identified. scatterplot or calculating a correlation coefficient.
Source See sink and source. Structural change (time series) [p. 260] A
sudden change in the established pattern of a time
Spanning tree: [p. 641] A subgraph of a
series plot.
connected graph that contains all the vertices of
the original graph, but without any multiple edges, Subgraph: [p. 611] Part of a graph that is also
circuits or loops. a graph in its own right.
Spread of a distribution: [p. 46 & p. 52] Summary statistics [p. 43] Statistics that
A measure of the degree to which data values give numerical values to special features of a
are clustered around some central point in the data distribution, such as centre and spread.
distribution. Measures of spread include the Summary statistics include the mean, median,
standard deviation (s), the interquartile range range, standard deviation and IQR.
(IQR) and the range (R). Symmetric distribution: [p. 21] A data
Square matrix: [p. 476] A matrix with the distribution in which the data values are evenly
same number of rows as columns. spread out around the mean. In a symmetric
distribution, the mean and the median are the
Squared transformations (x2 or y2 ):
same.
[p. 212] Transformations that linearise a
scatterplot by stretching out the upper end of the
scale on an axis.
Standard deviation (s): [p. 52] A summary
T
Time series data: [p. 253] A collection of
statistic that measures the spread of the data data values along with the times (in order) at
values aroundrtheP mean. 2The standard deviation is which they were recorded.
(x − x̄)
given by s = . Time series plot: [p. 253] A line graph where
n−1
Standardised (z) scores: [p. 80] The value the values of the response variable are plotted in
of the standard score gives the distance and time order.
direction of a data value from the mean in terms of Trail [p. 625] A walk with no repeated edges.
standard deviations. See also path.
The rule for calculating a standardised score is:
actual score − mean Transition matrix (T): [p. 551] A square
standardised score = matrix that describes the transitions made between
standard deviation
State matrix: [p. 559] A column matrix that the states of a system.
represents the starting state of a dynamic system. Transpose: [p. 475] The transpose of a matrix
Statistical question: [p. 10] A question that is obtained by interchanging its rows and columns.
depends on data for its answer. Tree: [p. 641] A connected graph with no
circuits, multiple edges or loops.

Glossary 773
Glossary
Trend: [p. 257] The tendency for values in the Univariate data: [p. 1] Data associated with a
time series to generally increase or decrease over a single variable.
significant period of time. Upper fence: [p. 63] See outliers.
Trend line forecasting: [p. 296] Using a line
fitted to an increasing or decreasing time series to
predict future values. V
Triangular matrix: [p. 477] An upper Vertex (graph): [p. 609] The points in a graph
U → Z
triangular matrix is a square matrix in which all or network (pl vertices).
elements below the leading diagonal are zeros.
A lower triangular matrix is a square matrix in
which all elements above the leading diagonal are W
zeros. Walk [p. 624] Any continuous sequence of
Two-way frequency table: [p. 110] A edges, linking successive vertices, that connects
frequency table in which subjects are classified two different vertices in a graph. See also trail
according to two categorical variables. Two-way and path.
frequency tables are commonly used to investigate Weighted graph: [p. 633] A graph in which
the associations between two categorical a number representing the size of some quantity
variables. is associated with each edge. These numbers are
called weights.
U
Unit-cost depreciation: [p. 345] Deprecia- Z
tion based on how many units have been produced Zero matrix (O): [p. 493] A matrix that
or consumed by the object being depreciated. For behaves like zero in arithmetic. Represented by
example, a machine filling bottles of drink may be the symbol O. Any matrix with zeros in every
depreciated by 0.001 cents per bottle it fills. position is a zero matrix.

1A → 1B
Answers
Answers
Chapter 1 b
Frequency
Shoe size Count %
Exercise 1A
8 6 50.0
1 a numerical b numerical c categorical 9 3 25.0
d categorical e numerical 10 2 16.7
f numerical g categorical
11 0 0
h categorical
12 1 8.3
2 a nominal b nominal c ordinal
d ordinal e ordinal f nominal
Total 12 100.0
3 a discrete b discrete c continuous 2 a categorical

d continuous e discrete b Frequency
f continuous State of residence Count %
4 B 5 D 6 B Victoria 6 54.5
SA 1 9.1
Exercise 1B WA 4 36.4
1 a Total 11 100.0
Frequency
c 7
Grades Count %
6
A 3 27.3 5
Frequency
4
B 5 45.5 3
C 3 27.3 2
1
Total 11 100.1 0
0 Victoria SA WA
State of residence

Answers 775
Answers
3 a categorical c Vehicle type
100
b Commercial
Frequency 90
Private
80
Car size Count % 70
60
Percentage
Small 8 40 50
Medium 9 45 40
30
1B
Large 3 15 20
10
Total 20 100 0
c
6 a 20, 55 b 5 c 20 d 55%
50 e Report: 20 schools were classified ac-
45
40 cording to school type. The majority of
35 these schools, 55%, were found to be
Frequency
30
25 government schools. Of the remaining
20 schools, 25% were independent while
15
10 20% were Catholic schools.
5 7 a 7, 45.5, 100.0
0
0 Small Medium Large b Report: When 22 students were asked
Car size the question, ‘How often do you play
4 a nominal sport’, the most frequent response was
b Place of birth ‘sometimes’, given by 45.5% of the
100
Overseas students. Of the remaining students,
90
Australia
80 31.8% of the students responded that
70
Percentage
60 they played sport ‘rarely’ while 22.7%

50 said that they played sport ‘regularly’.
40
30 8 Report: The eye colours of 11 children
20 were recorded. The majority, 54.5%, had
10
0 brown eyes. Of the remaining children,
5 a nominal 27.3% had blue eyes and 18.2% had hazel
b eyes.
Frequency
9 B
Type of vehicle Count %
Private 132 736 73
Commercial 49 109 27
Total 181 845 100

1C 776 Answers
Exercise 1C 4
11
10
1 a 9
Frequency
Answers
8
7
Frequency
Number Count %
6
0 6 30 5
1 4 20 4
3
2 3 15 2
3 3 15 1
0
4 2 10 0 200 400 600 800 1000
5 2 10 Population density
Total 20 100 5 a i 17% ii 13% iii 46% iv 33%
b i 6 ii 4
b 20%
c 15–19 words/sentence
c 0
6 a 21
2 a
Frequency b i 13 ii 8 iii 5 iv 0
Number Count % c i 4.8% ii 57.1%
2 1 2.5 7 a
3 0 0
4
4 17 42.5
5 13 32.5
count
6 9 22.5
2
Total 40 100.0
b 2.5% 0
c 4 62 66 70 74 78 82 86 90 94 98
Pulse
3 a b i 69 ii 3; 69, 70, 70
Height (cm) Frequency c
160−164 5
165−169 5 4
170−174 5
count
175−179 6
2
180−184 3
185−189 1
0
Total 25 60 70 80 90 100
b 175−179 Pulse
d 3
c 16%

Answers 777
Answers
8 a c
mode
6
20 symmetric
Frequency
count 4 15
10
2 5
1C
0
0 centre
0 1 2 3 4 5 6 7 8 9 10
children Histogram C
b 3.5, 5
d
c mode
9 20
6 15
count
Frequency
positively skewed
10
3
5
0
0 2 4 6 8 10 12
children 0
d i 2 ii 6 and 7 centre
9 a mode Histogram D
20 positively e
skewed
Frequency
15 potential mode
10 outlier 20
5
15
Frequency
0
10
centre
negatively
Histogram A 5 skewed
b 0
symmetric centre
80 Histogram E
60 f
Frequency
40 modes
80 symmetric
20
60
Frequency
0
40
centre
20
Histogram B
0
centre
Histogram F
1D 778 Answers
10 a All of the distributions are ap- b 21 years

proximately symmetric about their c approximately symmetric d 14%
respective centres. 6 a continuous
Answers
b There are no clear outliers in any of the b

Key: 1|6 = 16
distributions.
c In A the centre lies in the interval 8-10, 0 33699
in B it lies in the interval 24-26, and in 1 2267
C it lies in the interval 2 0225789
40-42. 3 15
d The spread is the lowest in B, since the 4
range is only 8, compared to 14 for A, 5 46
and 18 for C. 6
11 B 7
12 A 8
9 99
Exercise 1D
10 0
1 a discrete 7 a continuous
b b i
Key:16|5 = 16.5
4 5 6 7 8 16 579
Rooms
2 a discrete 17 0123667
b 18 245
19 39
1 2 3 4 5 6 7 8 ii
Number of children Key:16|5 = 16.5
c 4; mode is the most frequently 16 579
occurring number of children for these 17 0123
families
17 667
3 a continuous
18 24
b
18 5
19 3
70 71 72 73 74 75 76 77 78 79
Life expectancy (years) 19 9
c 75; mode is the most common life
8 a positively skewed
expectancy for these countries
b negatively skewed
4 a negatively skewed
9 a 40 b approx symmetric
b positively skewed
c 11
5
10 C
a
11 B
18 19 20 21 22 23 24 25
Age

Answers 779
Answers
Exercise 1E b M = 26
c Q1 = 17.5, Q3 = 30.5
1 a 0.4 b 1.4 c 2.4 d 3.4
d IQR = 13, R = 29
e −0.3 f −1.3 g −2.3 h −3.3
8 a positively skewed with a possible
2 a 0.0032 b 0.032 c 0.32 d 1.0
outlier at 6.
3 a 20
b M=0
Frequency
15
1E → 1F
c IQR = 1 d R=6
10
9 a M = 21 b Q1 = 10.5, Q3 = 28
5
c IQR = 17.5, R = 54
0
0 1000 2000 3000 4000 5000 6000 10 Median from 65 to less than 70, Q1 from
brain weight (gm) 60 to less than 65, Q3 from 75 to less than
The shape is positively skewed with 80.
outliers. 11 a Median in the interval 5.0-9.9.
b 8
b max IQR = 19.9
12 a n = 4, Σx = 12, x̄ = 3
Frequency
6
b n = 5, Σx = 104, x̄ = 20.8
4
c n = 7, Σx = 21, x̄ = 3
2
13 a x̄ = 3, M = 3, Mode = 2
0
–1.0 .0 1.0 2.0 3.0 4.0 b x̄ = 5, M = 5, Mode = 5
log(brain weight)
14 a i mean = 36.1 ii median = 36.0
The shape is slightly negatively
b The mean and median almost co-
skewed but closer to symmetric.
incide because the distribution is
4 a −0.4 b 3.8 c 100 g d 0.1 g approximately symmetric.
e i 5 ii 12 iii 24 15 a i mean = $3.65 ii median = $1.70
5 B 6 D b The median. The mean is inflated
because of the one large sale and not
Exercise 1F representative of the sales in general.
16 a strongly positively skewed distribution
1 a 5 b 12
b positively skewed distribution with
2 $850
outliers
3 M=1
17 a symmetric; either
4 a M=7.3 b R=6.4 b mean = 82.55 median= 82.5
5 a M = 2 b Q1 = 1, Q3 = 3 c IQR = 2 18 a IQR b range c standard deviation
d R=7
19 7.1, 0 20 b, d, f
6 a M = 11 b Q1 = 10, Q3 = 15
c IQR = 5, R = 18 21 a 20.1, 1.8 b symmetric
7 a approximately symmetric with no

outliers.

1G 780 Answers
22 e i 34 ii 30, 39 iii 9 iv 30, 55

TVs Cars Alcohol
v 55
mean 450 376 14.9
9 a i 55 ii 15
Answers
SD 100 107 5.1

b i 100 ii 20
23 C 24 C 25 C 26 B 10 a 30;
b no, 31 > 30, so inside the lower fence
Exercise 1G 11 a 25% b 75% c 25% d 50%
e 75%
1 4, 5, 6, 7, 9
12 a 25% b 25% c 50% d 25%
2 136, 148, 158, 169, 189
e 75% f 50%
3 a
13 Boxplot 1 matches histogram B, Boxplot
0 4 8 12 16 20 24 2 matches histogram D, Boxplot 3
b median: 258
matches histogram C, Boxplot 4 matches
histogram A.
135 145 155 165 175 185
14 a The distribution is negatively skewed
4 a
with no outliers. The distribution is
centred at about 42, the median value.
10 20 30 40 50 60 70 The spread of the distribution, as
b Q3 and the maximum are equal. measured by the IQR, is 15 and, as
5 b Upper fence = 26.5 + 1.5 × 26 = 50.5 measured by the range, 47.
c 55 is larger than the upper fence. b The distribution is postively skewed
d median: 21 with no outliers. The distribution is
centred at 800, the median value. The
0 5 10 15 20 25 30 35 40 45 50 55 spread of the distribution, as measured
unirate
by the IQR, is 1200 and, as measured
6 a lower fence = 7, upper fence = 95.
by the range, 3200.
b 6 and 99
15 a The distribution is negatively skewed
7 a median: 38.5
with an outlier. The distribution is
centred at 39, the median value. The
25 30 35 40 45 50 55 60 65
rtime spread of the distribution, as measured
b 25, 35, 38.5, 43, 64; 64 is a possible outlier. by the IQR, is 10 and, as measured by
8 a i 10 ii 5, 21 iii 16 iv 0, 45 the range, 45. There is an outlier at 5.
v none b The distribution is positively skewed
b i 27 ii 12, 42 iii 30 iv 5, 50 with outliers. The distribution is
v none centred at 16, the median value. The
c i 38 ii 32, 42 iii 10 iv 5, 50 spread of the distribution, as measured
v 5 by the IQR, is 6 and, as measured by
d i 16 ii 14, 21 iii 7 the range, 35. The outliers are at 5, 8,
iv 1.5, 50 v 1.5, 3, 36, 40, 50 36 and 40.

Answers 781
Answers
c The distribution is approximately 11 a 120 b 116 c 142 d 100
symmetric but with outliers. The e 72 f 50
distribution is centred at 41, the 12 $1.50
median value. The spread of the 13 101 g
distribution, as measured by the IQR, 14 mean = 3.5 kg, st dev = 0.5 kg
is 7 and, as measured by the range, 36. 15 mean = 66.0 marks, st dev = 7.7 marks
1H → 1 review
The outliers are at 10, 15, 20 and 25.
16 a 0.2 b 46.5 c 2.5% d 34%
16 The median time it takes Taj to travel
e 16% f 97.5%
to university is 70 minutes. The range
17 a i 16% ii 2.5%
is of the distribution of travel time is
b 130
60 minutes, but the interquartile range
c 133
is only 15 minutes. The distribution of
travel times is positively skewed with two 18 a i 84% ii 97.5%
outliers, unusually long travel times of b 184 cm
110 minutes and 120 minute respectively. c 144 cm
17 B 18 A 19 B 20 D d 150.4 cm
19 A 20 D 21 C 22 C
Exercise 1H 23 C
1 a 114 and 154 b 94 and 174

Chapter 1 review
c 74 and 194 d 154 e 94
f 74 g 134 Multiple-choice questions
2 a 68% b 99.7% c 16% d 2.5%
e 0.15% f 50% 1 A 2 B 3 D 4 C
5 D 6 B 7 C 8 C
3 a i 84% ii 50% iii 47.5% b 25
9 C 10 D 11 E 12 B
4 a i 99.7% ii 2.5% iii 81.5% b 800
13 C 14 B 15 A 16 B
5 a i 50% ii 34% iii 81.5% b 1994
17 A 18 D 19 D 20 A
6 a z = 1 b z = 2 c z = −1 d z = 0
21 B 22 B 23 A 24 D
e z = −3 f z = 0.5
25 B 26 D 27 C 28 B
7 a 1.0 b −1.0 c 1.4 d −1.4 29 C 30 E 31 A 32 A
8 33 E 34 B 35 E
Subject z-score Rating Written-response questions
English 2.25 Top 2.5%
Biology 3 Top 0.15% 1 a 30
Chemistry 0 Exactly average 25
Frequency
Further Maths 1.1 Top 16% 20

15
Psychology −2.25 Bottom 2.5%
10
5
9 a 2.5% b 15.85%
0
10 a 2.5% b 13.5% agree disagree don’t know
Response

2A 782 Answers
b b lower fence = 65 , upper fence = 185,

Frequency
no outliers.
Legalised? Count %
c i 68% ii 68 mmHg iii 3
Answers
Agree 18 34.6
iv 0
Disagree 26 50.0 5 a 18 cm
Don’t know 8 15.4 b 5.5%
Total 52 100.0 c
c
100
Response
90 Don’t know
80 Disagree
70 Agree
15 16 17 18 19 20 21 22
Percentage
60 hand span (cm)

50 There are two outliers at 21.0 and one
40 at 21.5.
30
20
10 Chapter 2
0
d Report: In response to the question,
‘Do you agree that the use of mar- Exercise 2A
ijuana should be legalised?’, 50% 1 a Two categorical
of the 52 students disagreed. Of the b One categorical and one numerical
remaining students, 34.6% agreed, c Two numerical d Two categorical
while 15.4% said that they didn’t 2 a EV: colour; RV: toxicity
know. b EV: diet; RV: weight loss
2 a i 50 ii 5 c EV: age; RV: price
b $105 − < $110 c 28 d 16% d EV: fuel; RV: cost
e i approximately symmetric e EV: location; RV: house price
ii $110 − < $115 iii $120 − < $125 3 a EV: age (numerical), RV: exercise level
3 a positively skewed. (categorical)
b M = 2.65 kg b EV: years of education (numerical),
c IQR = 1.25 kg RV: salary level (numerical)
d 15.6% c EV: temperature (numerical), RV:
e No, it is less than the upper fence comfort level (categorical)
(5.33 kg). d EV: time of year (categorical), RV:
4 a incidence of hay fever (categorical)
e EV: age group (categorical), RV:
musical taste (categorical)
f EV: state of residence (categorical),
RV: AFL team (categorical)
60 70 80 90 100 110 120 130 140 150 160 170 180
blood pressure (mmHg) 4 B 5 B 6 C

Answers 783
Answers
Exercise 2B c No, there is little difference in the
percentage of males and females who
1 a EV: gender, RV: intends to go to
are left handed, 9.0% compared to
university
9.8%.
b
5 a course b ordinal c 54.9%
Intends to go Gender d Yes; the percentage of Business stu-
2B
to university Male Female dents who exercise regularly (18.6%)
Yes 4 8 was much higher that the percentage of
No 4 4 Arts who exercise regularly (5.9%).
Total 8 12 6 a
Teacher (%)
Result Dr Evans Dr Smith
2 a EV: age group, RV: reduce university
fees? Fail 11.1 9.4
b Pass 61.1 62.5
Credit 27.8 28.1
Reduce Age group
university fees? 17-18 19-25 26 or more Total 100.0 100.0
Yes 8 6 6 b 100%
80%
No 3 3 4 Credit
60%
Result
Pass
Total 11 9 10 40%
Fail
c 20%
0%
Reduce Age group (%) Dr Evans Dr Smith
Teacher
university fees? 17-18 19-25 26 or more c There is no evidence students of Dr
Yes 72.7 66.7 60.0 Evans receive higher grades than
No 27.3 33.3 40.0 students of Dr Smith. The percentage
of students achieving each grade level
Total 100.0 100.0 100.0
is almost the same for both classes (eg.
3 a enrolment status 61.1% compared 62.5% for students
b No. The percentage of full-time and who received a Pass).
part-time students who drank alcohol 7 The data supports the contention that
is similar: 80.5% to 81.8%. This people who are satisfied with their job are
indicates that drinking behaviour is not more likely to be satisfied with their life,
related to enrolment status. with 70% of people who are satisfied with
4 a handedness their job reporting that they are satisfied
b with their life, compared to only 50% of
Gender (%)
people who are dissatisfied with their job.
Handedness Male Female
8 a EV: type of treatment, RV: treatment
Left 9.0 9.8 outcome
Right 91.0 90.2 b The data supports the contention that
Total 100.0 100.0 the special pillow is more effective
at treating snoring than the drug
ISBN 978-1-009-11041-9 © Peter Jones et al 2023 treatment, with just over
Cambridge
30% ofUniversity Press
2C 784 Answers
people who used the special pillow (34 years) admitted to the hospital
reporting they were cured, compared to was considerably higher than the
only 10% of people who used the drug median age of males (25.5 years). The
Answers
reported they were cured. variability of the ages was also higher
9 a 11.9% b 52.3% c marital status for the females (IQR = 28 years)
d ordinal compared that of the males (IQR = 13
e Yes. There are several ways that years).
this can be seen. For example, by 3 a hours online: numerical, year level:
comparing the married and widowed categorical
groups, we can see that a smaller b From this information it can be
percentage of those widowed found concluded that the median number
life exciting (33.8%) compared to of hours spent online was associated
those who were married (47.6%). with year level. The median time spent
Or: a bigger percentage of widowed online by the Year 10 students (20
people found life pretty routine (54.3% hours) was higher than the median
to 48.7%) and dull (11.9% to 3.7%) number of hours by the Year 11
compared to those who were married. students (16.5 hours). The variability
10 A 11 B 12 C of the hours spent online was lower for
the Year 10 students (IQR = 9.5 hours)
compared that of the Year 11 students
Exercise 2C
(IQR = 13 hours).
1 a EV: country of origin, RV: number of 4 a age at marriage: numerical, gender:
days away categorical
b The number of days these tourists b For this data there is an association
spend away from home was associated between age at marriage and gender.
with their country of origin. The me- The age at marriage is higher for
dian number of days spent away from men (M = 23 years) than for women
home for Japanese tourists (M = 17 (M = 20.5 years). The variability is
days) is considerably higher than also greater for the men (IQR = 12
for Australian tourists (M = 7 days). years) than for the women (IQR =
The variability for the number of 8.5 years). The distributions of age
days away is also higher for Japanese at marriage are positively skewed for
tourists (IQR = 16.5) compared to that both men and women. There are no
for Australian tourists (IQR = 10.5). outliers.
2 a age: numerical, gender: categorical 5 a pulse rate: numerical, gender:
b From this information it can be categorical
concluded that the median age of the b For this data there is an association
people admitted to the hospital during between pulse rate and gender.
this week was associated with their The pulse rates for males (M = 73
age. The median age of the females beats/min) are lower than the pulse

Answers 785
Answers
rates for women (M = 76 beats/min). 3
45
The variability is also lower for the
males (IQR = 8 beats/min) than for 30
runs
the women (IQR = 14 beats/min).
Both distributions are approximately 15
symmetric, with no outliers.
2D → 2E
6 a lifetime: numerical; price: categorical 0
5 10 15 20 25 30 35 40 45 50 55 60 65
b For this data there is an association
balls
between the lifetime of a battery 4 2.30
and its price. The lifetime of the
high price batteries (M = 51 hours) 2.25
is longer than that of the medium 2.20
diametet
price batteries (M = 35 hours), which
2.15
is in turn slightly longer than that
of the low price batteries (M = 32 2.10
hours). The variability in lifetime 2.02
increased as price decreased, from
IQR = 7 hours for the high price 0 40 80 120 160
temperature
batteries, to IQR = 12 hours for the
5 135
medium price batteries, and IQR = 17
hours for the low price batteries. All 120
number
three distributions are approximately

symmetric, with no outliers. 105
7 A 8 D 90
Exercise 2D 0 4 8 12 16 20 24
time
1 a number of seats 6 D
b numerical
c 8 aircraft Exercise 2E
d around 800 km/h
1 Note: There are no absolute right or
2
wrong answers to these questions as
36
answering them requires a degree of
max temp
personal judgment.
33 i no association ii yes, positive
iii yes, positive iv yes, positive
30 v yes, negative vi yes, negative
2 a i moderate, positive, linear association
17.5 18.5 19.5 20.5 21.5 22.5 23.5
min temp ii weak, negative, linear association
iii strong, positive, linear association

2F → 2H 786 Answers
iv no association life expectancies can be explained by

b i Those people who smoke more variation in birth rates.
tend to have a higher lung cancer d The coefficient of determination is
mortality rate. r2 = (0.818)2 = 0.669 or 66.9%; that
ii Older children tended to score is, 66.9% of the variation observed
lower on the aptitude test. in daily maximum temperature is
Answers
iii Intersections with higher levels of explained by the variability in daily

traffic volume also tended to have minimum temperatures.
higher CO levels. e The coefficient of determination is
r2 = (0.8782)2 = 0.771 or 77.1%; that
Exercise 2F is, 77.1% of the variation in the runs
scored by a batsman is explained by
1 a A: strong, positive, non-linear the variability in the number of balls
relationship with no outliers they face.
B: strong, negative, linear relationship 3 a r = 0.906 b r = −0.353
with an outlier
4 E 5 A 6 E 7 D
C: weak, negative, linear relationship
8 D
with no outliers
b A: non-linear
2 r = 0.73 Exercise 2H
3 a–c Answers given in question. Note: These answers are for guidance only.
4 a strong positive Alternative explanations for the source of an
b strong positive association may be equally acceptable as the
c moderate negative variables suggested.
1 Not necessarily. In general, older
Exercise 2G
children are taller and have been learning
1 a 45.6% b 11.9% c 32.1% d 45.3% mathematics longer. Therefore they tend
e 1.5% to do better on mathematics tests. Age
2 a The coefficient of determination is is the probable common cause for this
r2 = (−0.611)2 = 0.373 or 37.3%; that association.
is, 37.3% of the variation observed in 2 Not necessarily. While one possible
hearing test scores can be explained by explanation is that religion is encouraging
variation in age. people to drink, a better explanation
b The coefficient of determination is might be that towns with large numbers of
r2 = (0.716)2 = 0.513 or 51.3%; that churches also have large populations, thus
is, 51.3% of the variation observed in explaining the larger amount of alcohol
mortality rates can be explained by consumed. Town size is the probable
variation in smoking rates. common cause for this association.
c The coefficient of determination is 3 Probably not. The amount of ice-cream
r = (−0.807) = 0.651 or 65.1%; that
2 2
consumed and the number of drownings
is, 65.1% of the variation observed in would both be affected by weather

Answers 787
Answers
conditions. Weather conditions are the Written-response questions
probable common cause. 1 a Number of accidents and age; both
4 Maybe but not necessarily. Bigger categorical variables
hospitals tend to treat more people with b RV: Number of accidents; EV: age
serious illnesses and these require longer
c 470
hospital stays. A common cause could be
d
No of accidents < 30 ≥ 30
2I → 2 review
the type of patients treated at the hospital.
5 Not necessarily. Possible confounding At most one 21.7% 42.5%
variables include age and diet. More than one 78.3% 57.5%
6 There is no logical link between eating e The statement is correct. Of drivers
cheese and becoming tangled in bed aged less than 30, 78.3% had more
sheets and dying. The correlation is than one accident compared to only
probably spurious and the result of 57.5% of drivers in the older category.
coincidence. 2 a Numerical: conversation test score.
7 Not necessarily. For example, the more Categorical: completed weeks of
serious the fire, the more fire trucks course
in attendance and the greater the fire b There is an association between the
damage. A possible common cause is the students’ scores on the conversation
severity of the fire. test, and the number of weeks of
8 E the course they have completed.
The median score at the beginning
Exercise 2I
of the course (M=38) showed a
1 a segmented bar chart little improvement after six weeks
b scatterplot c parallel box plots (M =42), followed by a very large
d scatterplot e scatterplot improvement by the end of the 12
f segmented bar chart week course (M =72). The variability
g segmented bar chart of the scores changed little over the
h parallel box plots or back-to-back stem plots course (IQR=12 at the beginning,
2 E IQR = 12 at 6 weeks, IQR = 14 at 12
3 D weeks). The distributions of scores at
0 weeks is approximately symmetric
with an outlier at 66, positively skewed
Chapter 2 review with an outlier at 76 at 6 weeks, and
Multiple-choice questions approximately symmetric with no
outliers at 12 weeks.
1 A 2 D 3 B 4 D
3 a rate is the response variable,
5 E 6 B 7 D 8 A
experience is the explanatory variable.
9 C 10 E 11 D 12 E
13 C 14 C 15 C 16 D
17 A 18 E 19 B 20 C
21 E 22 C

3A → 3B 788 Answers
b 34 2 C
3 The data is numerical; the association is
32
linear; there are no clear outliers.
30 4 a x
rate ($/hr)
28 b y = 9.23 + 1.00x
Answers
5 a RV: pollution level; EV: traffic volume

26
b pollution level = −330 + 49 × traffic
24 volume
22 6 a RV: life expectancy; EV: birth rate
0 1 2 3 4 5 6 7 8 9 10 11 12 b life expectancy = 110 − 1.5 × birth rate
experience (years)
c There is a moderate positive linear 7 a y b r = −0.450
relationship; that is, people with more 8 a RV: distance travelled; EV: age of car
experience are generally being paid a b r = 0.947
higher hourly pay rate. 9 a r is also negative.
d 0.786. Note that this value is on the b Slope is zero: regression line is horizontal.
borderline between a moderate and c intercept = ȳ (mean of RV)
a strong linear relationship, but with 10–11 Answers given in question.
such a small data set it is difficult to 12 a Answers given in question.
assess the strength from the scatterplot. b runs = −2.6 + 0.73 × balls f aced
e Coefficient of determination = 0.618; 13 a RV: number of TVs
that is, 61.8% of variation in pay
b Answer given in question.
rate is explained by the variation in
c number of T V s =
experience.
61.2 + 0.930 × number of cars
4 a 60%
14 C 15 A 16 C
b For these people there is an association
between the person’s quality of sleep
Exercise 3B
and their participation in the course,
with 85% of people rating their sleep 1 mark = 80 − 4.3 × days absent
quality as good after the course, 2 a 2.9: On average a person who is 0 cm
compared to only 60% of people rating in height has a hand span of 2.9 cm -
their sleep quality as good before the not sensible!
course. b 0.33: On average a person’s hand span
increases by 0.33 cm for each 1 cm
increase in height.
Chapter 3 3 a 575: On average, the company will
have $575 in sales when their online
advertising expenditure is $0.
Exercise 3A
b 4.85: On average sales will increase by
1 A residual is the difference between a $4.85 for each additional $1.00 spent
data value and its value predicted by a on online advertising.
regression line.
Answers 789
Answers
4 a 80 cm, extrapolating e 49.7%: 49.7% of the variation in
b 92 cm, interpolating success rate in putting is explained by
c 98 cm, extrapolating the variation in the distance the golfer
5 a $487.50, extrapolating is from the hole.
b $1023.50, interpolating 14 a yes, linear relationship
c $1224.50, extrapolating b 0.9351 or 93.5%
3B
6 a 173 cm, reliable, interpolating c 93.5%
b 189 cm, unreliable, extrapolating d pay rate = 18.56 + 0.289 × experience
c 165 cm, reliable, interpolating e Intercept = $18.56. On average, a

worker with no experience will earn
7 a 20.3%
$18.56 per hour.
b 42.3%
f On average, the pay rate increases by
c The number of hours is more im-
29 cents per hour for each additional
portant as it explains 42.3% of the
one year of experience.
variation in exam score, much more
g i $20.87 ii $0.33
than IQ which explains only 20.3% of
the variation in exam score. h yes; no clear pattern in the residual
plot
8 Answers given in question.
15 a r = −0.608
9 a 9.7 b −0.8
b 37% of the variation in the hearing test
10 a 2 b −1 c 2
score is explained by the variation in
11 A: clear curved pattern in the residuals
age.
(not random), C: curved pattern in the
c score = 4.9 − 0.043 × age
residuals (not random).
d −0.043; the hearing test score, on
12 a 27.8: On average a packet of chips
average, decreases by 0.043 for each
with 0 gm of fat contains 27.8 calories.
one additional year of age.
b 14.7: On average, the calorie content
e i 4.04 ii −2.04
increases by 14.7 for each one
f i 0.3 ii −0.4
additional gram of fat included.
g yes; no clear pattern in the residual
c 75.7% of the variation in calorie
plot
content of the chips is explained by the
16 negative, drug dose,
variation in fat content.
−0.9492; 55.9; −9.31; decreases,
d 145.4 e −13.4
9.31; 55.9; 90.1, response time,
13 a −0.278: On average, for each addi-
drug dose; clear pattern
tional one metre the golfer is from
17 The scatterplot shows that there is
the hole the success rate decreases by
a strong positive linear relationship
27.8%.
between radial length and femur length:
b 73.5
r = 0.9876. There are no outliers.
c 3.54 m The equation of the least squares
d −0.705 regression line is:
radial length = −7.2 + 0.74 ×
f emur length
3C 790 Answers
The slope of the regression line predicts d

3.0
that, on average, radial length increased
1.5
residual
by 0.74 cm for each one centimetre
Answers
increase in femur length. 0.0

The coefficient of determination indicates −1.5
that 97.5% of the variation in radial
−2.0
lengths can be explained by the variation
8 9 10 11 12 13 14 15 16 17 18 19
in femur lengths. test a
The residual plot shows no clear pattern, 2 a
supporting the assumption that the
relationship between radial and femur 18
length is linear.
15
score
18 E 19 B 20 A
12
Exercise 3C
9
1 a
0 1 2 3 4 5 6 7 8
errors
17
b score = 17.5 − 1.08 × error,
r = −0.841, r2 = 0.707
test b
14
c
11
18
8 9 10 11 12 13 14 15 16 17 18 19
y = 17.5 + 1.08333x
test a 15
score
b test B score = 4.2 + 0.72 ×

test A score, r = 0.78, r2 = 0.61 12
c
9
0 1 2 3 4 5 6 7 8
17 errors
d
test b
14
1.5
13
residual
0.0
8 9 10 11 12 13 14 15 16 17 18 19
test a −1.5
−2.0
0 1 2 3 4 5 6 7 8
errors

Answers 791
Answers
3 a RV: adult weight; EV: birth weight 17 C 18 A 19 A 20 D
b 21 C 22 C 23 D
60
adult weight
56
1 a i 5 years
50
ii mean= 767, st dev = 35
3 review
1.0 2.2 2.6 3.0 3.4 3.6 4.2
birth weight b airspeed = 673 + 0.372 × number of
c i strong positive linear association seats
with no outliers c 74.1%
ii approximately 0.9 2 a days of rain b −6.88, 2850 c 2024
d adult weight = d decrease, 6.88 e −0.696
38.4 + 5.87 × birth weight, f 48.4, days of rain g i 1873 ii −483
r2 = 0.765, r = 0.875 h interpolation
e 76.5% of the variation in the adult 3 a cost
weight is explained by the variation in b There is a strong, positive, linear
birth weight. association between the cost of
f On average, adult weight increases by the meals and the number of meals
5.9 kg for each additional kilogram of prepared.
birth weight. c i $307.30 ii extrapolating
g i 56.0 ii 53.1 iii 61.3 d i 222.48: the fixed costs of preparing
h Yes. 76.5% of the variation in the adult meals is $222.48.
weight is explained by the variation in ii $4.039: The slope of the regression
birth weight. line predicts that, on average,
i 3.0 meal preparation costs increase by
1.0 $4.039 for each additional meal
residual
0.0
produced.
e Answer given in question.
−1.5
4 a RV: height; EV: femur length
−3.0
b height = 36.3 + 5.35 × f emur length
1.0 2.2 2.6 3.0 3.4 3.6 4.2
birth weight c On average, height increases by 5.35
The lack of a clear pattern in the cm for each cm increase in femur
residual plot supports the assumption length.
that the association between adult d r2 = 0.988; that is, 98.8% of the
weight and birth weight is linear. variation in height is explained by the
variation in femur length.
Chapter 3 review
e 97.6%
5 a RV: height; EV: age
1 C 2 D 3 A 4 C b strong positive association with no
5 E 6 C 7 B 8 B outliers.
9 D 10 A 11 A 12 A c Answer given in question.
13 D 14 E 15 A 16 C d i Answer given in question
ii Extrapolating 4 a ±4.7 b ±1.3 c 6 d −8

e On average, for each additional one 5 a y2
year of age, height increases by 6.4 30.0
cm.
20.0
f r2 = 0.995; that is, 99.5% of the
variation in height is explained by the 10.0
Answers
variation in age. 0 x
0
0
0
0
0
.0
g i 140.3 cm ii −0.7 cm
2.
4.
6.
8.
10
h i Answer given in question. b y2 = 1.5 + 3.1x
ii Residuals show a clear curved pattern. c y = ±5.4, but only the positive solution
6 a moderate, positive linear association applies here because the model is only
with no outliers defined for y > 0.
b i 142 6 a number of people = 0.0 + 4.1 ×
ii extrapolating diameter2
c −6.3 b 7
d i linearity 7 a time2 = 18 − 9.3 × amount b 3.8 min
ii the lack of a clear pattern in the 8 D 9 A 10 B
residual plot supports the linearity
assumption. Exercise 4B
1 a 6.6 b 1.1 c −3.1 d 138.5

Chapter 4 2 a y
Exercise 4A 10
9
8
1 a 19.5 b 11.7 c 23.8 d 126.7 7
6
2 a y 5
4
20 3 log x
50
1. 0
50
00
50
00
0
15
0.
1.
2.
2.
3.
b y = 1 + 3 log x c 7
10
3 a y
5 16
0 x2 14
0 5 10 15 20 12
b y = 16 − x2 10
8
c when x = −2, y = 12 6
4 log x
3 a y
1.0 1.5 2.0 2.5 3.0
60 b y = 20 − 5 log x c 5
50 4 a 100 b 218.8 c 1 000 000
40
30 d 0.8
20
10
0 x2
0 5 10 15 20 25
b y = 1 + 2x2
c when x = 6, y = 73
Answers 793
690
Answers
5 a log y 5 a horsepower = 22.1 +
consumption
2.0 b 99 to nearest whole number
1.8 1
6 a = −0.00024 + 0.050 × times
errors
1.6
to 2 sig. figs
1.4
b 3 to nearest whole number
1.2 x
4C → 4D
0.1 0.2 0.3 0.4 0.5 7 A 8 E 9 C
b log y = 1 + 2x c 158.5
6 a level = 1.8 + 2.6 log(time) to 2 sig. figs Exercise 4D
b 2.8 to 1 d.p. 1 1
1 a log y, , log x,
7 a log(number) = 1.314 + 0.08301 × y x
b None; trend needs to be consistently
month to 4 sig. figs
increasing or decreasing.
b 139 to nearest whole number 1
c log y , , x2
8 C 9 E 10 A y
2 2
d x ,y
Exercise 4C 2 a
2500
1 a 13.3 b 2.8 c 4.9 d 309.5
2 a y 2000
potato yield(kg)
60 1500
40
1000
20
500
1
0
x
0 0.1 0.2 0.30.40.5 0
120 5 10 15 20 25 30 35
b y= c 24 plot length (metres)
x b yield = −620.0 + 80.23 × length
3 a 0.17 b 0.07 c 0.16 d 0.06
c
4 a 1
y 400
300
5 200
residual
4 100
0
3
–100
2 –200
1 –300
5 10 15 20 25 30 35
0 x plot length (metres)
0 1 2 3 4 5
linear
No, the residuals show a clear curved
1
b =x pattern.
y
c 4
1
d log y , , x2
y
e yield = 3.983 + 2.030 × (length)2

4 review 794 Answers
f r2 = 97.5% 4 a
3 a 350
18
300
density (people/hectare)
16 250
smoking (cigarettes/day)
Answers
200
14
150
12
100
10 50
0 2 4 6 8 10 12 14
distance (km)
8 b density = 345.3 − 18.65 × distance
.60 .70 .80 .90 1.00 1.10 1.20 1.30 1.40
cost ($/cigarette) c
b smoking = 22.49 − 9.501 × cost 40
c
.50 20
residual
.25
0
residual
.00
–20
–25
–40
0 2 4 6 8 10 12 14
–50
distance (km)
No, the residuals show a clear curved
.60 .80 1.00 1.20 1.40 pattern.
cost ($/cigarette)
No, the residuals show a clear curved d x2 , y2
pattern. e density = 308.9 − 1.345 × (distance)2
1 1
d log x , log y , , f r2 = 99.1%
y x
1
e Either the log x and could be
x Chapter 4 review
recommended as both transformations
give very good results. That is
9.045
smoking = 3.420 + or 1 A 2 D 3 D 4 B
cost
smoking = 12.73 − 21.90 × log(cost) 5 A 6 B 7 E 8 D
1 9 D 10 D 11 D
The transformation is more intuitive
x
and easier to interpret. Written-response questions
1 1
f : r2 = 99.3% 1 a = 2.606 − 1.053 × length
x age
log x: r2 = 99.6% b 2.6 years
2 a literacy rate = −44.2 + 33.3 log (GDP)

Answers 795
Answers
b 4 a
30 120
20
100
10
residual
80
mortality
0
–10 60
4 review
–20
40
–30
2.0 2.5 3.0 3.5 4.0 4.5 20
log(GDP)
0
Residual plot shows no clear pattern 0 50 100 150 200 250 300 350 400
doctors
c 89% b
d 0.077 120
3 a 180
160 100
Distance (metres)
140
80
120
mortality
100
60
80
60 40
40
20 20
0
1 2 3 4 5 6 0
Time (seconds) .000 .005 .010 .015 .020 .025 .030 .035 .040
1/doctors
b 1
Time 0 1 2 3 4 5 6
c mortality = −1.194 + 3856 ×
Distance 0 5.2 18 42 79 128 168 doctors
d
2
Time 0 1 4 9 16 25 36 60
c 200 40
150
distance
residual
20
100
50 0
0 –20
0 5 10 15 20 25 30 35 40
(time)2 –40
.00 .01 .02 .03 .04
d distance = 0.45 + 4.8 × time2
1/(doctors)
e 236 metres The residual plot shows no clear struc-
f 7.5 ture indicating that the assumption of
5.0
2.5 linearity is justified.
residual
0
–2.5 e r2 =82.8%
–5.0 f 37
0 5 10 15 20 25 30 35 40
(time)2 g Since 100 is within the range of the
The residual plot shows no clear structure data we are interpolating, and the
indicating that the assumption of linearity prediction is reliable.
is justified.

796 Answers
Chapter 5
5A
5
Feature A B C
Irreg fluct X X X
Answers
Exercise 5A
Inc trend X
1 Dec trend X
70
Cycles X
60 Seasonality X X
50
6 Feature A B C
40
Sales
Irreg fluct X X
30
Struct change X
20 Inc trend X
10 Dec trend X
0 Seasonality X
2014 2015 2016 2017 2018 2019 2020 2021 2022
Year
7 Feature A B C
2 800
700 Irreg fluct X X X

600 Struct change X
Number of penguins
500
Inc trend X
400
300
Dec trend
200 Outlier X
100
8 The number of mobile phones per 100
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec people increases rapidly over the years
Month
3 25 2000-2008. The number continues to
24 increase from 2009 until 2019, but the
23
increase in the number of phones is at
22
much lower rate than in the preceding
Temperature
21
20 years.
19
9 a
18
Population of Australia (millions)
17 26
16
15 25
Mon Tue Wed Thu Fri Sat Sun
Day
4 24
Feature A B C
Irreg fluct X X X 23
Inc X
22
Dec X 2010 2012 2014 2016 2018 2020 2022
Year

Answers 797
Answers
b The plot shows a steady increase in the 1945 to 1975 but then decreased
population of Australia over the years at a similar rate to males over the
2012 - 2021. period 1975–1992.
10 a ii The difference in smoking rates
600 between males and females has
Theft rate(per 100,000 cars)
500 decreased over these years.
5B
400 b i
28.0
300
26.0
200 24.0
100 22.0
Smoking(%)
20.0
0
2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 18.0
Year
16.0
b The plot shows a steady decline in
14.0
the number of vehicle thefts over the
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
years 2003 -2010, after which the Year
number of vehicle thefts has remained ii Whilst both plots show irregular
reasonably steady, showing only fluctuation, overall the percentages
irregular variation. of male and females who smoke
have declined substantially over the
11 The number of cases of measles show an
years 2000-2018.
increasing trend between 1989 and 1992.
In 1993-1994 there is a rapid increase in iii The difference in smoking rates
the number of measles cases, followed between males and females has
be a rapid decrease in 1994-1995. The remained almost the same over
number of cases continued to decrease these years.
until 2000, since then have remained low, 14 E
showing only irregular variation over the 15 D
years 2001-2019.
12 The number of overseas arrivals (millions Exercise 5B
people per month) in Australia increased
1 a i 3 ii 1 iii 4
steadily from November 2011 until April
b i 3.2 ii 1.2 iii 2.2
2020. The number of arrivals is clearly
c i 2.6 ii 2.0
seasonal, with the peak time for arrivals
in the January quarter each year. The d 2.3
number of arrivals dropped suddenly to 2 a 24.4 b 20.0 c 23.2
almost zero in April 2020, and remained 3
at this level until October 2021. t 1 2 3 4 5 6 7 8 9
13 a i The percentage of males who
y 10 12 8 4 12 8 10 18 2
smoke has consistently decreased
3-mean − 10 8 8 8 10 12 10 −
since 1945, while the percentage of
females who smoke increased from 5-mean − − 9.2 8.8 8.4 10.4 10 − −

5B 798 Answers
4 a, c 45 • 3-moving mean exchange rate

40
35 • 5-moving mean exchange rate
Temperature
30 The exchange rate has a downward
Answers
25
20
trend over the 10-day period. This is
15 most obvious from the smoothed plots,
10
5
particularly the 5-moving mean plot.
0 b
0 1 2 3 4 5 6 7 8 9 10
Day 3-moving 5-moving
Raw data Day Exchange rate mean mean
3-mean smoothed
5-mean smoothed 1 0.743 − −
The smoothed plots show that the 2 0.754 0.745 −
‘average’ maximum temperature 3 0.737 0.747 0.742
changes relatively slowly over the 4 0.751 0.737 0.738
10-day period (the 5-day average varies 5 0.724 0.733 0.730
by only 5◦ ) when compared to the daily 6 0.724 0.720 0.729
maximum, which can vary quite widely 7 0.712 0.724 0.722
(for example, nearly 20◦ between the 8 0.735 0.721 0.720
fourth and fifth day) over the same
9 0.716 0.721 −
period of time.
10 0.711 − −
b
3-moving 5-moving 6 a 3.8 b 2.0
◦
Day Temp. ( C) mean mean 7 a 3.3 b 1.5 c 2.4 d 1.9
1 24 − − 8 a 13.1 b 12.2 c 10.7
2 27 26.3 − 9 a, c
3 28 31.7 28.2 25
20
4 40 30.0 28.0
15
5 22 28.3 27.0
10
6 23 22.3 25.6 5
7 22 22.0 22.6 0
Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec
8 21 22.7 23.4 Number of complaints 2-moving mean
9 25 24.0 − Two-mean smoothing of the plot has

10 26 − − not had a big effect, with both plots
showing that the that the number of
5 a, c
complaints between April and July is
0.76
0.75 considerably higher than the number in
0.74 the rest of the year.
0.73
0.72
0.71
0 1 2 3 4 5 6 7 8 9 10
Day
• exchange rate
Answers 799
Answers
10 a, c 6
90
Number of whales (000S)

80 70
70 60
60
50 50
40 40
30
20 30
10 20
0
Apr May Jun Jul Aug Sep Oct Nov Dec 10
rainfall 4-moving mean
5C
0
20
30
40
60
70
80
50
Four-mean smoothing of the plot shows
19
19
19
19
19
19
19
a steady increase in rainfall from June to Year
October. Raw data

3-median smoothing
11 A 12 C 13 E 14 D 5-median smoothing
7 a 4 b i, ii
Exercise 5C 6
5
Growth in GDP (%)

1 a (3, 3) b (2, 2) c (3, 2) d (3, 3) 4
◦ ◦ 3
2 a 30 C b 25 C
2
3 30 1
28 0
-1
26
-2
24
ar 0
Ye 11
Ye 12
13
Ye r 1
Ye r 3
Ye r 5
Ye ar 9
Ye 2
Ye 4
Ye 6
Ye r 7
Ye r 8
Ye r 1
ar
ar
ar
carsale($millions)
a
a
ar
ar
a
Ye
22
20
GDP growth
18 3-median smoothing
16
5-median smoothing
14 c The plot of GDP growth over 1 year,

12 shows a great deal of variability,
10 with no clear trend apparent. When
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Year smoothed over a 3-year period, GDP
The smoothed plot shows that sales were growth is still variable but to a lesser
quite consistent up to 2017 when they extent. No clear trend is apparent, but
dropped, and have remained at this lower GDP appears to be going through a
level from 2017-2022. period of below average growth during
4 a 25◦ C b 25◦ C the time period from Year 7 to Year 9.
5 0.76 When smoothed over a 5-year period,
0.75
Exchange rate
GDP growth is much less variable

0.74
0.73
but clearly shows the period of below
0.72 average growth during the from period
0.71 from Year 7 to Year 9.
0.70
0 1 2 3 4 5 6 7 8 9 10 11 8 D 9 C 10 D
Day
The smoothed plot shows that there was a
general decreasing trend in the exchange
rate over this period.
5D → 5E 800 Answers
Exercise 5D 11 b
Number of cars (deseasonalised)

200
1 a 1.0 190
180
170
b Sales in February are typically 30% 160
150
higher than sales in the average month. 140
130
120
c Sales in September are typically 10% 110
Answers
100
lower than sales in the average month. Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun
Day
2 a 7.8 b 6.7 12 a, d
250
3 a 3.9 b 6.9 240
230
4 a Increase by 42.9%. 220
210
b Decrease by 23.1% 200
190
5 a 1.2 b 1514 c 1437 d 1005 180
170
160
6 Sum Aut Win Spr 150
1 2 3 4 5 6 7 8 9 10 11 12
Quarter
Number of students: 56 125 126 96 Jobs vacancies Deseasonalised job vacancies
Deseasonalised: 112 125 97 80 b Q1 Q2 Q3 Q4

Seasonal index: 0.5 1.0 1.3 1.2
1.03 0.93 0.93 1.11
7 a, c Q1 Q2 Q3 Q4
c Year Q1 Q2 Q3 Q4
Deseasonalised: 152 142 148 153
Seasonal index: 1.30 1.02 0.58 1.10 1 206 209 211 205
b In quarter 1 the restaurant chain 2 214 212 211 215
employs 30% more waiters than 3 224 220 218 221
the number employed in an average
13 E 14 D 15 E 16 B
quarter.
17 B
8 Q1 Q2 Q3 Q4
0.89 0.83 1.12 1.16 Exercise 5E
9 Jan Feb Mar April May June 1 a There was a general increasing trend
0.89 0.96 1.04 1.26 1.33 1.11 in the number of university students in
July Aug Sept Oct Nov Dec Australia during the period 2010–2019
0.67 0.74 0.59 0.81 1.11 1.48 b students(000s) = −37563 + 18.927 ×
year. On average, the number of
10 Jan Feb Mar April May June university students in Australia has
1.06 0.96 1.18 1.18 1.03 0.86 increased by 18 927 per year.
July Aug Sept Oct Nov Dec c 859 000 (to nearest thousand)
0.96 0.79 0.74 0.54 1.18 1.50 2 a

13
11 a Mon Tues Wed Thur Fri Sat Sun
12
Sales
11
155 157 150 134 153 134 150
10
150 154 190 148 143 150 157 9
1 2 3 4 5 6 7 8 9 10 11 12
Year

Answers 801
Answers
b There was a general decreasing trend The deseasonalised sales appear to
in the percentage of retail sales made show an increasing trend over time.
in department stores. c deseasonalised sales
c sales = 12.5 − 0.258 × year = 80.8 + 23.5 × quarter
The percentage of total retail sales d forecasted actual sales
that are made in department stores is = 386.3 × 1.13 = 437
5 review
decreasing by 0.258% per year. 7 C 8 E
d 8.6%
3 a age = −147 + 0.0882 × year; On Chapter 5 review
average, the average age of mothers
increased by 0.0882 years (equivalent Multiple-choice questions
to 1 month) each year between 2010
1 E 2 E 3 A 4 B 5 E 6 C
and 2020.
7 D 8 D 9 C 10 D 11 B 12 C
b 32.0 years; Unreliable as we are ex-
13 B 14 A 15 E 16 D 17 D 18 A
trapolating 10 years beyond the period
19 B 20 E
in which the data were collected.
4 a earnings = −83 280 + 42.07 × year;
On average, average weekly earnings
increased by $42.07 each year between 1 a 18.5
18.0
2014 and 2021.
Carbon dioxide emissions
17.5
b $2122.10; Unreliable as we are extrap- 17.0
olating 9 years beyond the period in 16.5
which the data were collected. 16.0
15.5
5 a deseasonalised number
15.0
= 50.9 + 1.59 × quarter number 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Year
b deseasonalised number = 76.34 b Carbon dioxide emissions decreased
reseasonalised (actual) number = 90 between 2009 and 2014, then remained
(to the nearest whole number) reasonable steady over the years
6 a Year Q1 Q2 Q3 Q4 2014–2017, showing only irregular
1 122 128 118 130 fluctuations. Between 2017 and 2018
there was a small decrease in carbon
2 250 245 263 236
dioxide emissions.
b 400
sales c CO2 emissions = 612.0 − 0.2958 × year
300 d 12.7
Sales
200 deseasonalised e Unreliable as we are extrapolating 8

sales years beyond the period in which the
100
data were collected.
0 2 a i inflation= 332 − 0.164 × year
0 1 2 3 4 5 6 7 8
Quarter b i inflation= 339 − 0.167 × year

6A → 6E 802 Answers
a ii, b ii 9 D 10 A 11 E 12 E
6 13 B 14 A 15 E 16 C
5 17 D
inflation(%) 4
Exercise 6D
3
1 A 2 A 3 A 4 B
Answers
2
5 E 6 B 7 A 8 B
1
9 D 10 E 11 D 12 E
0
2009 2011 2013 2015 2017 2019 13 E 14 B 15 B 16 D
year
17 A
c The trend lines are parallel. As such,
they will never cross, so the inflation Exercise 6E
rate for China will remain higher than 1 a age, distance
the inflation rate for Australia. b mean = 7.17 km, sd = 3.46 km
d 1.7 c z = 1.7
3 a Sum Aut Win Spr d
Gender
SI 0.29 0.36 1.37 1.98
b Study mode Female Male
Sum Aut Win Spr On campus 3 3
Deseas 269 239 255 273 Online 4 2
Total 7 5
Chapter 6 e i 60%
e ii Yes, there is an association between
Exercise 6A
study mode preference and course.
A higher percentage of students
1 A 2 B 3 B 4 B business chose to study online
5 D 6 B 7 D 8 E (60%), compared to only 36% for
9 E 10 E 11 D 12 A both students of Health and Social
13 D 14 A 15 A 16 E Science.
17 C 18 B 19 B 20 B 2 a The distribution of distance is pos-
21 C 22 E 23 D 24 C tively skewed, with outliers at 17 km,
25 A 26 D 18 km, and 19 km.
Exercise 6B b 30
c i Lower fence = -2, Upper fence =
1 E 2 A 3 E 4 C
14.
5 C 6 B 7 E 8 B
ii A distance of 1 km is within the
9 B 10 E 11 B 12 E
fences.
13 D 14 B 15 E 16 D
d i 1 km ii 1.5 km
17 A 18 D
3 a On average, height increases by 0.815
Exercise 6C
cm for each additional 1 cm increase in
1 C 2 C 3 A 4 D
arm span.
5 B 6 B 7 E 8 B

Answers 803
Answers
b i Females: r2 = 64.6% c i Slope = $1525.80. On average the
ii Males: r2 = 69.9% price of bitcoin is increasing by
iii Since the value of the coefficient of $1525.80 each month.
determination for males (69.9%) is ii $92 837
higher than the value for females iii Unreliable as we are extrapolating
(64.6%), then we can say that arm several years beyond the period in
7A
span is a better predictor of height which the data were collected.
for males than for females.
c i The models predict that when both Chapter 7
have arm span measurements of
160 cm, a male will be 1.8 cm Exercise 7A
taller than a female.
1 a 2, 8, 14, 20, 26 b 5, 2, −1, −4, −7
ii The models predict that when both c 1, 4, 16, 64, 256 d 64, 32, 16, 8, 4
have arm span measurements of 2 a 6, 14, 30, 62, 126
190 cm, a female will be 4.6 cm b 24, 16, 12, 10, 9 c 1, 2, 5, 14, 41
taller than a male. d 124, 60, 28, 12, 4
iii The differences predicted are 3 a 4, 6, 8, 10, 12 b 24, 20, 16, 12, 8
not reliable for a height of 160 c 2, 6, 18, 54, 162 d 50, 10, 2, 0.4, 0.08
cm as this value is outside the e 5, 13, 29, 61, 125
range of height data for males. f 18, 16.4, 15.12, 14.096, 13.2768
The prediction is not reliable for a 4 a 2, 5, 8, 11, 14 b 50, 45, 40, 35, 30
height of 190 cm as this value is c 1, 3, 9, 27, 81 d 3, −6, 12, −24, 48
outside the range of height data for e 5, 9, 17, 33, 65 f 2, 7, 17, 37, 77
females. g −2, −1, 2, 11, 38
4 a There is a moderate strength, non- h −10, 35, −100, 305, −910
linear association between expenditure 5 a 12, 57, 327, 1947, 11667
and score. b 20, 85, 280, 865, 2620
1
b y2 , log x, c 2, 11, 47, 191, 764
x
c i The linearity assumption. d 64, 15, 2.75, −0.3125, −1.078125
ii No, there is a clear structure in e 48000, 45000, 42000, 39000, 36000
the residual plot. If the linearity f 25000, 21950, 19205, 16734.50, 14511.05
assumption had been met the resid- 6 a A2 = 6 b B4 = 2 c C3 = 27
uals would have been randomly d D5 = 95
scattered around a horizontal line at 7 a V0 = 4, Vn+1 = Vn + 2
y = 0. b V0 = 24, Vn+1 = Vn − 4
d i score = 12.99 + 120.6 × c V0 = 2, Vn+1 = 3Vn
log(expenditure) 8 a V0 = 5, Vn+1 = Vn + 5
ii 495 b V0 = 13, Vn+1 = Vn − 4
5 a i $12 000 c V0 = 1, Vn+1 = 4Vn
ii $11 000 d V0 = 64, Vn+1 = 0.5Vn
b $52 208.29
804 Answers
8
7B → 7C
9 5 8 a × 2800 = 244 b 2800

100
10 0 11 E 12 D 13 E
c W0 = 2800, Wn+1 = Wn − 224
9 a $2100, $1700, $1300 b 4
Exercise 7B 10 a $22 195, $21 390, $20 585 b 17
11 a i $1500 ii $102 iii 6.8%
1 a Vn
Answers
10 b $684
c 14 years
8
12 a $32 600
6 b M0 = 32 600, Mn+1 = Mn − 10
13 a 450, 449.95, 449.90, 449.85, 449.80
4
b $449
2 14 a $47 800, $47 600, $47 400
n
b $45 000 c 25 000 km
0 1 2 3 15 a 7200 b 72 000 c $720 d 10%
b Vn 16 C 17 C 18 B
50
Exercise 7C
40
1 a An = 4 + 2n, A20 = 44
30 b An = 10 − 3n, A20 = −50
20 c An = 5 + 8n, A20 = 165
d An = 300 − 18n, A20 = −60
10
2 a 5000 b $270 c Vn = 5000 − 270n
n d $7430
0 1 2 3
3 a 12 000 b $864
2 a V0 = 8000 b $320
c Vn = 12 000 + 864n d $19 776
c V0 = 8000, Vn+1 = Vn + 320
4 a $8000 b $512
3 a H0 = 41 000 b $2542
c i $14 144 ii 16 years
c H0 = 41 000, Hn+1 = Hn + 2542
5 a $2000 b $70
4 a V0 = 2000 b 14 years
c i $2420 ii 29 years
V1 = 2000 + 76 = 2076 6 a $5600 b $1260 c Vn = 5600 − 1260n
V2 = 2076 + 76 = 2152 d $1820
V3 = 2152 + 76 = 2228 7 a $7000 b $1225 c Vn = 7000 − 1225n
5 a $7518, $8036, $8554 b 6 years d 5 years
6 a i $15 000 ii $525 iii 3.5% 8 a $1700 b $212.50 c $850
b 29 years d $212.50 e 8 years
7 a $12 300 9 a $65 000 b $3250 c 5%
b C0 = 82 000, Cn+1 = Cn − 12 300 d $42 250 e 11 years
10 a $29 000 b $0.25 (25 cents)
c $24 000 d 96 000 km

Answers 805
Answers
11 a $9700 b $0.388 per km 2 a V0 = 6000
c Vn = 35 400 − 0.388n V1 = 1.042 × 6000 = 6252
d 74 000 km
V1 = 1.042 × 6252 = 6514.58
12 a i $0.026875 ii $69 687.50
V2 = 1.042 × 6514.58 = 6788.20
iii $20 156.25
b 7 years
b $9218.75
3 a V0 = 20 000
7D → 7E
c 1 486 400
V1 = 1.063 × 20 000 = 21 260
13 D 14 C
V1 = 1.063 × 21 260 = 22 599.38
Exercise 7D V2 = 1.063 × 22 599.38 = 24 023.14
b 7 years
1 a 2, 4, 8, 16, 32 4 a $5000 b 1.068
Vn
40
c V0 = 5000, Vn+1 = 1.068 × Vn
35 d $6947.46 e $1947.46
30
5 a $18 000 b 1.094
25
20 c V0 = 18 000, Vn+1 = 1.094Vn
15 d $25 783.50 e 4 years
10
5 6 V0 = 9800, Vn+1 = 0.965Vn
0
n 7 M0 = 28 600, Mn+1 = 0.926Mn
1 2 3 4
8 a V0 = 18 000, Vn+1 = 0.955Vn
b 3, 9, 27, 81, 243
b $17 190, $16 416.45, $15 677.71,
Vn
250 $14 972.21, $14 298.46
c $15 677.71 d $3701.54
200
9 a W0 = 4000, Wn+1 = 0.959Wn
150 b $3527.90 c 755.46
10 a S 0 = 13 420, S n+1 = 0.888S n
100
b $11 916.96, $10 582.26, $9397.05,
50 $8344.58, $7409.99
n
c $7409.99 d $1185.21
0 1 2 3 4 11 C 12 E 13 E 14 C
c 100, 10, 1, 0.1, 0.01
Vn
Exercise 7E
100
90
1 a Vn = 2n × 6, V4 = 96
80
70 b Vn = 3n × 10, V4 = 810
60 c Vn = 0.5n × 1, V4 = 0.0625
50
40 d Vn = 0.25n × 80, V4 = 0.3125
30 2 a i 3000 ii 10%
20
b Vn = 1.1n × 3000
10
n c $4831.53
0 1 2 3 4

7F → 7 review 806 Answers
3 a i $2000 ii 6% 12 a 6.38% b 8.76% c 4.91% d 13.10%

b Vn = 1.06 × 2000
n
13 6.78%
c $2524.95 d $837.04 14 E 15 C 16 E 17 A
4 a Vn = 1.125n × 8000
b $11 390.63 c $3390.63
Chapter 7 review
d $1265.63
5 a i $1200 ii 12% Multiple-choice questions
b Vn = 0.88 × 1200
n
c $490.41
Answers
1 C 2 E 3 D 4 C 5 A
6 a Vn = 0.905 × 38 500
n
6 A 7 D 8 A 9 B 10 A
b $23 372.42 c $15 127.58 11 D 12 C 13 C 14 B 15 A
7 6 years 8 100 years 94 16 B 17 C 18 D 19 B 20 C
10 6% 11 $9223.52 12 $32 397.18
13 C 14 E
Exercise 7F 1 a V0 = 20 000, Vn+1 = Vn + 1880
1 a 0.40% b 2.08% c 0.40% d 0.14% b $29 400

e 0.03% c W0 = 20 000, Wn+1 = 1.094Wn
2 a 6.48% b 5.8% c 14.82% d Wn = 1.094n × 20 000
d 9.88% e 8.03% e $31 341.27
3 a V0 = 8000, Vn+1 = 1.048 Vn 2 $328.03
b V0 = 8000, Vn+1 = 1.012 Vn 3 a i V0 = 22 500, Vn+1 = Vn − 2700
c V0 = 8000, Vn+1 = 1.004 Vn ii $9000
4 a V0 = 20 000, Vn+1 = 1.005 × Vn b i V0 = 22 500, Vn+1 = 0.84Vn
b Vn = 1.005 × 20 000
n
c $26 977
ii $9409.77
5 a V0 = 8000, Vn+1 = 1.012 × Vn c 25000
b Vn = 1.012 × 8000
n
20000
c $9231.16
Value
15000
6 a V0 = 7600, Vn+1 = 1.005 × Vn 10000
b Vn = 1.005 Vn × 7600
n
5000
c $7791.91 d 139 months
O 1 2 3 4 5
7 a V0 = 3500, Vn+1 = 1.02 × Vn n
b $3788.51 4 a $0.20
8 a 4.68% b 4.70% c Monthly b Let Vn be the value of the vacuum
9 a 8.25% b 8.24% c Monthly cleaner after cleaning n offices.
10 a A – 8.62%, B – 8.11% V0 = 650, Vn+1 = Vn − 0.20
b A – $3018.10, B – $2837.08 c $250
c B – this loan will be charged less interest 5 a $6575 b $6722.75 c 6.9%
11 a A – 5.43%, B – 5.61% 6 a V0 = 30 000, Vn+1 = 1.0075 Vn
b A – $7603, B – $7860 b $31 142.00
c B – this investment will earn more interest c $32 814.21 d $34 318.81

Answers 807
Answers
7 $234.57 10 8%
8 3.5 11 10%
12 C 13 B 14 C
Chapter 8 Exercise 8B
1 V0 = 5000, Vn+1 = 1.054Vn − 1400
8A → 8B
Exercise 8A
2 a V0 = 2000, D = 339
1 a 2, 5, 11, 23, 47
b R = 1.005
b 50, 90, 170, 330, 650
c V0 = 2000, Vn+1 = 1.005Vn − 339
c 128, 96, 80, 72, 68
3 a B0 = 10 000, Bn+1 = 1.03Bn − 2600
2 a $500 b $100 c 1.03
b $5331
d V0 = 500, Vn+1 = 1.03Vn + 100
4 V0 = 3500, Vn+1 = 1.004Vn − 280
3 a $300 000 b $50 000
5 V0 = 150 000, Vn+1 = 1.0014Vn − 650
c 1.052
6 a V0 = 235 000, Vn+1 = 1.0001Vn − 150
d V0 = 300 000, Vn+1 = 1.052Vn + 50 000
b $234 620.46
4 a 1.003
7 a $2500 b $626 c 8%
b V0 = 3500, Vn+1 = 1.003Vn + 150
d $1117.03
c $3821.48
8 a $5000 b $865 c r = 12%
5 a V0 = 1700, Vn+1 = 1.008Vn + 100
d $3361.85
b $2395.38
9 a V0 = 20 000, D = 3375
6 V0 = 1500, Vn+1 = 1.0002Vn + 4
b R = 1.072
7 V0 = 24000, Vn+1 = 1.005Vn + 500,
c V0 = 20 000, Vn+1 = 1.072Vn − 3375
$27 766.81
10 a V0 = 750 000, D = 4100
8 a $2000 b $1000 c $4412.80
b R = 1.0045
d Vn ($)
c V0 = 750 000, Vn+1 = 1.0045Vn − 4100
4500 11 a V0 = 40 000, Vn+1 = 1.015Vn − 10 380
4000
3500 b $10 217.70
3000
2500
12 a $5000 b $1030 c 12%
2000 d $2030.50 e $3090
1500
1000 13 a $3052.65 b $6000
500
14 a $1 000 000 b $400 c 2.88%
n (years)
0 1 2 d $996 796.16
9 a $20 000 b $2000
15 a $18 400 b 6.6%
c $27 689.06
c $9762.84
d Vn ($) 16 D 17 E 18 A
30,000
25,000
20,000
15,000
10,000
5,000
n (quarters)
0 1 2 3
8C 808 Answers
Exercise 8C
1 a $14 000 b $1540 c $260

Answers
d $13 740
e Payment number Payment Interest Principal reduction Balance
0 0.00 0.00 0.00 14 000.00
1 1800.00 1540.00 260.00 13 740
f Payment number Payment Interest Principal reduction Balance

0 0.00 0.00 0.00 14 000.00
1 1800.00 1540.00 260.00 13 740.00
2 1800.00 1511.40 288.60 13 451.40
3 1800.00 1479.65 320.35 13 131.05
2 a $12 000 b 0.005
c $60 d $240 e $11 760
f Payment number Payment Interest Principal reduction Balance
0 0.00 0.00 0.00 12 000.00
1 300.00 60.00 240.00 11 760.00
g Payment number Payment Interest Principal reduction Balance

0 0.00 0.00 0.00 12 000.00
1 300.00 60.00 240.00 11 760.00
2 300.00 58.80 241.20 11 518.80
3 300.00 57.59 242.41 11 276.39
3 a $36 000 b 0.02
c Payment number Payment Interest Principal reduction Balance
0 0.00 0.00 0.00 36 000.00
1 1000.00 720.00 280.00 35 720.00
2 1000 714.40 285.60 35 434.40
3 1000 708.69 291.31 35 143.09
4 a 1%
b A = 16.75, B = 334.85, C = 342.17
5 a $4000, $100 b 2.5%
c A = 64.81, B = 505.16, C = 1076.65, D = 26.92, E = 530.73
6 A = 421.26, B = 458 736.22
7 a i $15.00 ii 0.25%
b A = 12.53, B = 495.47, C = 4517.30
8 a 1%
b A = 53.02, B = 153.02, C = 5454.52
9 D
ISBN 978-1-009-11041-9
10 D 11 D © Peter Jones et al 2023 Cambridge University Press
Answers 809
Answers
Exercise 8D c $6061.91
4 a $21 867.22 b $12 095.13
1 $554.16
5 a $225 788.13 b $5452.89
2 $1262.60
6 a $34 093.96 b $344.64
3 $692.58
7 a Negative b $28 674
4 $771.27
c $6825.74
8D → 8F
5 a 180.53 b 1380.53 c $624.67
8 a $15 133.81 b $1732.49
6 a A = 345.69, B = 1.72, C = 343.97
9 a $7627.37 b $298.51
b $4131.23 c $131.23
10 E 11 E 12 B
7 a A = 3903.19, B = 34.82, C = 3868.37
b $31 227.69 c $1227.69
Exercise 8F
8 $12 165.50, $165.50
9 1 8.39%
350
2 a 2.7% b $741.19
300
c 60 months
250 3 a $500
Amount ($)
200 b i $6000 ii 6491.15

iii $491.15
150
c 23 months
100
4 55 payments
50 5 a $197 793.85
b i $2158.06 ii $774 237.82
0 1 2 3 4 5 6
iii $534 237.82
Payment (n)
10 6 a $617.80 b $617.72
160
140 c $18 533.92 d $1533.92
120
Amount ($)
7 a i $5253.40 ii $210 135.61

100
80 b i $4175.11 ii $4174.54
60 iii $250 506.03
40
20 8 $446 972.41
0 9 $2591.94
1 2 3
Payment (n) 10 a 9.24% b $2348.50
11 E 12 C c 57 months
11 96 quarters
Exercise 8E 12 a $4000 b $22 965.28
1 a Negative b Negative 13 E
c $123 003.55 d $733 636.83 14 C
2 a $22 327.26 b $37 864.50 15 B
3 a Positive b Negative

8G → 8 review 810 Answers
Exercise 8G 12 3.6%
13 4.8%
1 $46 615.21
14 a $1312 b $78 720 c $86 400
2 $178 558.60
d $1440 e 5.4%
3 a $81 939.67 b $67 141.09
15 C 16 B 17 D
4 $416.37
5 $338 807.90
Exercise 8I
6 a $5312.50 b $6500.67
Answers
c $6495.55 d $2 268 945.88 1 a $9790.50 b $642 000

e $1 518 945.88 c $642 000
7 a 153 months b $229.96 2 a $4800 b $57 600
8 a V0 = 125 000, Vn+1 = 1.0045Vn − 850 3 a $2160 b It will increase
b $11 966 4 $100 000
9 a S 0 = 150 000, S n+1 = 1.005 × S n 5 $50 000
b $272 909.51 c 2.72% 6 a $540 000 b $380 000
10 a $29 697.05 b $34 378 c $451 866.88
11 a $3504.76 b $104.76 7 3%
c $18929.68 d 58 payments 8 6.25%
12 B 9 6.3%
13 B 10 a $17 850 b $350 000
14 E c M0 = 350 000, Mn+1 = 1.00425Mn − 1487.50
11 a $600 b $7600 c Option B
Exercise 8H d Z0 = 200 000, Zn+1 = 1.038Zn − 7600
12 E
1 $600
13 D
2 $350
3 $462
4 $44 550
Chapter 8 review
5 a $2775 b $32 775 Multiple-choice questions
6 a $2555 b $306 600
1 C 2 D 3 D 4 B
c $1 006 600
5 D 6 C 7 D 8 B
7 a $182 b $35 000 c $272.48
9 C 10 C 11 D 12 E
d $64 892.48
13 D 14 E 15 C 16 D
8 $40 000
17 E 18 C 19 E 20 D
9 $16 000
10 $28 000 Written-response questions
11 2.9% 1 a V0 = 250 000, Vn+1 = 1.004Vn − 1800

b $240 185.96 c 56 months

Answers 811
Chapter 10
Answers
d i $1000 ii $12 000
iii $250 000
2 a $781.25 b $147 298.48 Exercise 10A
c 38 months
d i 41 payments ii $3323.07 1 a 2×3 b 1×3 c 3×2 d 3×1
e 3×3
3 a $656.65 b $13 134
9A → 10A
c $3134 2 a 2×3 b 4×1 c 1×3
4 a 40 months b $320.78 3 a 12 b 15 c 28
c $4770.78 4 1 × 12, 12 × 1, 6 × 2, 2 × 6, 4 × 3, 3 × 4
 
9 8
5 a $247.04 b $83 713.37    
1 0 1 9
6 a $1 175 244.58 b 290 months 5 a   b 3 5 c  

2 3 0 1
  
c $3300
 
7 5
6 a square; 2 × 2; 4 b column 3 × 1; 3
c row; 1 × 4; 4
Chapter 9 
1 3 5
 
1 0 0

   
7 a 0 4 7 b 0 1 0
Exercise 9A
   
0 0 2 0 0 1
   
1 E 2 C 3 C 4 E 5 C 6 D 
1 0 0

 
7 E 8 D 9 A 10 D 11 D 12 D   3 0
c 0 1 0and  
13 E 14 E 15 D   
0 5

 
0 0 1
d First 3 are symmetric.
Exercise 9B
8 a C, E b 3 c A d B
1 a $8500 b $222.44 e 4, 2 f 3, 3 g 1, 5 h 3, 1
c A0 = 8500, An+1 = 1.013 × An i 4, 2 j 9 k 5 l 0
d 5.2% e 13 quarters m 1 n 0 o 4 p −1
2 a V0 = 25 000, q 3 r 3 s 1
V1 = 25 000 − 936 = 24 064,
9 
1 2 10 3 11 
   
V2 = 24 064 − 936 = 23 128, −2 −5
   
V3 = 23 128 − 936 = 22 192
 
2 4 4 −1 −4
     
b 36 
3 6

5
  
0 −3

3 a $260 000  
6
1170
b × 100 = 5.4% 12
13 
4 9 

260 000 2 5 10
c A = 1156.71, B = 993.29, C = 256 053.46  
9 16

4 a 204 b $29516.73 c The first
50 14 a   b   c  
5 a $4400 b = $2.50 1 2   4 −2  1 
20      
c Hn = 4450 − 2.50n 0 −2 −4 6 −1
d $3950    
e 581 3 1 2
 
d

6 a $5520 b 22 months
2 3

10B → 10C 812 Answers
15 C 16 D 17 B 18 C 6  
21 5 5
19 B 
 8

 2 3
 4 
1 1
Exercise 10B
 

14 8 6
 
1 a  
0 1 2
4 2 1
 
 
6 2 3 , 3 × 3 7 a   b  
0 1 0 0 1 0 1
Answers
     

2 1 0 1 0 1 1 0 1 0
   
0 1 0 0 1 0 1

b 6 2 3 , 1×3

 
c 1 0 1 0
  c  
1 0
0 1 1
   
3 , 3 × 1; the total number of computers 0
0 0 1
   
1
0 0 1

0
 

owned by the three households

1 1 0 1
2 a 
 
8  
24 32 11 0 0 1
  , 2 × 3; 
0 0 3

32 34 9  
 
b 24 32 11 , 1 × 3 1 3 0
c 9 a Girls 3 and 4 are friends.
  b Girls 2 and 5 are not friends.
24
  , 2 × 1; the total number of small cars c 3: girl 3 has three friends.
32
 
d girl 1, girl 3
sold by the car dealers
10 a i Polar bears eat cod.
3 a 4×4 b
;
430 380 950 900 ii Nothing eats polar bears.
1 × 4;The total exports of B b
P S C Z
c 
370; 4 × 1
   
   0 1 1 0  P
950  
  W =  0 0 1 1 S

150 
   0 0 0 1 C

470
  
0 0 0 0 Z
4  
200 110

180 117
 11 C 12 B
 
135 98 
Exercise 10C
 
 
110 89 
1 a = 9, b = 7
 
 56 53 
 

28 33
 2 a C, F
5   b A and B, C and F, D and E
3 5 8 7 0 2 3 6
  c A and B, C and F, D and E
4 2 2 9 0 0 0 9

Answers 813

Answers
d i 4 4 ii   8 a
1 1    
  16 104 86 24 124 100
1 1 A =   B =  
75 34 94 70 41 96
  
iii   iv −2 2
0
0 0 
 40 228 186

b C = 
 
0 0 0 ; the total
145 75 190

v   vi 9 3
10D
−1
1 number of females and males enrolled
−3


3 in each of the three programs for the
vii   viii   two years
 0 4 16  0 4 16
 
 
 8 20 14
c D =  ; the increase in the
 
12 8 4 12 8 4
 
−5 7 2
ix −2 10 x not defined number of females and males enrolled
4 a   b    in each of the three programs for the
5 5 −3 −1
    two years; a decrease in the number of
5 5 3 1
men enrolled in weights classes
c   d e  
 
0 0  48 248 200
9 8 1 d E =  
   
6 7 1 140 82 192

f     g     h not defined 9 C 10 C 11 E
2 −2
   
3 3
5 a    b   Exercise 10D
−2.2 1.1 −0.2 −13.8
1 a i, ii, iv, v, vi, vii
   
7.7 4.4 1 −3.7  
c 
 5 8 7 d 0.6 2
    1 
1 3.2 b i [6] ii [2] iii   iv −3 7 
 

−1
   
 16 0 3 1 0 −0.6 2
c i [6] ii  iii [0]
   

−1 5 4
 2 0 −2
6 x = 2, y = 4, z = 6, w = −4
 
1 0 −1
 
7 a 
0 0 0

 
2.4
 
2.8
 
2.5
 
3.4 iv undefined
       
[0] b [1] c [3]
A = 3.5 B = 3.4 C = 2.6 D = 4.1 2 a
    d  
1
         
1.6 1.8 1.7 2.1 3
b 
  e 
5 5 3
 f  
11.1    
 
1 2 0
13.6; the total (yearly) DVD sales    
 
2
 
7.2
 
for each store 3 a 
 1 2 3
 b
  6 7 


−3 −6 −9
  
  19 18
5 10 15
 

10D 814 Answers
 
c   d   110 000
4 0  −5 15 9  
    116 000
2 −2 −2 8 4
11 XY =   It gives the total sales of

   
154 000
1 4 −5 8 7
   
Answers
 
4 a   post-multiply by this matrix. 58 000
1 each
   of the  dealers.
1  29 
pre-multiply by this matrix. 12 a  , John took 29 minutes to eat
b

8.50

1 1 1
     food costing $8.50
7 1 2 1 10  
       29 22 12 
5 1 2 2 1 =  5  . b  ,
8.50 8.00 3.00
      
8 1 4 1 13
    
John’s friends took 22 and 12 minutes
 

9 0 2 to eat food costing $8.00 and $3.00

6 1 1 1 1 7 3 = 18 10 9 respectively

   

8 3 4
   0.2
79 78 80  
7 a   b
13 a   b 0.3
22 8 10 6 14 11 80 78 82  
0.5
   
12
  c Semester 1: 79.2; Semester 2: 80.4
18
 
  d Semester 1: 83.8; Semester 2: 75.2
12 e No, total score is 318.6
 
30 f 3 marks
14 5 5  15 20 50 75 , 2250 3625
       
8   9  
9 3000   ,   ,    
 
5 10 20 35 75 125 3625 5875
         
7 2800
15 5 7   2 19 17 40 , 23 97 
           
6
2600
 
    ,   ,    
7 26 19 59 40 137 97 314
           
4 2200
  16 
24 30 36

3  
  38 59 64
1  
33 54 51
 
10 a 2 × 3
b i  17 a   b  
−1 5 −3 8 

184.50    

236
 5 2 6 −3
ii the total revenue from selling c   d  
 17 8 

  16 2 

products A, B and C at Energy and 
8 17
 
2 39

Nourishing respectively e  
 29 −5
c number of columns in P , number of  
−5 13
rows in Q
18 A 19 C 20 D

Answers 815
Answers
Exercise 10E f  
−8
 
1 a i   ii   10
1 0 1 0 0
6  
−1 2
   
0 1 0 1 0
  
 
−3 5
   
0 0 1
iii   7 a X = A−1C
1 0 0 0
10E → 10F
  b X = (AB)−1C = B−1 A−1C
0 1 0 0
  c X = A−1CB−1 d X = A−1C − B
0 0 1 0
 
  e X = A−1 (C − B)
0 0 0 1 f X = (A − B)A−1 = I − BA−1
8 x = −5000, y = 15 000, z = 0
    
1 2 1 0 1 2
b AI =    
   =   ;
0 3 0 1
 
0 3 9
Spray P Q R
    
1 0 1 2 1 2 8
46 12
IA =     =   Barrels
0 1 0 3
  
0 3 13
39 13
∴ AI = IA = A 10 a  
 0.1 0.25 −0.4
c 
 0.3 −0.75 0.8 

      
1 2 0 1 0 0 1 2 0 −0.2 0.5 −0.2
 

CI = 3 1 0 = 3
    
1 0 0 1 0 ; b
      Product P Q R
0 1 2 0 0 1 0 1 2 Number per day 13.5 0.5 13
    
1 0 0 1 2 0 1 2 0 
  11 Brad 20; Flynn 10; Lina 15
IC = 0 1 0 = 3
   
1 0 3 1 0
      12 A 13 E 14 D
0 0 1 0 1 2 0 1 2
CI = IC = C Exercise 10F
3 a 3 b −3 c 0 d −8
1 B only
4 a  10 2  b  20
 1 
 11 − 3  2 a X= H U T S b n=4
  
   9 18 
1   −50 1 
 
0   3   4  
3 9 9 0
0 1 0  X 
Matrix has no inverse, det (D) = 0
 
c  
0
0 0 1 W 
 
d
1 1 1   
 − −  1
0 0 0  Z 
 
 2 2 2   
  
1 0 0 0 Y

 0 1 1 
 
5 a b
0 0 1     
0 1 1 2 1 1
5 a b 
      
−9 −3
  
 −28 −15 
 C = 1 0 1 C 2 = 1 2 1
   
1 3 39 22
   
1 1 0 1 1 2
c 
−14 9
 d   e  
c 2
  1  3.5 
  
 2.5 
 6 a There is no direct communication link
−9 8 0
 
between the towers.
 
−1.5
 

10G 816 Answers
b T 1 and T 3 2 a
c 1, 0 losers
d There is a 2-step communication link A B C D E
Answers
between T 3 and T 1.  
A 0 1 1 1 1 
e 6  
B 0 0 1 0 0 
f 
T1 T2 T3 T4

winners C 0
 0 0 0 0 
  
 1 1 1 0  T1 D

0 1 1 0 0 
   
T =  1 2 1 1  T2 E 0 1 0 1 0

 
 1 1 2 1  T3
  A : 4, B : 1, C : 0, D : 2, E : 2
0 1 1 1 T4
A; D and E equal; B; C.
g T 1 and T 4
b 
0 2 2 1 0

7
A B C D E  
  0 0 0 0 0
A  0 0 0 0 1   
  D2 = 0 0 0 0 0
B  1 0 0 0 1   

 0  0 0 1 0 0
C 1 0 0 1   
0 1 2 0 0
 

D  0 0 1 0 1 
c  
  0 3 3 2 1 9
E 0 1 1 0 0  
0 0 1 0 0 1
8 D 9 D  
T = 0 0 0 0 0 0


0 1 2 0 0 3

Exercise 10G 

0 2 2 1 0 5
1 a
A B C D
A

 0 1 1 0 
 The tie can be broken using two-step
  dominances to give the ranking
B  0 0 1 0 
  A, E, D, B, C.
C  0 0 0 0 
  A B C D E Score
D 1 1 1 0 3 a  
A  0 0 1 1 0 2
D, A, B, C
 
B  1 0 1 0 1 3

 

D = C  0 0 0 1 0
 
b 1
A B C D  
  D  0 1 0 0 0 1
A  0 0 1 0   

 0
 E 1 0 1 1 0 3
B 0 0 0 
  b 
0 1 0 1 0

C  0 0 0 0 

 
  1 0 2 3 0
D 0 1 2 0  
D = 0 1 0 0 0
2  
D, A, B, C  
1 0 1 0 1
 
0 1 1 2 0
 

Answers 817
Answers
c A B C D E Score Chapter 10 review
 
A  0 1 1 2 0 4
 Multiple-choice questions

B  2 0 3 3 1 9
 

D + D2 = C  0 1 0 1 0
 
2 1 C 2 D 3 B 4 D
 
D  1 1 1 0 1

 
 4 5 D 6 A 7 B 8 C
 
10 review
E 1 1 2 3 0 7 9 D 10 E 11 A 12 B
The matrix D + D gives the following 13 D
2 14 E 15 A 16 D
ranking: 17 D 18 E 19 D 20 C
21 A 22 C 23 C 24 C
Rank Player Score
25 A 26 C 27 C 28 C
First Bea 9 29 E
Second Eve 7
Equal Ann and 4
third Deb 1 a   b  
0 1 2 0 2 1 1
Fifth Cat 2 
    
1 0 1 2 0 2 0
4 a 10    
2 1 0 1 2 0 1
b Ash defeats Carl and Dot
 
1 0 1 0

Ben defeats Ash, Carl and Dot
c  
Carl defeats Dot 0 0
 
0 0
Dot defeats Elle
2 a 2×1 b 1×2
Elle defeats Ash, Ben and Carl
c Yes; number of columns in C equals
c Ben = Elle, Ash, Carl = Dot

0 1 0 1
 the number of rows in J.
0

0 0 1 0  d [162.41] ; 5 × 30.45 + 4 × 2.54 = 162.41
1
e 
  
172.57

5 a 1 0 0 1 0
   
0 1 0 0

0 185.24
3 a 456 b 2×2
 
1 0 1 1 0
b E, B, A = C, D c  
354 987
B = 
 
0 2 2 1  
314 586
  
0 0 1 0
6 a 
 
b A d 
688 1863

0 0 0 0 C =  ; the total number of
527 1042
  
0 1 2 0
 
  books of each type in the two stores
0 1 1 0 e i 2×1
 
0 0 1 1
ii  
7 a  b A, B, D, C 31 236
 
0 0 0 1  
  18 021
1 0 0 0
 
iii total value of fiction and
8 A, B, D, C non-fiction books at bookshop 1
9 E, B, A = C, D
 
668 1752
f 2A =  
10 C 11 A 12 D 426 912

4 a 1×5 d
A B C
b i 
90 135 165 150 60
  
A 0.45 0.35 0.15
R =    
48 72 63 88 32 B 0.25 0.45 0.20

 
ii the number of students expected to C 0.30 0.20 0.65
get a C in Mathematics .
M P 2
O B
c i F = 220
h i
Answers
197
 
O 0.96 0.98
 


600
 B 0.04 0.02
ii FN = 220 197   = [195 040]
 
320 3
F P
The total fees paid are $195 040.  
F 0.80 0.14
5 a N= 8 6 1 b NG = [575]  
P 0.20 0.86
c total number of points scored by Daniel
6 a 80 tonnes b 100 tonnes
4 B
c $186, 000
d i 3×1
Exercise 11B
ii The price per tonne of each of
the minerals 1 a  
0.85 0.25
T = 

1000 
0.15 0.75
  
iii  700 
0.15
 
300
 
0.85 P I 0.75
0.25
Chapter 11
b 0.85 × 80 = 68 c 0.25 × 60 = 15
Exercise 11A d 0.15 × 120 + 0.75 × 40 = 48
2 a i 10% ii 80% iii 10%
1 a
A B b i 680 ii 85
 
A 0.40 0.55 c i 1150 ii 0 iii 0
 
B 0.60 0.45 d All (100%) of the sea birds who nest at
site A this year will nest at site A next
b
X Y year.
 
X 0.70 0.25
 
Y 0.30 0.75
3 a i 76440 ii 7560
c b i 5500 ii 1210 iii 266
X Y Z
  4 a i 18 ii 6 iii 6
X 0.6 0.15 0.22
  b i 84 ii 66 iii 30
Y 0.1 0.7 0.23

  84
 
Z 0.3 0.15 0.55 c 66
 
30
 
Answers 819
Answers
d i 21 ii 3 iii 6 b i 
207
 ii 
 225 

e 180
   
S 2 136 S 3 = 132.1
5 B 6 E    
257 242.9
iii 
224.9

Exercise 11C  
S 7 = 129.7
11C
 
1 a 
380
 b  
398 225.4
S 1 =   S 2 =  
c See solutions
220 202
 
c   6 a 
0.80 0.25

399.8
S 3 =    T =   
0.20 0.75

200.2

2 a   b   b  
400
399.998 400
S 5 =   S 7 =   S 0 =  
200.02 200
 400
 
c 
400
 420
S 12 =   c S 1 =  , 420 to Jill’s and 380 to
380

200

Pete’s
3 a 
5460
 b 
5456
 
442.2
S 4 =   S 3 =  
d S 5 =   , 442 to Jill’s and 358 to
4780 4784

357.8

4 a i  
130 ii  
151 Pete’s
S 1 =   S 2 =  
 
444.4
170 149 e steady state solution: S s =  , 444
iii   355.6
165.7
S 3 =    to Jill’s
 and 356 to  Pete’s
134.3
 0.90 0.60
7 a T =  
b 
0.72269

0.10 0.40

0.55462
T 5 =    
0.27731 0.44538 1500
b S 0 =  
c i  
151 ii  
165.7 500
S 2 =   S 3 =     
149 134.3 1650

c S 1 =  , 1650 are happy and 350
iii 
191.8

350
S 7 =   are unhappy
108.2  
d See solutions 1712.55
d S 4 =  , 1713 are happy and
5 a i   ii   287.45
180 207
    287 are unhappy
S 1 = 130 S 2 = 136 
1714.3

e steady state solution: S =  ,
   
290 257 285.7
iii 
 225 

1714 are happy and 286 are unhappy
   
S 3 132.1 1200
8 a S 0 =  600 
   
242.9  
200
 

11D → 11E 820 Answers
ii S 2 = T S 1 − B
 
1270
 
b S 1 =  440 , 1270 are happy
     
0.6 0.2 100 −20
=   ·   − 
  
290
 
0.4 0.8 100 20
    
 
1310.33   
 80  −20 100
  
 
c S 5 =  429.82 , 1310 are happy =   −   =  
120 20 100
  
 
259.85 
11500

Answers
   
1311.7
  2 a i S 1 =  8500  ii 7300
 
d steady state solution:  429.1 , 1312 
10000

 
259.1
 
b A: 30 000, B: 0, C: 0; While the sea
are happy birds move between nesting sites each
9 A 10 E 11 A year, the ‘1’ in the transition matrix
indicates that, once a sea bird nests at
Exercise 11D site A, it continues to nest at this site.
    Meanwhile, some of the birds who nest
 80   68.8 
1 a i   ii   at sites B and C each year will move to
120 131.2
  
site A until, in the long term, all birds
b i are nesting at site A.
S1 = T S0 + R
c i 
 9500  ii  9000  iii 8507.5
          
0.6 0.2 100 10
=   ·   +        
 9500   9150  8912.5
0.4 0.8 100 5
          
11000 11850 12 580
           
 80  10  90 
=   +   =   3 C 4 B
120 5 125
ii Exercise 11E
S2 = T S1 + R
     
0.6 0.2  90  10 1 a i 1.9 ii 0.6
=    · 
   +  
0.4 0.8 125
  
5 b
1.1
     
 79  10  89 
=   +   =   2.1
136 5 141 1.9
1 2 3 4
c i 0.7 0.5 0.6
S1 = T S0 − B
c i 
510 ii 784.8 iii  208 276 
          
0.6 0.2 100 −20
=   ·   − 
       
 70    
212.8   103 876 
0.4 0.8 100 20
      
     
       50  178.5  36 984 
 80  −20 100      
=   −   =   60 21 15 815.8
120 20 100
  

Answers 821
Answers
d   e 37 f 90
3483

1829
 g i 111 ii 158 iii 235
 
 600  h i 99 ii 145 iii 233
  It appear that the population rate of
291
 
increase approaches 10% per year.
2 a
Further investigation confirms this.
11E
2.1
5 a  
 0  b  0
 
3.1 0 1000
2.9    
 0  0.02 0 0
1 2 3 4

0.8 0.7 0.5
   
50 0 0.05 0
   
b 0.42 c 1000
0 0
1 2 3 1 2 3
0.6 0.75 0.02 0.05
c
d i   ii  
8 50 000    0 
3
 0   
 0 
 
0    
0 50
 
1 2 3 4
0.4 0.5 0.25
iii 
50 000

     
 0 1.3 2.4  0 2.3 3  0 
     
3 a 0.7 0 0  b 0.6 0 0
0
  
  
0 0.6 0 0 0.3 0

e i 
50 000
 ii  
   50 
 0 1.4 2.6 0.6  
   1   
100
0.5 0 0 0     
c  5 50
     
 0 0.4 0 0 
  iii 
50 000

0 0 0.05 0

 
 1 
4 a    b  
15 0 0.2 0.9 1.1 0
 
5
 
   
20 0.8 0 0 0 0
    6 a i   ii
269 356 iii 622
   
30 0 0.9 0 0 0
        
    127 168 294
15 0 0 0.7 0 0      
 
30 40 70
       
10 0 0 0 0.8 0
i 427 ii 565 iii 986
c
1.1 0 b At this stage the rate of increase is
0.9
0.2 approximately 5.7%
1 2 3 4 5
0.8 0.9 0.7 0.8 7 a i 
1400 ii 700 iii
    
2160
 
d i   ii      
 240  840  420 
48   58      
12  
  37 
100
 
60
 
210

   
18 22
 
 
21 21
 
   
12 19
11 review 822 Answers
b i 
976 ii 1241 iii
    
1579 Chapter 11 review
     
460  586   745  Multiple-choice questions
     
91 115 146
     
1 B 2 A 3 C 4 B 5 B
c i 
974 ii 1237 iii 1571
    
6 C 7 C 8 A 9 B 10 B
     
460  584   741  11 B 12 C 13 E
Answers
     

90 115 145
     
The population appears to be
iincreasing at a rate of 27% 1 a i   ii 748
504
 
d
244
 
     
 0   0  200
b 
 
517
     
8 a i 300 ii  0  iii  0   
      257
0 100 0
     
c Diiscoo attendance is expected to
b Population cycles through three states
increase to around 520 and stay at that
c i Population decreases by 50% every
level.
three time periods
d Diiscoo attendance is expected to
ii Population increases by 25% every
decrease to around 20 and stay at that
three time periods
      level.
2800 1080 4440
      Y M O
 200 1400  540 2  
9 a 

b
 
c
   Y 0.1 0.2 0.4
 200  100  700  
      M 0.9 0 0 
40 20 10
       
O 0 0.8 0.6
10 a i Every 3 years, the population 3 a  
returns to 1000 newborns 0 0.1 0.9 0.2 0 0 0 0
 
ii Every 3 years, the population 0.98 0 0 0 0 0 0 0

increases by 50% and returns to 0 0.95 0 0 0 0 0 0


only newborns  
0 0 0.95 0 0 0 0 0
iii Every 3 years, the population  
0 0 0 0.9 0 0 0 0
decreases by 40% and returns to  
0 0 0 0 0.7 0 0 0

only newborns  
b Long-term growth rate 1.37; long-term 0 0 0 0 0 0.5 0 0
 
ratio of age groups 818 : 130 : 52 0 0 0 0 0 0 0.1 0
   
11 D 12 D 13 A 104.5   28.83 
   
107.8  102.41 
   
  
 0  102.41 
   
 90.25  0 
b S2 =  , S =  
 85.5  3  81.225
   
 31.5   59.85 
   
   
 0   15.75 
   
0 0
Answers 823
Answers
c 1.035 b  
4000
S 0 = 
  
0.2 0.5 0.6 0.4 980 
6000
    
0.7 0 0 0  692
4 a 
 
b 
  c
0  488
 0.7 0 0   
 
4360
S 1 = 
 ; 4360 fish in Lake Blue and
0 0 0.7 0 344
  
5640

   
972 7467 5640 fish in Lake Green
12A → 12B
   
680 4762
d
c  , 0.992 d 
   
484 3038  
4555.156
    S 3 =   ; 4555 fish in Lake Blue
342 1938 5444.844

e Day 56 and 5445 fish in Lake Green
e
 
4590.2
S s =   ; 4590 fish in Lake Blue and
Chapter 12 5409.8

5410 fish in Lake Green
Exercise 12A 3
a 30
1 E 2 D 3 A 4 A 5 B 6 C
7 D 8 C 9 C 10 D 11 B 12 A 20
13 A 14 B 15 C 16 C 17 C 18 E
E J Y A
19 D 20 C 21 A 22 D 23 B 24 D 0.5 0.1 0.05
25 A b 50 c 920 d 5 weeks
e 5%
Exercise 12B f i 1667 ii 1706
g 2%
1 a 3×1 b 1×3  ii 95
4a i 
c HC; because the number of columns in 100
 
H equals the number of rows in C 
50

d i [696.72] b S n = T nS 1 c 5 weeks
ii the number of Australian dollars d 90
($696.72) that you would receive 5a i 3 × 1 ii k = 1.2
by converting your foreign b i A and B and A and C
currency into Australian dollars; ii D → B → A → C
HC = 102 × 1.316 + 262 × 1.818 + iii C → A
 →B and C → D → B
385
 516 × 0.167  
566.21 c i S 1 =  75  ii 230
   
e 137.46 140
 
 

647.21
 iii 70.56%
2 a  
0.67 0.28
 
0.33 0.72

13A 824 Answers
 
20 d C
 
d B = 15
  D
25
 
Answers
6a 4 × 2 H
2 a b
b i 435
  ii 72.4%
 60 

120

c
c L =   Q × L = 5115

 50 
 
85 d
   
39 41.4
   
36 35.1
7a S 1 =   S 2 = 
   
21 20.1
   
24 23.4
  3 a i b ii c ii
42.76

34.64
 4 a B C
A
b S 10 = 
 
19.72
  F
E D
22.87
b A B
(P) 43, (S ) 35, (D) 20, (W) 23
c 3
C E
d 36 D
e 3.9 c B
C
f 47.4
A D
8a 550
b The number of sandwiches sold in
F E
week 3.
c Hamburgers $15, fish and chips $14 d not possible
sandwiches $12 5 a i v = 8, f = 6, e = 12
b i v = 6, f = 8, e = 12

d L= 1 0 1
c i v = 7, f = 7, e = 12
d i v = 5, f = 3, e = 6
e i v = 5, f = 6, e = 9
Chapter 13 f i v = 6, f = 4, e = 6
6 a 4 b 12 c 19
Exercise 13A 7 7 8 14 9 15
1 a i 3 ii 2 iii 1 10 C 11 E 12 C 13 B
b 14 14 E 15 C
c town D and town H

Answers 825
Answers
Exercise 13B 2 a A B b B
A
1 a
A B C D
A

 0 1 1 0 
 C C
D
 
c A D
B  1 0 1 1 
 
C  1 1 0 0 

13B → 13C

D 0 1 0 0 C
B
b
A B C D
  3 C is an isolated vertex.
A  0 1 1 0 
  4 Leading diagonals will all be ‘1’.
B  1 0 0 1  5
A B C D E
 
C  1 0 0 1 

 
  A  0 1 1 1 1
D 0 1 1 0  
B  1 0 1 1 1
c
A B C D 
 1

  C 1 0 1 1
A  0 1 0 0 
 
D  1 1 1 0 1

 
B  1 0 0 0   
  E 1 1 1 1 0
C  0 0 0 1 
 
D 0 0 1 0 6 E 7 A 8 E 9 A
d 10 C 11 B 12 E
A B C D
 
A  0 1 1 1 

 1
 Exercise 13C
B  0 1 1 

C  1

1 0 1  1 a path b trail c path d walk

D

1 1 1 0 e trail f path
e 2 a walk b cycle c path d walk
A B C D E F
  e path f walk g h
A  0 1 1 0 0 0 
3 a i Euler trail
 
B  1 0 0 1 0 0 
ii A−B−E−D−B−C−D−A−E
 

C  1 0 0 1 0 0 
  b neither
D  0 1 1 0 0 0 
  c i Euler trail
E  0 0 0 0 0 1 

  ii A−C−E−C−B−D−E−F
F 0 0 0 0 1 0 d i Euler circuit
f ii A−B−C−E−D−C−A
A B C D
A

 0 0 0 0 
 e i Euler circuit
  ii E−F−D−E−A−B−D−C−B−E
B  0 0 0 1 

 0
 4 a A−B−C−F−I−H−E−G−D−A
C 0 0 2 
  b A−B−C−D−E−F−A
D 0 1 2 0
c A−B-D−C−E−A
5 F−A−B−C−D−E−H−G

13D → 13F 826 Answers
6 a 2 b 7 c B Exercise 13E
A
1 a A − B − C − H, 160
b A − C − F − E − G − H, 53
C D
c A − D − E − F − H, 385
d Vertices are not all even.
d A − B − E − F − I − H, 87
7 a v = 9, e = 12, f = 5
2 23 minutes
v−e+ f =2
Answers
b i Hamiltonian cycle
Exercise 13F
ii Lake Bolac − Streatham − Nerrin
Nerrin−Woorndoo−Mortlake 1 a 6
−Hexham−Chatworth− b 4 5 3 2
Glenthompson−
2 6
Wickliffe−Lake Bolac.
5 3 2
and the reverse of this 2 2
c i Eulerian circuit 6
3 5 2
ii Not all vertices have an even 3
degree 6
2
d i Lake Bolac - Wickliffe c 22, 20, 21 (Answers will vary)
ii LWMHCGWCWNSLWL 2 a 2
A B 2 E
8 a Yes b Yes c No d Yes 3 F
e No f Yes g Yes 2
1 10
9 3 C D
b C
10 C 11 D 12 E 13 B B
16
14 A E 16
A D
12 11
10
15 80
Exercise 13D G F
c B 18 C
1 a D–E b 17 minutes 10
c 8 minutes d 36 minutes 10 9
E
2 11 A D 47
3 a 34 b 56 c E–B–A–E, 22 d D
H
d A–E–F–G–I or A–C–F–G–I 70
C 100 200 G
4 a S –B–D–F, 12 100
b S − A − C − D − F, 10 80 90
F
B
c S − B − D − F, 15 90
d S − A − E − G − F, 19 730
A
5 19 km 3 490 m
6 B 7 C 8 A 9 E

Answers 827
Chapter 14
Answers
4 A 5 B 6 C 7 E
8 B
Exercise 14A
Chapter 13 review
1 a 3 b 8
2 C1 = 14, C2 = 12, C3 = 21
1 C 2 C 3 A 4 D
3 C1 = 12, C2 = 16, C3 = 16
13 review → 14B
5 A 6 C 7 B 8 B
9 E 10 B 11 A 12 A 4 a 9 b 11 c 8 d 18
13 C 14 B 15 B 16 B 5 a A, 14; B, 23; C, 12; D, 16; E, not a

cut that can be used to determine the
maximum number of available seats
1 a A−B−C−F−G−E−X, 11 b It does not prevent flow from source
b A−C−E−I−H−X, 127 (Arlie) to sink (Bowen).
c A−D−E−G−H−J−M−X, 55 c 12
d A−B−D−E−G−I−L−M−X, 49 6 a sink 1=10,sink 2=11
2 a i B b sink 1 = 8, sink 2 = 18
6 9
C 7 a 9 b 18
A 8 D 4 8 a Cut passes through edges with weights
5 F 7
4 5 20,10,30,30
8 5
b 59 c 25
E
ii 24 9 D 10 C 11 B
iii A B C D E F
1  Exercise 14B
 
A  0 1 0 1 1
 
B  1 0 1 0 0 0  1 a Worker 1−Task 3; Worker 2−Task 1;
 
C  0
 1 0 1 1 0  Worker 3−Task 2

b Worker 1−Task 5; Worker 2−Task 6;

D  1 0 1 0 1 1 
Worker 3−Task 4
 
E  1 0 1 1 0 1 

2 Niranjan − Cake; Nishara − Candles;

F 1 0 0 1 1 0
Dinesh − Serviettes; Dhishani − Balloons
b i 45 km 3 Two answers possible
ii Some vertices are visited more than Player 1 −Right Wing; Player 2 − Left
once. wing; Player 3 −Centre; Player 4 − Right
iii F–E–D–C–B–A–F Defence; Player 5 − Left Defence
iv 33 km (for route above; other Or
answers possible) Player 1 − Centre Player 2 − Right wing
c C and F Player 3 − Left Defence Player 4 − Right
Defence Player 5 − Left Wing
3 a 4 b 18
c v = 6, e = 9, f = 5; v − e + f = 2 4 a two distinct groups of vertices (people
d-l See solutions and flavours)

14C 828 Answers
b chocolate f A
Gloria start B D E G
vanilla F
Minh C J
H
finish
Answers
Carlos peppermint I K
2 a H
Trevor butterscotch F
I
start m my
strawberry du finish
c 5 G
J
5 a W–D, X–A, Y–B, Z–C b D
B
b e.g., minimum cost is 11; W–A, X–B, A dummy finish
start E
Y–D, Z–C C
6 Dimitri 800 m, John 400 m, Carol 100 m, c R
U
P
Elizabeth 1500 m dummy finish
start S
7 Joe C, Meg A, Ali B V
8 A–Y, B–Z, C–X, D–W Q T
9 Champs Home, Stars Away,
d D
Wests Neutral; or Champs Neutral, Stars B G
H finish
A
Away, Wests Home. start dummy
F
Cost = $20 000 C
E
10 A Mark, B Karla, C Raj, D Jess; or A
3
Karla, B Raj, C Mark, D Jess; 55 km a
11 D 12 A 13 E 14 D
A −
B −
Exercise 14C
C A
1 a B D D A
A
start finish E B, C
C E F D
b P R
G E
T
start finish b
Q S
c T V P −
X Y Z
start Q P
U W R P
d F I S Q
G L
start J finish T Q
K
H U S, V
e K M V R
R S
start N finish W R
Q
L O P X T, U

Answers 829
Answers
c f
Activity Immediate predecessors Activity Immediate predecessors
J − A −
K − B A
L J C A
M N D A
N K E B
14C
O K F C, D
P N G D
Q L, M H E, F, G
R P I G
S O, R J I
T Q K H
d 4 a Remove panel.
b ‘Order component’ and ‘Pound out dent’
A −
5 a
B −
C A Activity Immediate predecessors
D A A −
E D, B B −
C −
F C, E
D A
G D, B
E B, F
H B F C
e G B, F
P − H D, E
I H
Q −
J I, K
R P
K G
S P
L G
T Q M H
U R N J, L
V S O N
W S, T b A−D−H−M
X U A−D−H−I−J−N−O
Y W c B−E−H−M
Z V, X, Y B−E−H−I−J−N −O
B−G−K − J −N −O
B−G−L−N −O
d C−F−E−H−M

14D 830 Answers
C−F−E−H−I−J−N−O 8 a D, F, G b 13
C −F −G−K − J −N −O c Activity H lies on the critical path and
C −F −G−L−N −O if delayed, the completion time of the
Answers
6 a D, F, H project will be extended.

b A, B, C, D, E, F, G, H d 15 e F
7 aG b K 9 a i 25 ii 29 iii 30
c G is the immediate predecessor of both b 5 c 2
I and K, however activity K has other 10 a B, C b A, B, C, D, E, G
immediates predecesors not common c B, F, H, J d 4 e E, G, I
to activity I 11
8 D 9 A 10 A a
A −
Exercise 14D
B −
1 a p = 12 b w = 10 C −
c m = 8, n = 8 D B
d a = 10, b = 18, c = 11 E B
e f = 9, g = 12 F A, D
f q = 8, r = 3, p = 5, n = 9 G C, E
2 a 3 b A−C c 5 d 13 H F, G
e 2 I F, G
3 a 12 b 10 c 9 d 1 J G
e 3 f 9 K H
4 a D−E−F b A: 1, B: 1, C: 15 L I
5 a B−E−F−H−J M J
b A: 1, C: 14, D: 1, G: 1, I: 1 b 3 c 9
6 a d B−E −G−H −K
Duration Immediate
Activity B−E −G−I −L e 6
(weeks) predecessors
12 a
A 3 −
B 6 − S, 3
P, 4 U, 3
C 6 A, B
Q, 5 T, 6 W, 8 Y, 6
D 5 B start finish
dummy
E 7 C, D X, 13
R, 12 V, 4
F 1 D
G 3 E
H 3 F
I 2 B
b B−C−E−G
7 a 8 b 11 c C −F −G
d B

Answers 831
Answers
b b B − E − G − I, 21 hours
Activity ES T LS T
c 18 hours
P 0 11
2 a A−B−F −H b 21 days
Q 0 10
c 20 days d $100
R 0 0
3 a B−E−H−J b 2 hours
S 4 15
c 6 hours d 14 hours
T 5 15
14E → 14 review
4 a 4 b 17 hours c $1200
U 7 18
V 12 12 5 a 22 days b $870
W 12 21 6 a C, D, H
X 16 16 b B, E, H, I, J
Y 29 29 c i 21 days ii $450
c R−V−X−Y d 35 weeks 7 a 21 days b 10 days c 5 days d K

e $6000
13 a L, 4
I, 2
O, 6 8 a 29 days b 6 c 4
start J, 3 M, 8 d Two answers possible:
S, 1 H, 2 J, 0 K, 2 L, 1 M, 1
K, 5 P, 6
H, 2 J, 0 K, 1 L, 1 M, 2
N, 1 Q, 7 finish
9 B 10 B 11 D
T, 9
R, 5
b Activity ES T LS T
I 0 9 Chapter 14 review
J 0 3
K 0 0
L 2 11 1 D 2 D 3 A 4 A
M 6 7 5 C 6 B 7 E 8 D
N 5 5 9 D 10 E
O 14 15
P 6 16
Q 6 6 1 a Alvin should write the Body par 3
R 5 8 Billy should write the Body par 2
S 20 21 Chloe should write the Body par 1
Danielle should write the Conclusion
T 13 13
Elena should write the Introduction
c K−N−Q−T d 22 weeks
b 43 minutes
14 E 15 D 16 C 17 A 2 a 26 b 15
18 D 3 Rob – breaststroke, Joel – backstroke,
Henk – freestyle, Sav – butterfly or Rob
Exercise 14E – breaststroke, Joel – butterfly, Henk –
backstroke, Sav – freestyle. Time = 276
1 a A − D; B − E − F; B − E − G − I;
C−H−I 4 a 26 b 15

C, 6
5 a A, 4
3 a Cut 1 does not isolate the source from
E, 10
Start dummy H, 5 the sink.

dummy Finish
B, 5 G, 4
b 26 c 22
A C H J
D, 7 F, 5 4 Start Finish
b 12 c 1 hour d 4 D G I
e B−D−E−H f 27 hours B F
E
g i B−D−F −G−H
Answers
ii 22 hours
5 a 9 b 7 c 1
h D, H must be in that order
d B−D−E −G e 15
6 a A → 1, D → 4, F → 10, K → 12,
Chapter 15 b B–C–E–G–J–K
7 a i 2.1 km
Exercise 15A ii PQRT S U or PRQS T U or
PRQT S U or PRT QS U
1 A 2 E 3 E 4 B
5 C 6 B 7 A 8 B b i R−Q−P−R−T−Q−S−T−U−S or
9 E 10 C 11 E 12 D R−Q−P−R−T−S−Q−T−U−S
13 D 14 B 15 E 16 A ii travel each road only once
17 C 18 C 19 C 20 E 8 a None of the edges overlap.
21 B 22 A 23 D 24 C b 7 + 6 − 11 = 2 c C d 297 km
25 B 26 C 27 C 28 E e no f 79 km g 127 km
h 187 km
Exercise 15B 9 a 5 b 24 hours c 7 hours
10 11 megalitres/day
1 a 14 b 3 c 3
11 a 112 km
e 5 + 5 = 7 + 2, 9 = 9
b i minimum spanning tree
2 ii M
P Q R S
  38
P  1 1 0 1  L 31
 
Q  1
 0 2 1  35 S N

R  0 2 0 1 

  24
S 1 1 1 0 R
47 O
Q
63
P
55
Q
P
R iii 293 km
c 306 km
12 a A, B, C

Answers 833
Answers
b LST for B is 1, EST for E is 10, LST b time
for I is 18 c i IQR = 6.2 seconds
c i A−D−F−I−J ii 27 months ii Upper fence = 28.2 + 1.5 × 6.2 =
d i B−C−D−F−I−J ii 25 months 37.5
13 a A–Z, B–W, C–X, D–Y, or A–Z, B–X, d 10 people
C–W, D–Y e From this information it can be con-
16A → 16B
b $130 cluded that the time taken to complete
14 a 15 weeks b $8500 c 3 the task is associated with the number
of distractions. The median time
Chapter 16 taken by the group who completed
the task with no distractions was 25.0
Exercise 16A seconds, faster than the group with a
few distractions which has a median
time of 26.2 seconds, which was in
1 A 2 C 3 B 4 D 5 C turn faster than the group with many
6 B 7 B 8 C 9 A 10 A distractions which took a median time
11 E 12 E 13 B 14 B 15 A of 29.2 seconds to complete the task.
16 D 17 B 18 B 19 A 20 E 3 a r2 = 84.8%
Recusion and financial modelling b 84.8% of the variation in fuel con-
sumption can be explained by the
21 C 22 B 23 D 24 A 25 B
variation in speed.
26 C 27 B 28 D 29 A 30 C
c 9.0 litres/100 km
Matrices
d slope = 0.0218. On average, for each
31 B 32 B 33 A 34 E 35 D additional 1 km/hr increase in the
36 D 37 E 38 D 39 E 40 A speed of the car, the fuel consumption
41 B increases by 0.0218 litres/100 km.
Networks e predicted value = 8.40, actual
value= 8.30. Thus residual= −0.10.
42 C 43 D 44 B 45 D 46 E
4 a There is a strong, non-linear re-
47 B 48 C 49 C 50 D 51 C
lationship between efficiency and
enthusiasm.
1
Exercise 16B b log y, , x2
y
c log(efficiency) = 0.0205 + 0.0860 ×
enthusiasm
1 a mean = 54.042, stand dev = 2.717 d 6.6
b z = −1.1 5 a
c i 48.6 kg ii 2.5%
Q1 Q2 Q3 Q4
2 a EV: number of distractions, RV: time
SI 1.01 1.15 1.32 0.52

16B 834 Answers
 
Year Q1 Q2 Q3 Q4 2791.33
b  
2022 62 60 61 63 ii S 50 = 2577.88
 
1830.79
 
6 a 130.6 cents/litre
Answers
iii The long term customer numbers

b 141.1 cents/litre
are: HSL (2791), Radcliffs (2578)
c i Victoria, slope = 2.09. On average,
Cottonworths (1831)
the price of fuel in Victoria is 
0
  
0.9 0.7   0
increasing by 2.09 cents/litre each    
11 a L = 0.9 0 0  b S 0 = 100
year.    
0 0.8 0 0
 
ii NT, slope = 2.20. On average, the
     
price of fuel in the NT is increasing 90 56 73
     
by 2.20 cents/litre each year. c S 1 =  0, S 2 = 81, S 3 = 50,
d i 156.7 cents/litre.
     
80 0 65
ii 169.7 cents/litre.  
91
e The difference is predicted to increase  
S 4 = 66;
over time. The cost of petrol in the NT  
is already higher than the cost of petrol 40
number of cases is increasing and
in Victoria, and the cost is increasing
spreading through the three stages
at a higher rate in the NT (on average    
2.20 cents/litre each year) than it is 5309 5957
S 40 = 4258, S 41 = 4778, 1.122
   
increasing in Victoria (on average 2.09 d
   
cents/litre each year). 3036 3406
e Not sufficient to eradicate disease;
Recusion and financial modelling
growth rate after 40 weeks is
7 a $6468.13 b 6.2% approximately 1.027
c i $5225 ii 9% f Sufficient to eradicate disease;
8 a $39.23 b 6.42% growth rate after 40 weeks is
c V0 = 2600, Vn+1 = 1.0025vn + 140 approximately 0.94
9 a $79 b $344.78 
29.6
 
96.11

c $375.48, $375.60 
14.4
 
55.97

   
Matrices 14.4 48.88
12 a S 1 =  , S 40 =  ,
   
10 a 1 × 3  21.6
 42.69
 150 250 320  
 6.4
 
33.15

 
b Q =  300
   
500 640 
0.0 19.30
    

1050 1750 2240
  
99.04
The number of shoppers who shopped 
 
57.67
from several sections at Radcliffs.  
50.37
c T ×A   S 41 = 
 

2835 43.99

   
d i S 1 = 1940 34.16
 
  
19.89

2425
 

Answers 835
Answers
   
715 735.9 iii A, E, F, H
   
416 429.0 iv 18 months
   
363 
374.4
 16 a A − B − D − E − C − A
c S 0 =  =
 
, S
  
1
b B − C − E − D − E − C − A − B − D;
  
317  326.7

   
247 Must start and end at a vertex of odd
253.6
 
 
    degree.
16B
144 148.2
  
c 9:54 am
13 a 3 × 1
17 a i 4
b C × A = [17900]; the total cost of the
ii
seats.
A Office
c The product = [11600]; the total cost 20 20
of the stalls and dress circle. 35
60 20 30
Networks 20
B 30
14 a i 40 40 15
1
45 35
C D
2 30
iii A, B, D
b i
A 20 Office
3 20
5
4
20
ii 1 + 4 + 1 + 2 + 2 = 10
40 B
b i Vertices D and E are odd. 15
40
ii E and F
D
iii E − F − D − E − A − B − C − D C
30
c i Capacity = 20 + 25 + 30 = 75
ii $44 400
ii Maximum flow = minimum cut
c i A Office
= 15 + 15 + 30 = 60 20
15 a i 2
ii C 20
B
b A on breastroke, 15
B on backstroke,
C on butterfly D
C
c i 7 months 30
ii ii $20 400
E 8 10
D 8 8 D,7 G 15 15
18 a
C,2
E, 10
C 6 6
G,4 G, 4
B, 8
A,3 E,9
A, 5 D, 15
A 0 3 dummy,0
I,2 21 21 H, 1 I, 3
I 19 19
H,9 C, 3
B 0 0 B,6 F, 6
F 6 7 F,3
H 9 10 b 32 weeks
16B 836 Answers
c i d i The project would be completed in

a minimum of 30 weeks.
Activity Duration EST LST Float
ii Nothing. This activity has a float of
A 5 0 0 0
Answers
14 and so it could be extended in

B 8 5 5 0 duration by 14 weeks.
C 3 5 19 14 iii The project would be completed in
D 15 13 13 0 a minimum of 37 weeks.
E 10 13 15 2
F 6 8 22 14
G 4 23 25 2
H 1 28 28 0
I 3 29 29 0
ii A − B − D − H − I


General Mathematics VCE Units 3&4 Second Edition

Uploaded by

Copyright:

Available Formats

General Mathematics VCE Units 3&4 Second Edition

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

General Mathematics VCE Units 3&4 Second Edition

Uploaded by

Copyright:

Available Formats

GENERAL

CAMBRIDGE SENIOR MATHEMATICS VCE

CAMBRIDGE SENIOR MATHEMATICS VCE

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Introduction and overview vii

1 Investigating data distributions

2 Investigating associations between two variables. . . . . . . . . . . 103

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

3 Investigating and modelling linear associations

3B Using the least squares regression line to model a relationship

5 Investigating and modelling. time

6 Revision: Data analysis

6B Exam 1 style questions: Associations . . . . . . . . . . . . . 316

7 Modelling growth and decay using recursion . . . . . . . . . . . . . 333

8 Reducing balance loans, annuities and investments

pound interest investments with additions to the principal . . 392

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

8B Using recurrence relations to analyse and model reducing

9 Revision: Recursion and financial. modelling

10 Matrices 10A What is a matrix? . . . . . . . . . . . . . . . . . . . . . . . 473

10B Using matrices to represent information . . . . . . . . . . . 485

11 Transition matrices and Leslie matrices

12B Exam 2 style questions . . . . . . . . . . . . . . . . . . . . 602

13 Graphs, networks and trees: travelling

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

13C Exploring and travelling . . . . . . . . . . . . . . . . . . . . 624

14 Flow, matching and scheduling

15 Revision: Networks and decision. mathematics

16 Revision of Chapters 1–15 . . . . . . . . . . . . . . . . . . . . . . . 741

Online appendices accessed through the Interactive Textbook or PDF Textbook

Appendix A Guide to the TI-Nspire CAS calculator in VCE mathematics

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Separate revision chapters containing Multiple-choice (Exam 1-style questions) and

The Review section at the end of each chapter includes:

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Numbers refer to descriptions above.

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

INTERACTIVE TEXTBOOK POWERED BY THE HOTmaths

WORKSPACES AND SELF-ASSESSMENT

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Overview of the Online Teaching Suite powered by the

More about the Exam Generator

Filtering by question-type, topic and degree of difficulty

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Cover: c Getty Images / m_pavlov

Images: c Getty Images / OsakaWayne Studios, Chapter 1 Opener / Andriy Onufriyenko,

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Study mode Fitness level

height (in centimetres) study mode (C = on-campus, O = online)

Categorical variables classify or name a quality or attribute– for example, a person’s

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Nominal variables have data values that are simply names.

ISBN 978-1-009-11041-9 © Peter Jones et al 2023 Cambridge University Press

Comparing numerical and categorical variables

Example 1 Types of data

Classify the following data as nominal, ordinal, discrete or continuous.

Types of variables: categorical or numerical

Categorical variables: nominal or ordinal