A Handbook MBA Thesis PDF

A Handbook on Thesis Writing
MBA Program
Graduate School of Business
Assumption University
by
Patricia Arttachariya, Ph.D.
A Publication of the Graduate School of Business, Assumption University, Bangkok,

Thailand, 2005.
PREFACE
The primary objective of this Thesis Handbook along with the accompanying
appendixes containing guidelines for content, style, and process is to assist those
presently enrolled or contemplating enrollment for a thesis (Plan A) in the Masters of
Business Administration program at Assumption University. The thesis is intended to
provide the student with an opportunity to synthesize and apply the knowledge gained
from his or her related coursework while at the same time affording him or her to
focus on a particular area of interest. It also affords the student a way to generate a
level of expertise not otherwise possible by delving deep into an area of specific
interest.
There is no chance that all of the thesis questions that you might have will be
answered here, but a thorough reading of this handbook should provide you with
some pointers in the right direction. Additionally, I hope this serves as a useful
reference throughout the process. Please read it carefully and particularly note all the
forms and other requirements you are supposed to complete. Finally, the guidelines
are reviewed regularly and occasionally change, so please be sure that you have a
current copy of the handbook.
Dr. Patricia Arttachariya,

Chairwoman, Thesis Committee.
ii
TABLE OF CONTENTS
Preface
Chapter
1
Page No.
Basics of Technical Writing and Thesis Writing

Key Features of Technical Writing
Three elements of technical writing
Three general stages of technical writing
Key Features of Thesis Writing
Definition of a thesis
Purpose of a thesis
The Research Process
Nature and limitation of a thesis
Practical Guide in the Selection of a Thesis Topic
Criteria in choosing a thesis topic
Three demands for a thesis
Characteristics of a good thesis research problem
Preliminaries or Front Matters
Thesis Title Page
Committee Approval Sheet
An Abstract
Acknowledgments
Table of Contents
List of Tables
List of Figures
Generalities of the Study
Introduction of the Study
Statement of the Problem
Research Objectives
Scope of the Research
Limitations of the Research
Significance of the Study
Definition of Terms
Review of Literature and Related Studies
Brief introduction of the chapter
Definition and Features of the Independent Variable (s)
Theories Related to the Independent Variable (s)
Critical Analysis/Discussion of the Theories Related to the
Independent Variable
Discussion on the Dependent Variable
Relationship of the Independent Variable to the
Dependent Variable
Previous Studies
iii
1
1
1
1
2
2
2
3
4
4
4
4
5
6
6
6
6
7
7
7
7
8
8
8
9
9
9
9
9
10
10
10
10
11
11
11
11
10
11
12
Research Frameworks
Theoretical Framework
Conceptual Framework
Research Hypothesis
Operationalization of the Independent and Dependent Variables
Example of a section of the Operationalization Table
Research Methodology
Methods of Research Used
Respondents and Sampling Procedures
Research Instruments/Questionnaire
Collection of Data/Gathering Procedures
Statistical Treatment of Data
Presentation of Data and Critical Discussion of Results
The First Independent Variable
The Second Main Independent Variable
The First Main Dependent Variable
Summary Findings, Conclusions and Recommendations
Summary of Findings
Discussion/Conclusions
Recommendations
Further Research
References and Back Matters
Appendixes
The use of references
Bibliography and references
Choosing the Right Statistics
Simulations
Research Situations and Solutions
Statistical Requirements
Level of Measurements
Characteristics of Choosing Correct Statistical Test
Definition of Terms in Statistics
Choosing the Correct Statistical Test for Nominal Data
Choosing the Correct Statistical Test for Ordinal Data
Choosing the Correct Statistical Test for Interval Data
Kinds of Conceptual Frameworks
Appendixes
Appendix A: Physical Specifications
Appendix B: Samples
Appendix C: Forms and Letters
Appendix D: Frequently asked Questions
References
iv
12
12
12
12
13
14
15
16
16
16
19
21
22
22
23
23
23
23
23
25
25
25
25
26
26
27
27
27
27
35
35
51
60
63
64
65
67
68
69
70
74
75
78
83
95
98
1
BASICS OF THESIS WRITING
Thesis Writing is a requirement for graduating students particularly those who opted
for Plan A (Thesis) instead of a comprehensive examination.
This handbook focuses on the mechanics on how to develop a thesis and is made
simple for easy understanding of the processes. Many students are confronted with
tasks that demand good writing skills. They need guidance on how to go about the
complex job of writing a thesis or dissertation. This book will therefore serve as a
guide.
Thesis Writing is subsumed under Technical Writing. All professional fields
require technical documents that help readers perform tasks and understand specific
tasks. As mentioned by some authors, a thesis is classified under technical writing as
it provides vital information to specific readers.
Technical Writing refers to the provision of specific information to clearly identified
readers who will use the information for a specific purpose.
KEY FEATURES OF TECHNICAL WRITING

Three Elements of Technical Writing.
The elements are reader, purpose and writing situation. The reader seeks information
for a specific purpose. The purpose is to design a document that will serve the
reader's need and help the reader understand the information. The writing situation
consist of both reader and purpose, as well as such factors such as the sponsoring
organization's size, budget, ethics, deadlines, policies, competition and priorities.
Three General Stages of Technical Writing.
The stages are planning, multiple drafting and revising/editing. In the Planning
stage a writer analyzes the reader, purpose and writing situation; gathers information
and tentatively organizes the document. All these may recur several times during the
writing.
Multiple drafting a rough draft is the next step. It focuses on thoroughly
developing the information you have gathered without worrying about grammar,
punctuation, spelling and fine points of style. At this stage, your main focus is on the
content and organization. Follow your initial plan of organization and write quickly.
Your document may require many drafts from the beginning to the end.
Revising and editing it takes place throughout the writing process but particularly
after you have begun drafting. Read your draft and rethink the following elements:
content - do you need more facts? are your facts relevant for the reader and
purpose?
organization - have you grouped the information into topics? have you put the
details in an order that readers will find easy to understand and use?
headings - have you written descriptive headings that will guide your readers to
specific information?
opening - does your opening establish the documents purpose and introduce the
main topic?
closings - does your closing provide a summary, offer recommendations or
suggest actions?
graphic aids - do you have enough visual aids to understand the data? are they
appropriate?
language - have you used appropriate language? too much technical jargon? have
you defined terms your readers may not know?
reader usability - can your readers understand and use the information
effectively? does the document format help your reader find specific information?
After you are satisfied, then the final editing will focus on the form and style , such
as, grammar, punctuation...and others and finally proof reading.
KEY FEATURES OF THESIS WRITING

.
Definition of a Thesis
A Thesis is an idea or theory that is expressed as a statement, a contention for which
evidence is gathered and discussed logically (Turabian, 1987). It normally represents
the culmination of a substantial piece of original work over a period of at least one
year.
Purpose of a Thesis
The Thesis is one option for fulfilling the final requirement for the Master of Business
Administration degree at Assumption University. In general, a thesis involves
formulating an original idea or area of inquiry which is either quantitative
(e.g., typically involves either an empirically based, provable hypotheses) or is
qualitative (e.g., includes explorative outcomes, along with data collection and
analysis). It is expected that a thesis contributes one or more of the following for it to
be deemed acceptable: a new perspective, a new application, uncover new

implications or make fact-based predictions. The central part of the thesis proceeds to
methodically substantiate a researcher's position. A disciplined approach to
conducting research is the basis for any acceptable thesis. The thesis must be the
students original work and should be based on factual data, which may be
quantitatively or qualitatively derived and verifiable. To do an acceptable job of
analysis, you will have to have laid a solid foundation based on methodical and
exhaustive research.
Thesis writing is demanding and challenging work for students as it unfolds and is
gratifying once it is finished. It is a tough test of intelligence and endurance.
The Research Process.

3. Conceptual Framework
(Theory, Literature)
6. Data Analysis
2. Proposition
4. Research Questions
and Hypothesis
1. Empirical Observation
5. Data Collection
Inductive
Deductive
The Research Wheel

The Research Process involves the following steps:
1. Empirical Observation the entry point is an infinite array of possible topics.
2. Proposition the inductive logic serves to relate specific topic to a broader context
and begins with hunches wonder if.
3. Conceptual Framework it consists of statements that link abstract concepts to
empirical data. A Theory allows us to move from observation to observation and
make sense of similarities and differences.
4. Research Questions and Hypotheses based on the conceptual framework, the
questions and hypotheses can be formed.
5. Data Collection refers to the gathering of information relevant to the subject
under study.
6. Data Analysis data collected are analyzed and the results/findings are interpreted
as bases for further action and decision making.
Nature and Limitations of a Thesis
Experts have reached no absolute agreement regarding details of form in thesis
writing. In a any particular situation, several forms may be acceptable. Some
universities require students to follow special forms quite different from those
elsewhere. An example: A department of sociology at a particular university might
require students to use a particular referencing technique that would not be acceptable
to the department of history at the same university. Regardless of a particular system
adapted, students may have the option to choose which ever is applicable in their
study.
PRACTICAL GUIDE IN THE SELECTION OF A THESIS TOPIC
Criteria in Choosing a Thesis Topic
The following are some recommended criteria in the selection of a title: Using
personal experience for topic discovery reflecting on your personal experiences for
a topic that touches your life-style or career; talking with others to find a subject talk
with other people because collaborative learning can broaden your vision of the
issues. Reading further on your subject to discover ideas speculate about the subject
and discover ideas by listing issues, asking questions, free writing and other
techniques.
Three Demands for a Thesis
It must examine a significant issue; It must address a reader and carry that reader to
another plateau of knowledge; It must have a serious purpose, one that demands
analysis of the issues, argues from a position, and explains complex details.
Characteristics of a Good Thesis Research Problem

According to (Turabian, 1987), there are ten important characteristics of a good
research problem for a thesis:
The problem can be stated clearly and concisely. It is tested by writing it as a

concise sentence or paragraph and sharing it with others.
The problem generates research questions. It refers to the formulation of
specific questions which represents the various aspects of the problem.
It is grounded in theory. Good problems have theoretical and or/conceptual
frameworks for their analysis. It relates to the specifics of what is being
investigated to a more general background of theory which helps the results and
links it to the field.
It relates to one or more academic disciplines. It should be based on
disciplines, such as, sociology, psychology or management science or have clear
links to one or two disciplines.
It has a base in the research literature. It often relates to a well-defined body of
literature written by a selected group and published in journals to establish
connections.
It has potential significance or importance. The problem must have importance
to the researcher and the others as well.
It is do-able within the time frame and budget. Logistic factors are needed to
carry out a thesis.
Sufficient data are available or can be obtained. Data to address the problem
should be accessible. There may be some restrictions on the environment and
some other factors.
The researcher's methodological strengths can be applied to the problem.
Some problems are related to standard methodology. The researcher should have
the inclination toward the methodology to be used
The problem is new; it is not already answered sufficiently. Once the field is
known, it becomes clear what has been done and what needs to be done.
2
PRELIMINARIES OR FRONT MATTERS
This chapter presents the chronological sections of a Thesis and its contents. It
contains information on the title page, approval sheet, abstract, acknowledgments,
table of contents, list of tables and list of figures.
The following sections make up the front matter of the thesis:
i. Thesis Title Page

It is usually the first page of the Thesis. It includes the following: Title: It should be a
concise statement of the main topic and should identify the actual variables or
theoretical issues under investigation and the relationship between them. It should be
fully explanatory when standing alone. The recommended length for a title is from 10
to 12 words. It is centered at the top third of the page. Most universities and colleges
have their own style of title page for theses, and this should follow exactly in matters
of content, capitalization, centering and spacing. Author's name and institutional
affiliation: The preferred author's name is first name, middle initial and last name
because this reduces the likelihood of mistaken identity. Use the same form for
publication throughout your career; that, is, do not use initials on one manuscript and
the full name on a later one. The affiliation identifies the location where the author
conducted the investigation, which is usually an institution. It is centered at the middle
of the page; name, title and university/school of the writer, centered at the bottom
third of the page and date of the research/report, centered directly below the writer's
name. The title page is not listed in the table of contents but it is counted as page i.
ii. Committee Approval Sheet
It is the page where the members of the Committee of the Oral Defense sign.
iii. An Abstract
It is a brief, comprehensive summary of the contents of the study. It allows readers to
survey the contents of a study quickly. A well-prepared abstract can be the most
important section of the study. It is a synopsis, of the most important points in a report
and provides readers with a preview of the full contents. It can be specified in 200
words. It consists of a short statement of the problem, a brief description of the
methods and procedures adopted and a condensed. summary of the findings of the
study.
iv. Acknowledgments
Citations accorded to the persons who in one way or another had helped in the
realization of the study. This includes names of individuals or companies/agencies
whom the researcher was indebted to, such as reading materials, facilities used or
financial support to the researcher.
v. Table of Contents
It alerts the reader to pages that contain specific topics, the overall organization and
content of the research and specific and supplemental materials, such as appendices.
All front matter for a thesis paper is numbered in small Roman numerals. The first
page of the thesis proper is numbered as page 1 in Arabic numbers, and all pages after
will have Arabic numbers.
vi. List of Tables
The list of tables appears directly after the list of figures. List each table by number
and title and indicate page numbers. In a list of tables, the table numbers (in Arabic,
followed by a period) are placed in a column left under the heading "Table," and
the page numbers are listed right under the heading "Page." The table titles should
begin two spaces after the period following the table number and should agree exactly
with the wording of the titles as they appear above the tables themselves. The titles
are capitalized in either sentence or headline style, and run-over lines are indented
three spaces. Double-space between items, single-space within.
vii. List of Figures
Any graphic aid, such as bar graph, map or flowchart that is not a table with numbers
or words in columns is called a figure. The list of figures follows the table of contents.
List each figure by both number and title and indicate page numbers. The figure
numbers in the list are aligned by their periods under the word "Figure" and page
numbers are listed right under the word Page,". Captions are capitalized headline
style.
3
CHAPTER 1
GENERALITIES OF THE STUDY
The following are the suggested format and contents of the Thesis as outlined below:
A new form and style was introduced (such as center and side headings) to guide the
readers how to go about the formulation of the different chapters of the thesis.
1.1
Introduction of the Study
In paragraph form, this section includes the following information: 1. Global Context
- this portion describes the situation associated in the global context of the study, in
general, the world situation. How does your study link with the current problems of
today? Does the research focus on groups, organization or an individual? 2. National
Context - state how the country or nation is currently experiencing the present
problem and with which your study has some kind of relatedness 3. Local Context state how the study relates or is relevant to the area in focus. 4. Overview/history of
the Organization (subject of the study) - make brief statements for the company; it
should also include in brief some information on the main topics of the study of which
the organization is a focus.
1.2
Statement of the Problem
The statement of the problem is the focal point of your research. It is just one sentence
(with several paragraphs of elaboration).
You are looking for something wrong.
....or something that needs close attention
....or existing methods that no longer seem to be working.
Example of a problem statement:
"The frequency of job layoffs is creating fear, anxiety, and a loss of productivity in
middle management workers."
While the problem statement itself is just one sentence, it is always accompanied by
several paragraphs that elaborate on the problem. Present persuasive arguments why
the problem is important enough to study. Include the opinions of others (politicians,
futurists, other professionals). Explain how the problem relates to business, social or
political trends by presenting data that demonstrates the scope and depth of the
problem. Try to give dramatic and concrete illustrations of the problem. After writing
this section, make sure you can easily identify the single sentence that is the problem
statement.
1.3
Research Objectives
In paragraph form, this refers to the general objective of the study, It simply indicates
the intent or purpose why the study is conducted. It must be stated in declarative
sentences or in operational terms what the research seeks to achieve: what to find out?
to determine ..., to identify ..., to assess Follow the SMART Formula. SMART
means S-specific, M-measurable, A-attainable, R-realistic and T - time bounded.
1.4
Scope of the Research
In paragraph form, include the following information: Topic in focus state the
different main variables of the study. State what are the main independent and
dependent variables; Target respondents - state who are your main respondents. Are
they Managers? Employees? or Middle Managers? The area in focus refers to the
coverage of the study whether it is nationwide, regional, provincial or district-wide,
where the subjects can be located.
1.5
Limitations of the Research
It implies limitations on the research design that you have deliberately imposed. It
restricts the population to which the results of the study can be generalized.
Limitations refer to the restrictions over which you have no control. The extent/or
magnitude to which the study covers the topics or issue. What can be done...or cannot
be done.
1.6
Significance of the Study
This part should explain the relevance or usefulness of the study. Who would be
benefited? What importance does it generates for certain individual, groups or
organization. It would address the 3 W's (what, who and why) plus how?
1.7 Definition of Terms
.
This section would define the terms, which are either in operational form or literal
meanings to express clarity and correct understanding of some terms with which
readers are not familiar. It should contain a brief and concise definitions about two or
three sentences. It should define all terms, which are presented in the conceptual
framework. and other relevant terms included in the discussion of the text.
Thesis Writing Made Simple
4
CHAPTER 2
REVIEW OF RELATED LITERATURE AND STUDIES
Chapter 2 is a review of the literature. It is important because it shows what previous
researchers have discovered. It is usually quite long and primarily depends upon how
much research has previously been done in the area you are planning to investigate. If
you are planning to explore a relatively new area, the literature review should cite
similar areas of study or studies that lead up to the current research. Never say that
your area is so new that no research exists. It is one of the key elements that proposal
readers look at when deciding whether or not to approve a proposal.
Words that describe similarities/differences are: supports, presents, confirmed,
illustrates, shares, positive relationships, revealed, pointed, argued, concluded,
negative relationships and contrasting results. It is necessary to show how the problem
under investigation relates to previous research studies. It is important to locate the
problem within a theoretical framework and in such cases, the theory needs to be
reviewed as well. It would give the researcher a wider range of ideas as to the
similarities in the studies which will contribute to the conceptualization of the
research paradigm. This chapter should include the following information:
Brief introduction of the chapter What would be the contents of this chapter?
What topics are to be discussed in detail?
2.1
Definition and Features of the Independent Variable
Suppose your Independent Variable is Leadership. First, provide definitions of

Leadership. Cite at least two to three definitions from authorized authors; discuss the
role of Leadership. It includes other topics related to Leadership, such as,
classification of leadership; the distinction between a leader and a manager, and other
related topics of the variable.
2.2
Theories Related to the Independent Variable (s)
In this section, describe the major theories written on Leadership. For instance,
Theory of Hershey and Blanchard (199-); Theory of Mintzberg (199-); Theory of
Yukl (199-); Theory of Page (199-); Theory of McGregor (199-). Document the main
points of the theory. The authors should be presented in sequence as to year, from the
traditional to the contemporary. Also present graphs or figures explaining the theories.
Present at least five or more theories per variable. If you have two or more
independent variables, the format for discussion must follow the same mode as cited
above.
10
2.3
Critical Analysis/Discussion of the Theories Related to the Independent

Variable
This is a useful exercise whereby you can compare the findings of several authors
who have written on your Independent Variable.
Tabular presentation/analysis/comparison of theories/authors and their subVariables in one big table, indicate all the authors/theories side by side and analyze
by looking for similarities and differences of each theory as to sub-variables.
2.4
Discussion of the Dependent Variable: As mentioned in 2.1, document the
main points of the theory. The authors should be presented in sequence as to year,
from the traditional to the contemporary. Also present graphs or figures explaining the
theories. Present at least five or more theories per variable.
2.5
Relationship of the Independent Variable (s) to the Dependent

Variable:(Related Studies coming from Journals, Magazines, Theses,
Dissertations & Internet)
For instance: A Study on Leadership by (Adams, 1997) a study which relates

Leadership with Job Performance.
Note: If there are two or more independent and dependent variables, the process
would follow the same sequence as above.
2.6 Previous Studies: In the final section of Chapter 2, it is necessary that empirical
work related to the topic under study is cited. In general, a previous study must briefly
discuss the 4 standard sections:
o
Introduction & background introduces the research topic and

question
Methods how the study was conducted
Results summary of the data that was collected
Discussion Discusses the implications and significance of the

findings and relationship of results to a theory.
11
5
CHAPTER 3
RESEARCH FRAMEWORKS
This chapter discusses the Research Frameworks. It included the theoretical

framework, the conceptual framework, research hypotheses and operationalization of
the variables. This chapter of the thesis will include the following:
Brief introduction of the chapter in one paragraph, specify what will be the
contents of this chapter.
3.1
Theoretical Framework
The Theoretical Framework refers to the theories being used as a basis or reference
for the study which are drawn from the literature. It contains the major ideas of
authors or theorists, experts and specialists. It is a big map of ideas in global terms.
A framework is a model, which allow the researcher to explore the relationship
of variables in a logical and prescribed fashion. It clarifies questions and it
summarizes the overall concept being investigated.
It should include the following information: Link the theories and studies on which
you pattern your conceptual framework; Show specific models from which most of
the variables are taken; Write the main variable and its sub variables and summarize
the authors together with the year, which you are employing in your framework.
3.2
Conceptual Framework
It is the specific map of ideas of your topic or study. In this section, the researcher has
to explain the research framework (model) together with the explanation of the
variables. It is a framework which a researcher designs, and conceptualizes for the
specific needs of the study. The development of a suitable framework is part of the
process of planning, clarifying the research problem and conducting the analysis.
There are two different kinds of frameworks which are popularly used by researchers:
One - dimensional framework: The starting point for the research is often the
formulation of a one-dimensional framework. Most common is the pre-test-post test
experiment. Something is measured before an educational intervention and again
afterwards. Another example: Input-process-output model. It is popularly known as
the Black box model.
12
Two-dimensional framework: permits analysis of the interrelationships between

sets of related variables. It is formed by dividing the data sample into groups with
same characteristics. It would suggest various types of inputs, which bear logical
relationships to components of institutional development.
Present your conceptual framework showing the independent and dependent
variables; Discuss in brief the main variables and sub-variables and the theory (ies)
on which they are patterned.
3.3
Research Hypotheses
A hypothesis is a claim or statement either about a value of a single population

characteristics or about the values of several characteristics.
A test of hypothesis is a method for deciding which of the two contradictory claims is
the correct one. In carrying out a test, we initially assume that a particular one of the
two is the correct one. This claim will be rejected in favor of the second (alternative)
claim if sample evidence is incompatible with the initial assumption.
The null hypothesis, denoted by Ho, is the claim that is initially assumed to be true.
The other hypothesis is referred as the alternative hypothesis and is denoted as Ha.
In carrying out a test of Ho versus Ha, the hypothesis Ho will be rejected in favor of
Ha only if sample evidence strongly suggests that Ho will not be rejected.
How are hypotheses stated? Hypotheses should synchronize with the question posed
in the statement of the problem, particularly focusing on the relationship of variables.
The section on hypotheses should include the following information:
List all the statements/hypotheses which you wanted to prove in the study.
There are only two hypotheses that can be statistically tested the hypothesis of
difference and the hypothesis of association. Whenever the research is
experimental, then the hypothesis of difference is the one that must be tested. This
hypothesis states that the populations from which the sample groups have been
selected are in some way different from each other. If however the research is postfacto, then the hypothesis under scrutiny might be one of either difference or
association.
The hypothesis of association states that a correlation exists in the population
from which the sample has been selected. The correlation may exist between
different measures taken on the same group of subjects (for example, a single group of
subjects being measured on both height and weight) or between the same measure
taken on different subjects (for example, obtaining IQ scores from pairs of identical
twins). Testing the hypothesis of association requires different statistical tests than
does testing the hypothesis of difference.
If the Hypothesis of Difference has been tested, are the samples independent or
correlated? Whenever the hypothesis of difference is tested, whether in experimental
or post facto research, it must be clearly determined whether the sample groups are
independent or correlated. If the selection of one sample is in no way influenced by
13
the selection of another, then the samples are independent. This occurs when each
sample is randomly selected. If on the other hand, the subjects to be measured are
in any way paired off, either by using the same subject more than once or by
equating subjects on the basis of some relevant variable then the groups are
correlated. Attempting to isolate differences between correlated sample measures
requires different statistical tests than when analyzing differences between
independent sample measures.
3.4 Operationalization of the Independent and Dependent Variables
The construction of actual, concrete measurement techniques; the creation of
operations that will result in the desired measurements. The development or choice
of specific research procedures (operations) that will result in representing the
concepts of interest.
An operational definition is a procedure for classifying, ordering, or quantifying
something
Classifying - crowded or not crowded

Ordering - uncrowded, mildly crowded, severely crowded
Quantifying - measure crowdedness in terms of the number of residents per square
kilometre.
Levels of measurement:
A zero point
Distances between categories equal
Categories can be rank-ordered
Nominal measures
Reflects only categories
The variable gender has two attributes, male and female
They are distinct from one another, but they have no additional
structures
Also political party affiliation, birthplace, etc.
Ordinal measures
Include the above
Plus we can logically rank-order the attributes
e.g. social class, prejudice
14
Interval measures
Include the above
Plus the intervals between the attributes have meaning
e.g scores on an intelligence test
zBUT:
We cannot say that a person with an IQ of 120 is TWICE as
intelligent as one with an IQ of 60
Ratio measures
Include all of the above
Plus they have a true zero point
The following is an example of a section of the Operationalization Table:
Concept
Conceptual
Operational Component
Definition
Level of
Measurement
Information
- External
Q1: Read available information
Sources
information
from printed media (such as
sources
brochures, pamphlets, point-ofsale display, others information

given by service providers, thirdparty report, magazine,
newspaper)
Q2: Pay attention to advertising
through TV and radio
15
Ordinal scale
6
CHAPTER IV
RESEARCH METHODOLOGY
Brief introduction of the chapter what would be the different sections to be

discussed in the chapter.
4.1
Methods of Research Used
( Note: In the thesis proposal, use future tense and at the finalization stage use past
tense). Select a research method and explain how you will employ it. Will you use
Experimental or Post Facto? What will be the research design?
The following are Research Methods used by thesis and dissertation writers: (From
among the methods, which would be applicable to your research?)
Historical Research (DR) it is concerned with natural behavior in a real situation
and the focus is on interpretation of what it means in the context. The historian
attempts to discover data that already exists in some form. Problems arise from
personal interest, which are kindled by exposure to a person, event or logical source
of unused original data.
Descriptive Research (DR) refers to as any approach that attempts to describe data.
It is either quantitative or qualitative. The use of descriptive statistics is merely a
convenient way of description. Data are reported in tables organized to give a suitable
overall picture at a glance. They are presented in matrices or frameworks, which
convey data characteristics for sub-groups or different cells in the framework.
Summary statistics are used such as means, standard deviations and other measures of
normalcy. It also includes tracer studies - the tracking down of people who were
participants in prior educational experiences. (studying phenomena as attrition and
brain-drain). It also covers sociometry. It assesses social structure in a group of
individuals by ascertaining the status of each individual in the group. Sociograms are
used to report findings of sociometry.
Experimental and Quasi-Experimental Methods (EM). It is assumed that the
experimental method is the only method of research that can truly test for cause and
effect relationships. The researcher manipulates at least one variable, controls some
variables and observes the effects. The experimental variable that is manipulated is
called Treatment and is presumed to be one making the difference.
Example:
R
16
The top line experimental group, which receives a treatment, is X. Subjects are
randomly assigned to this experimental group or to control group at the bottom line
(randomization denoted by R); O indicates observations generally referred to as pretests and post-tests. Comparisons between the test scores of the groups theoretically
indicates the effect of the treatment. The fundamental principle of experimental
design is that the participants be randomly assigned to experimental and control
groups.
Quasi-experimental methods are those where there is also an artificially manipulated
treatment, but where randomization is not possible. It is often called as the "Nonequivalent control group design".
Example:
O
O
O
O
The participants have been randomly assigned to receive the treatment or not. If we
looked at the changes between pre-test and post test scores, we can make comparisons
of experimental and control group.
Example:
Children who attended French Immersion it means IQ for similar groups higher
educational aspirations may be strictly the result of their home environment and have
less to do with the nature of their educational experience. Experience are formulated
by 10 test hypotheses derived from theory and prior research. It is a useful tool for
establishing generalizations.
Correlational Research. It involves the calculation of a correlation coefficient,
which is a measure of the extent to which variables vary.
Correlation Coefficients generally range -1.00 to +1.00
-1.00
Perfect negative correlation
- .95
Strong negative correlation
- .50
Moderate negative correlation
- .10
Weak negative correlation
.00
No correlation
+ .10
Weak positive correlation
+ .50
Moderate positive correlation
+ .95
Strong positive correlation
+1.00
Perfect positive correlation
Correlational Research is one way of describing in quantitative terms, the degree to
which variables are related. It investigates a number of variables, such as independent
variables or predictor variables or criterion variables. Multiple Correlation combines
two or more independent variables to enhance the relationship with a dependent
variable. The correlation coefficient thus described denotes a linear relationship
between variables.
17
There are two major approaches to addressing research questions using correlational
methods: "Looking and Looking for."
Example: Battery of Tests
Looking Approach generates correlation coefficients among all variables and then
examines them to see which ones are related. It is a sort of astrological approach to
research in which natural and unnatural happenings add credence to a theory of
generalities.
Looking for Approach is more rigorous as it involves making a prediction of
significant correlations based on theory of prior research. This is akin to the
hypothesis testing approach. If the theory is correct, then variable A should be related
to variable B. Multiple Correlation is the technique used for combining independent
variables and relating them to a dependent variable.
Ethnographic Research (ER). It is a research technique of direct observation of
human activity and interaction in an on-going and naturalistic setting. It aims to
discover and describe the culture in an educational setting. This approach emphasizes
inductive analysis, description and perception in the natural setting rather than the
concerns with measurement and manipulation characterized by the experimental
method.
Participant observation is one of the distinctive characteristics. It is inherently
descriptive. It takes the forms of words or pictures rather than numbers. Observers are
concerned with the process. They tend to go looking for something. There are two
basic approaches: first open research design little is known about the subject under
consideration. It is used to develop conclusions and hypotheses and to create an
overall picture of the situation. Second focused research used to study areas in
which a great deal is already known and there exists an already identified problem,
question and testable hypotheses.
Case Study (CS). It investigates a contemporary phenomenon, in real life contexts
when boundaries between phenomenon and context are not clearly evident and in
which multiple sources are used.
Some people have confused historical research with case study. Historical deals with
the past and case study with contemporary events and also with Evaluation.
Evaluation finds out what happened and its relationship with what was planned. Case
study is concerned with how things happen and why. It does not control by using
events. CS attempts to understand what is happening and link the contextual realities
to the difference between what was planned and what actually occurred.
Program Evaluation. Evaluation is one way of assessing whether or not what we are
doing is achieving what is intended to achieve. Program evaluation is to improve a
new program or activity. Formative evaluation its results are intended to feed back
and improve on-going practice while summative evaluation is directed at evaluating
the consequences of a program and is often seen as the principal mechanism for
making crucial resource allocation decisions about a program, including its
continuation or termination.
18
Policy Research - It explores social issues and problems within real life context,
attempting to account for the political and practical dimensions of a problem as well
as exploration of the technical dimensions of possible solutions. It often assumes a
"what if stance" Ex: What if we jailed all drinking drivers? It extends technical
analyses to existing contexts, which the researcher must understand sufficiently to be
able to forecast the effects of possible strategies and decision alternatives. It is firmly
grounded in the policy-making arena and is best tackled by those familiar with that
arena.
It is defined as the process of conducting research on or analysis of a fundamental
social problem in order to provide policy makers with pragmatic, action-oriented
recommendations for alleviating the problem. It explores a course of action not yet
put into practice. It is conducted by think tanks or institutes set up to consider social
issues. Example: Social/Behavioral Researchers use Descriptive Methods or
Experimental Research and other methodologies which are applicable to the nature of
the study.
4.2
Respondents and Sampling Procedures
Respondents of the Study

The following are the information to be included:
-Who are your primary respondents? How many levels of respondents do you intend
to use (2 or 3)?
-Who constitutes the respondents, how many per level of respondents?
-What is your target population or sampling frame?
-How many employees in the company? How many top level managers? middle
managers? Staff/rank & file?
-Out of your population or sampling frame, how many are your sample respondents?
-Is it total enumeration? Is it 75 %? Is it 30%?
Present the process on the determination of sample size. You can compute by using a
formula or use Andersons (1996) Table of Sample Size (see next page)
19
Table __ Theoretical Sample Sizes for Different Sizes of Population and a 95

percent level of certainty
Population
(Sampling Frame)
100
500
1,000
5,000
50,000
100,000
1,000,000
25,000,000
Required Sample for Tolerable Error

5%
4%
3%
2%
79
217
277
356
381
382
384
384
85
272
375
535
593
596
599
600
91
340
516
897
1,044
1,055
1,065
1,067
96
413
705
1,622
2,290
2,344
2,344
2,400
Anderson, G. Fundamentals of Educational Research, 1996 p.202.
Sampling Procedures
As to Sampling Procedures, what would be the ideal number of respondents in order
to get reliable responses for the study? What sampling strategies to apply? Simple
random sampling, systematic, stratified or cluster sampling?
Here are some Random Sampling Methods:
Simple Random Sampling each person has an equal chance of being selected for
participation and where each combination of participants is equally similar.
Systematic Random Sampling its employs intervals between prospective
respondents, such as taking every nth person on a list.
Stratified Sampling sub-dividing into groups and selecting a given number in each
of the groups.
Cluster Sampling dividing the target population into groups and then choosing the
groups from which to collect your data.
Here are some Non Random Sampling Methods:
Accidental Sampling it is based exclusively on what is convenient for the
researcher. He/she includes the most convenient cases in his sample and excludes the
inconvenient cases from his/her sample.
Quota Sampling diverse characteristics, such as age, sex, social class or ethnicity
are sampled in the proportions which they occupy in the population.
Judgment or Purposive Sampling the basic idea involved in this type is that logic,
common sense or sound judgment can be used to select a sample that is representative
of a larger population
20
4.3
Research Instruments/Questionnaire
This section discusses the structured instrument/questionnaire which contain the

questions to be asked of the respondents. All information with respect to what is asked
feed back to the statement of the problem. It consists of sections which will cover all
the variables used in the conceptual framework
Steps for Constructing Effective Questionnaires
Determine your research questions: what do you intend to find out; how will the
information be helpful; which issues will relate to the questionnaire.
Specify your sub-questions: list all the issues you want to find; indicate those sub questions to be included in the questionnaire and refine your list.
Draft the items: translate questions into items; formulate fill-in the blanks, multiple
choice, comments, lists, Likert and rank-order questions.
Sequence the items: group into topic sections; group question type and rewrite as
necessary.
Design the questionnaire: order and number questions; layout booklet formats and
arrange questions on pages.
Pilot test the questionnaire: The questionnaire should be administered to a group of
at least 20 30 respondents to identify any problems they have with completing the
questionnaire.
Measures of internal consistency estimate how consistently individuals respond to
the items within a scale. The internal consistency of the measurement scales within
the questionnaire should be assessed using SPSS to calculate Cronbachs Alphas for
each dimension and scale within the instrument. Cronbach's Alpha is the most
commonly used estimate of internal consistency of items in a scale. The Alpha
measures the extent to which item responses obtained at the same time correlate
highly with each other. Alpha is a measure of level of mean intercorrelation weighted
by variances, or a measure of mean intercorrelation for standardized data.
The widely accepted social science cut-off is that alpha should be .60 or higher for a
set of items to be considered a scale, but some use .75 or .80. An alpha is calculated
for the whole scale and for each item within the scale. An alpha is also given which
indicates what the overall scale alpha would be if a particular item is removed. If the
Alpha will be higher when an item is deleted, the researcher infers that that item is not
tapping the same construct as all of the other items and therefore it should be removed
from the scale.
21
4.4 Collection of Data/Gathering Procedures

This section discusses the methods that are used either to collect primary data
(through personal interview with the aid of a questionnaire or secondary data from
books, journals, articles, etc.) Is it through personal interview?, group or panel
discussion?, mailed questionnaire? or through the email/Internet?
The following information should be included: How did you collect your primary
data? How did you collect your secondary data? What is the action plan to collect
your data from the beginning to the end of the study. How will you retrieve your data?
Who will edit? encode? analyze up to its finalization?
4.5
Statistical Treatment of Data
This portion will present what statistics are to be used? Descriptive or inferential
statistics? Example ANOVA, Multiple Regression and/or others.
Using statistical computers for statistical analysis, such as Statistical Package for
Social Sciences (SPSS), Minitab and Statgraphics are generally used to treat the data
that has been obtained. Another tools or programs from personal computers are Stata,
GBStat, SysStat, Stat Pac Gold and others.
The above mentioned statistical tools are often used by colleges, universities,
government institutions and business organizations. However, the statistical treatment
will depend on the nature of the research study. If in doubt about what statistics to
use, please consult a statistician.
The following information should be included: What are the statistical tools that need
to be used to answer each of the questions on the statement of the problem and
hypotheses? What are the tables to be constructed/presented for the results? Will you
use SPSS and other computer software?
22
7
CHAPTER V
PRESENTATION OF DATA AND CRITICAL DISCUSSION OF
RESULTS
The chapter should start with a brief introduction as to its contents. The data should be
presented as outlined in the conceptual framework.
5.1
The First Main Independent Variable
Present all data relevant to the First Main Independent Variable. It should also
synchronized with the questions asked in the statement of the problem and hypothesis.
5.2
The Second Main Independent Variable
The same process as in 5.1 .

5.3
The First Main Dependent Variable
The same process as above.

Presentation, Analysis and Interpretation of Data
Because of the diversity of research topics within disciplines, it is not possible to
specify exact directions for this section of the thesis. Here are some general
principles: organize the presentation of the argument or findings in a logical and
orderly way, developing the aims stated or implied in the introduction; substantiate
arguments or findings; be accurate in documentation.
All data, which pertains to the study, should be presented chronologically and
analyzed with correct interpretation.
The results could be presented in tabular form, graphical presentations such as pie
charts, bar graphs, histograms and other visual aids for easier understanding.
Figures and Tables: Unless a set of findings can be stated in one or two numbers,
results that are sufficiently important to be stressed should be accompanied by a figure
or table summarizing the relevant data. The basic rule of presentation is that a reader
be able to grasp your major findings either by reading the text or by looking at the
figures and tables. Thus, figures and tables must be titled and labeled clearly and
completely, even if that means constructing a very lengthy title or heading. In the
text, the explanation or interpretation should come first before the table or figure. In
the interpretations: On relationships: state whether it is significant or not significant.
In the hypothesis testing: state the Statistical Rule or Rejection Rule. There are two
23
approaches used in interpreting the statistics: The use of the p-values or the test
statistic and its critical values. Choose one or both. If both, it should be explained in
separate paragraphs, tables or figures. Do not mix data or values in one big or lengthy
table.
An example of a Table:
Table 5. Chi-square test of difference between DTAC mobile phone users
switching behavior and sources of information search
Chi-Square Test
Monte Carlo Sig. (2-sided)

Value
df
Asymp. Sig.
(2-sided)
95% Confidence Interval

Sig.
Pearson Chi-Square
Likelihood Ratio
Fisher's Exact Test
Linear-by-Linear
Association
57.879(a)
70.246
57.700
20
20
.000
.000
.000(b)
.000(b)
.000(b)
Lower
Bound
.000
.000
.000
1.165(c)
.280
.271(b)
.262
N of Valid Cases
400
a 20 cells (47.6%) have expected count less than 5. The minimum expected count is .46.
b Based on 10000 sampled tables with starting seed 2000000.
c The standardized statistic is -1.079.
24
Upper
Bound
.000
.000
.000
.280
8
CHAPTER VI
SUMMARY FINDINGS, CONCLUSIONS AND
RECOMMENDATIONS
Brief introduction to the chapter what details are contained in the chapter.
6.1
Summary of the Findings
This section shows the summary of findings in brief statements based on the problems
and hypotheses. There must be a specific answer to each of the questions and
hypotheses posed in the earlier chapters of the study.
6.2
Discussion/Conclusions
The discussion section can either be combined with the conclusions, or appear
separately. In either case, it forms a cohesive narrative with the introduction, and you
should expect to move materials back and forth between these two sections as you
rewrite and reshape your thesis. Topics that are central to your thesis will appear in
the introduction and probably again in the discussion.
Begin the discussion by telling the reader what you have learned from the study. Open
with a clear statement on the support or nonsupport of the hypotheses or the answers
to the questions you first raised in the introduction. But do not simply reformulate and
repeat points already summarized in the results section. Each new statement should
contribute something new to the reader's understanding of the problem. What
inferences can be drawn from the findings? What are the theoretical, practical, or even
the political implications of the results?
It is also appropriate at this point to compare your results with those reported by other
investigators and to discuss possible shortcomings of your study, conditions that
might limit the extent of legitimate generalization or otherwise qualify your
inferences. Remind readers of the characteristics of your participant sample, the
possibility that it might differ from other populations to which you might want to
generalize; of specific characteristics of your methods that might have influenced the
outcome; or of any other factors that might have operated to produce atypical results.
thesis.
25
6.3
Recommendations
This section offers vivid and specific solutions based on the results of the study. The
recommendations should be action words or interventions which would partially or
wholly eliminate the problem areas found in the results. Major findings should be
stated using bullets followed by a brief explanation of the interventions
6.4
Further Research
Research often indicates the need for more research. The results may point to a new
topic or hypothesis. The conclusions may provide some insight that dictate further
study. The current project may suggest other possible topics or avenues of
investigation. Others may want to continue working on some phase of the project.
Include a few statements indicating what further research could be done on the topic.
(This may be a variation on the original theory, variables used, sample size, or
statistical model that has been refined and modified to obtain a more accurate
reflection of reality).
26
9
REFERENCES AND BACK MATTERS
7.1 Appendixes
Appendixes are final elements in formal reports. They contain supplemental
information, information that is too detailed and technical to fit well into the body of
the report, or information that some readers need and others do not. It includes
documents, interviews, statistical results, case histories, list of pertinent items,
specifications, or lists of legal references.
Here are some guidelines: Label Appendixes with letters, such as "Appendix A" and
"Appendix B" if you have more than one; Provide a title for each appendix, such as
"Appendix A. Questionnaire Sample."; Indicate in the body of the report that an
appendix provides supplemental information on a particular topic, such as, "See
Appendix C for cost figures."
7.2 The Use of References
References are records of the sources of information in the report and follow the final
section of the report body. It is popularly known as Documenting sources. This refers
to the practice of citing original sources of information used in formal reports, journal
articles, books, or any document that includes evidence from published works.
Cite information sources for the following reasons: readers can locate the original
sources and read them if they want; you are not personally responsible for every fact
in the document; you will avoid charges of plagiarism. Plagiarism is the
unacknowledged use of information discovered and reported by others or the use of
their exact words, copied verbatim.
If your research relies on other sources, remember to document information such as
using a direct quotation or paraphrasing information from another source.
The documentation system frequently used are: The American Psychological
Association (APA) System and the Number -Reference System (NR). Although all
document systems are designed to help readers find original sources, the systems do
vary slightly. The APA System includes author and the date of publication while in
the NR System - references are written as shown above, but the reference list is
numbered: list the items alphabetically by last name of author and list the items in the
order they cited in the text.
7.3 Bibliography and References
(1) For those in the social sciences and versions similar to it, such as, the biological
sciences, business, and the earth sciences, the style followed is The Publication
Manual of the American Psychological Association (APA).
27
(2)
Since the Master of Business Administration is a social science, the APA Style
is recommended.
APA System: Citation in the Text. An example:

Television's portrayal of women has been called "the best of recent years"
(Steenland, 1986, p.17). Since reader-response theory (Allen, 1987) suggests that
viewers interpret images individually, research is now focusing on testing viewer
perceptions. Atwood, Zahn, and Webber (1986) used random telephone dialing
to test viewer response to women in television and found no difference between
male and female viewers. A British study (Gunther & Wober, 1982) found that
heavy viewers of crime actions shows had lowered perceptions of women as
traditionally feminine. Durkin (1983) provides a comprehensive survey of all
relevant studies of audience perception involving children. A recent study
(Howard et al., 1992) reported significant differences in the ratings of female
television characters based on the gender of the viewer. Howard et al., also
provided comparisons within and among groups of male and female viewers.
The Number-Reference (NR) System.: Citation in the Text. An example:
Television's portrayal of women has been called "the best of recent years" (1:17).
Since reader-response theory (2) suggests that viewers interpret images
individually, research is now focusing on testing viewer perceptions. Reference
(3) used random telephone dialing to test viewer response to women on television
and found no differences between male and female viewers. A British study (4)
found that heavy viewers of crime action shows had lowered perceptions in
women as traditionally feminine. A comprehensive survey (5) of all relevant
studies of audience perceptions involving children was completed in 1985. A
recent study (6) reported significant differences in the ratings of female television
characters based on the gender of the viewer. The same study also provided
comparisons within and among groups of male and female viewers.
For readers, the APA system is easier to use because it provides dates for information
and the names of authors that readers may recognize as experts on the subject. The
NR system requires readers to flip back and forth to the reference list to find dates and
authors.
7.3.1 Writing in the Proper Tense for an APA Styled Paper
1. Verb tense is an indicator that distinguishes papers in the humanities from those in
tile natural sciences. APA style requires the use of the past tense or present perfect
tense ("Marshal stipulated" or the "work of Elmford and Mills has demonstrated").
APA style requires the present tense when you discuss the results (e.g., "the
results confirm: or the study indicates"') and when you mention established
knowledge (eg., the therapy offers some hope or salt contributes to
hypertension).
28
2. APA style requires that you use present tense for generalizations and references to
stable conditions, but it requires the past tense for sources cited: the sources have
tested (present perfect) a hypothesis or the sources reported (past tense) the
results of the test.
7.3.2 Using In-Text Citations in APA Style

1. APA style uses these conventions for in-text citations.
1. Cites last names only.
2. Cites the year, within parentheses, immediately after the name of the
author.
3. Cite page numbers only with a direct quotation, not with a paraphrase.
4. Uses "p". or "pp." before page numbers.
5. Uses double spacing side headings and permits triple-spacing before and
after illustrations and tables
2. APA Style requires an in-text citation to the last name of the author and the year
of publication
The study of Conniff (1990) showed that one federal agency, the Bureau of
Land Management, failed to protect the natural treasures of public land holdings,
and Struble (1991) offered evidence that BLM serves the needs of ranchers, not
the public.
3. If you do not use the author's name in your text, place the name within the
parenthetical citation:
It has been shown that the Bureau of Land Management often sacrifices
wildlife and the environment to benefit miners and ranchers (Conniff, 1990;
Struble, 1991).
4. Provide a page number only when you quote the exact words of a source, and do
not use p. or pp. with page numbers:
Conniff (1990, p. 33) explained that the bureau must figure out how to keep
the land healthy while also accommodating cowpokes, strip miners, dirt bikes,
birdwatchers and tree huggers, all local, all willing to sue for their conflicting
rights.
5. Put the year immediately after the name of the authority; the page number may
appear at the end of the quotation:
Jones (1984) found that these data of psychological development suggest
that related adolescents are atypical in maturational growth (p. 215).
6. Write a quotation of 40 words or more as a separate block, indented an additional
five spaces. Because it is set off from the text in a distinct block, do not enclose it
with quotation marks:
Albert (1983) reported the following:
29
Whenever these pathogenic organisms attack the human body and

begin to multiply, the infection is set in motion. The host responds to this
parasitic invasion with efforts to cleanse itself of the invading agents. When
rejection efforts of the host become visible (fever, sneezing, congestion), the
disease status exists. (pp. 314-315)
7. When one work has more authors, use & in citations only, not in the text:
It has been reported (Werner & Throckmorton, 1990) that toxic levels
exceeded the maximum allowed levels each year since 1983.
Werner and Throckmorten (1190) offered statistics on their analysis of water
samples from six rivers and announced without reservation that the writers are
unfit for human consumption, pose dangers to swimmers, and produce
contaminated fish that may cause salmonella (pp. 457-458).
8. For three to six authors, name them all in the first entry, (Torgerson, Andrews,
Smith, Lawrence, & Dunlap, 1989), but thereafter use (Torgerson et al., 1989).
For seven or more authors, imply (Fredericks et al., 1989) in the first and all
subsequent instances.
9. Use small letters (a, b, c) to identify two or more works published in the same year
by the same author; for example Thompson (1986a) and Thompson (1986b).
Then use 1986a and 1986b in your list of references.
Horton (1986; Thomas, 1982, p. 89) suggested an intercorrelation of these
testing devices. But after multiple-group analysis, Welston (1989, esp. p: 211)
reached an opposite conclusion.
10. If you make an in-text citation to an article or chapter of a textbook, casebook, or
anthology, use the in-text citation to refer only to the person(s) you cite:
"One writer stressed that two out of every three new jobs in the 1990s will
go to women (Bailey, 1988).
11. The list of references will clarify .the nature of this reference to Bailey (see
Textbook, Casebook, Anthology," 286). Corporate authors may be abbreviated
after a first, full reference:
"One source has questioned the results of the use of aspirin for arthritis
treatment in children (American Medical Association (AMA), 1991).
12. Thereafter, refer to the corporate author by initials: (AMA, 1991).
13. When a work has no author listed, cite the title as part of the in-text citation (or
use the first few words of the material):
"The cost per individual student has continued to rise rapidly ("Money
Concerns," 1991, p. 2).
30
7.4
Preparing the List of References
Use the title "References" for your bibliography page. Alphabetize the entries and
double-space throughout. Every reference used in your text should appear in your
alphabetical list of references at the end of the paper. Type the first line of each entry
flush left, and indent succeeding lines three (3) spaces. Use two spaces after each
period in a reference (except between initials of names).
Books
Book (Basic Format)
Carter, J. (1988). An outdoor journal: Adventures and Reflections. New York:
Bantam.
1. List the authors (surname first with initials for given names), year of publication
within parentheses, title of the book underlined and with only the first word of the
title and any subtitle capitalized (but do capitalize proper nouns), place of
publication, and publisher. In the publisher's name, omit the word Publishing
Company, (If Inc. but otherwise a full name: Harcourt Brace Jovanovich; Florida
State University Press; Harper Collins).
2. List chronologically two or more works by the same author; for example,
Fitzgerald's 1989 publication would precede his 1991 publication:
Fitzgerald, R.F. (1989). Controlling oil spills.
Fitzgerald, R. F. (1991). Alaska and its oil reserves.
3. References with the same author in the same year are alphabetized and marked
with lowercase letters (a, b, c) immediately after the date:
Fitzgerald, R. F. (1990a). Water purification systems.
Fitzgerald, R. F. (1990b). Water waste today.
4. Entries of a single author precede multiple-author entries beginning with the same
surname without regard for the dates:
Fitzgerald, R. F. (1990). Water purification systems.
Fitzgerald, R. F., & Smithson, C.A. (1988). Water maps.
5. References with the same first author and different second or third authors should
be alphabetized by the surname of the second author:
Fitzgerald, R. F., & Smithson, C.A. (1988). Water maps.
Fitzgerald, R. F., W.R. (1989). Alaska pipelines.
31
Part of a Book
Hartley, J. T., Harker, J. O., & Walsh, D.A. (1980) Contemporary issues and
new directions in adult development of learning and memory. In L. W.
room (Ed.), Aging in the 1980s: Psychological issues (pp. 239-252).
Washington, DC: American Psychological Association.
6. List author(s), date, chapter or section title, editor (with name in normal order)
preceded by "In" and followed by "(Ed.)" or "(Eds.)," the name of the book
(underlined, page numbers to the specific section of the book cited (placed within
parentheses), place of publication, and publisher).
Textbook, Casebook, Anthology
Make a primary reference to the anthology;
Vesternman, W.(1991). Readings for the 21st century. Boston: Allynn &
Bacon. 5th edition.
7. Thereafter, make cross-references to the primary source, in this case to reference
pages in alphabetical order so that cross-references may appear before or after the
primary source. The year cited should be the date when the cited work was
published, not when the Vesternman book was published; such as information is
usually found in a headnote, footnote, or list of credits at the front or back of the
anthology.
Bailey J. (1988). Jobs for women in the nineties. In Vesternman, pp. 55-63.
Fallows, D. (1982). Why mothers should stay home. In Vesternman, pp.6970
Steinem, g. (1972). Sisterhood. In Vesternman, pp. 48-53.
Vesternman, W. (Ed.) (1991). Readings for the 21stcentury. Boston: Allynn
& Bacon
8. The alternative to the style shown above is to provide a complete entry for
everyone of the authors cited from the casebook (in which case you do not need a
separate entry to Vesternman):
Bailey, J. (1988). Jobs for women in the nineties. In W.
Vesternman (Ed.) (1991), Readings for the 21st century
(pp.55-63). Boston: Allynn & Bacon.
Steinem. G. (1972). Sisterhood. In W. Vesternman (Ed.),
(1991) Readings for the 21st century (pp. 48-53).
Book with Corporate Author, Third Edition
American Psychiatric Association. (1980). Diagnostic
statistical manual of mental disorders (3rd ed.).
Washington, D.C.: Author.
32
Encyclopedia
Woodley, D. J. (1990). Acne. World book encyclopedia.
Chicago: World Book.
9. List author, year, title of the article, title of the encyclopedia (underlined), place,
and publisher. If no author is listed, begin with the title of the article.
Brazil. (1970). Harper encyclopedia of the modern world.
New York: Harper.
Periodicals/Journals
Mielke, K. W. (1988). Television in the social studies classroom. Social
Education, 52, 362-365.
10. List author, year, title of the article without quotation marks and with only the first
word capitalized, name of the journal underlined and with all major words
capitalized, volume number underlined, inclusive page numbers not preceded by
"p." or "pp."
Magazines
Conniff: R. (1990, September). Once the secret domain of
miners and ranchers, the BLM is going public.
Smithsonian, pp. 30-47.
11. List author, data of publication (year, month with abbreviation, and the specific
day for weekly and fortnightly magazines), title of the article without quotation
marks and with only the first word capitalized, name of the magazine underlined
with all major words capitalized, and inclusive page numbers preceded by "p." or
"pp."
Newspaper
Raymond, C. (1990, September 12). Global migration will
have widespread impact on society, scholars say. The
Chronicle of Higher Education, pp. A1, A6.
12. List author, date (year, month and day), title of article with only the first word and
proper nouns capitalized, complete name of newspaper in capitals and underlined,
and the section with all discontinuous page numbers.
Abstract (Citing from an Abstract Only). An abstract of a published
article is cited as follows:
Misumi, J., & Fujita, M (1992). Effects of PM
organizational development in supermarket organization.
Japanese Journal of Experimental Social Psychology, 21,
93-111. (From Psychological Abstracts, 1982, 62,
33
Abstract No. 11474)

13. An abstract of an unpublished work is cited as follows:
Havens, N. B. (1982). Verbalized symbolic play of
pre-school children in two types of play environments.
Abstract of doctoral dissertation, Temple University.
(From Dissertation Abstracts International, 1982, 42,
5058A)
Review
Jones, S. L. (1991, January 6). The paper of motivation
(Review of Body Heat). Contemporary Film Review, p. 18
Report
Lance, J. C. (1990). Housing regulations (KU No. 90-16).
Lawrence, KS: Media Center.
Nonprint Material
Corborn, W. H. (1990, November 3). "On facing the fears
caused by nightmares" (Interview). Lexinton, KY.
Purple, W. C. (Producer). (1990). Hitting the backhand
(Videotape). Nashville: Sports Network.
Landers, J., Woolfe, R. T., & Balcher, C. (1990).
Geometry games: Level two (computer program. Emporia.
KS: media works.
34
Thesis Writing Made Simple
10
CHOOSING THE RIGHT STATISTICS
Statistical Treatment of Data. The following section illustrates some simulations
and exercises on how to choose the correct statistical test. These simulations and
exercises act as a guide to assist you in reviewing and selecting what is the
appropriate statistical treatment for your study.
1. Simulations by Sprinthall, R. Basic Statistical Analysis, 5th Edition,
1997 pp. 476-497.
SIMULATION A
It has been traditional for the man rather than the woman to receive the check when a
couple dines out. A researcher wondered whether this would still be true if the woman
was clearly in charge, asking for the wine list, and so on. A large random sample of
restaurants was selected. One couple was used in all restaurants, but half the men
assumed the traditional in-charge role, and in the other half, the women was in charge.
The data were in the form of the numbers of times the check was presented to each
member of the couple.
ANALYZING THE METHODOLOGY
The is experimental research; the independent variable (man or woman in charge) was
clearly manipulated. The dependent variable was whether the man or the woman
received the check. Thus, if differences in the dependent variable occur here, they can
be attributed to the action of the independent variable. Of the experimental designs,
this was an after-only, no pretest was given, and no matching took place.
ANSWERING THE CHECKLIST QUESTIONS: THE CRITICAL DECISIONS
1. Scale of measurement? The measurements in this study are in the form of nominal
data the responses being categorized on the basis of which member of the couple
received the check.
2. Hypothesis? The hypothesis being tested is that of difference-whether or not a
difference as to who received the check would occur as a function of who
appeared to be in charge. (Since this was the experimental method, it has to test
the hypothesis of difference.)
3. If the hypothesis of difference is tested, are the groups independent or correlated
samples? The subjects (waiters or waitresses) were independently selected, since
the restaurants were randomly chosen.
4. How many sets of measurements? There are two sets of nominal measurements
based on whether the man or the woman gets the check.
35
SOLUTION
Use chi-square, in this instance a 2 x 2 chi-square, with the independent variable
(who is in charge) in the rows and the dependent variable (who gets the check) in the
columns.
SIMULATION B
A researcher, noting the positive correlation between socioeconomic status and
amount of education, assumes there will be a difference in the amount of TV viewing
by the principal wage earners in high and low-socio economically placed families. A
random sample of families is selected and categorized as having either high or low
socioeconomic status on the basis of a number of measures (income of principal wage
earners of all families were then contacted and asked for their hours per week of TV
viewing). The hours per week for each socioeconomic class were compared.
This is an example of post-facto research; the independent variable (socioeconomic
status) is a subject variable, not manipulated. Regardless of whether the statistical test
is significant, no definitive cause and effect statement can be made. Even if the upperclass group is shown to watch significantly less, the reason for this difference can be
suspected (Perhaps upper-class persons read more, or work more hours, or have
expensive time-consuming hobbies, or ... the list of possibilities is endless).
ANSWERING THE CHECKLIST QUESTIONS: THE CRITICAL DECISION
1. Scale of Measurement? The measurement here are in at least interval form. If one
person watches TV for 25 hours and another for 24, we know not only that one
person spends more time watching TV, but also how much more time. With no
evidence to the contrary, the distribution of number of hours spent watching TV is
assumed to be normal.
2. Hypothesis? The hypothesis tested is that of difference-that the two groups would
differ as to the extent that they watched TV.
samples? The two groups in this case are independent of each other, each being
randomly selected and then assigned to the high or low socioeconomic status.
4. How many set of measurements? There are two sets of measurements, one for
upper status and one for lower status.
SOLUTION
Use the independent t test. Had it been predicted that one group would watch more
TV than the other, then test as a one-tail t.
36
SIMULATION C
A researcher wished to test the hypothesis that taller men are more likely than shorter
men to be judged as leaders. A random sample of 30-year old men was selected and
measured for height. The men were then brought before a panel of personnel
managers and ranked ordered on the basis of perceived leadership qualities. Each
subject was assigned the median rank of the panel's decisions.
This is a post facto research. The independent variable (height) was not manipulated
(although it could have been by using elevator shoes, or even hidden stilts). Therefore
the issue of causation is not relevant here, although the probability of making accurate
leadership predictions is very much at issue.
1. Scale of measurement? Despite using the panel's consensus on rankings, the data
are still in ordinal form. Although we may know that Subject 1 is perceived to be
imbued with more of the leadership image than is Subject 2, we do not know how
much more. Also height is clearly an interval measurement, it must be converted
to an ordinal rank before the statistical analysis can be completed (the men are
simply rank ordered as to height).
2. Hypothesis? The hypothesis being tested here. is that of association. It is
gratuitous to try and establish a difference between such already disparate
measures as height and perceived leadership qualities.
samples? Not applicable; the hypothesis of association is being tested.
4. How many sets of measurements? There are two sets of measurements, height and
leadership.
SOLUTION
Use the Spearmans rank correlation coefficent. Note that in this problem, both age and
sex were ruled out as variables, since only men of the same age were selected.
SIMULATION D
A market researcher, working for a manufacturer of hair coloring, wished to establish
whether blondes do indeed have more fun. A large random sample of female college
sophomores was selected and categorized as to hair color blonde, brunette, or red.
Each subject was then asked to answer yes or no to the question, "On balance, would
you say you've been having fun, this semester?"
37

This is post facto research. The independent variable (hair color) was a subject
variable, not manipulated. (Here again, this could have been designed as an
experiment, perhaps randomly dividing the brunettes into other groups and then
giving all members of one group a blonde tint and allowing the effect statement to be
made). The groups may differ on a variety of other variables related to happiness.
number of dates, grade point average, etc. Also, it may be a function of the woman's
personality. Perhaps socially-oriented women are more apt to use a tint, and be more
optimistic.
1. Scale of Measurement? Since the women are measured on the basis of answering
the question either yes or no, and then the frequencies of these answers are tallied,
the data are nominal.
2. Hypothesis? This is one of the studies in which the hypothesis could conceivably
be classified either way-difference or association. It is likely that the researcher
meant this as a difference study since the hypothesis was stated as, " blondes have
more fun." (Had it been the hypothesis of association, it probably should have
been stated as, "There is a correlation between hair color and having fun.")
samples? The samples (blondes, brunettes and redheads) are independent of one
another. The fact that one woman was placed in the blonde group neither caused
another woman to be placed in, nor precluded her placement in, a different group.
4. How many sets of measurement? There are two sets of scores based on whether
the subjects report having or not having fun.
SOLUTION
Use chi-square, in this case a 3 x 2 chi- square. Since this is a 3 x 2 chi-square, the
special 2 x 2 computational method cannot be used. .This means that all of the values
for fe (frequencies expected) must be separately calculated. Had this been a test of
association, the chi-square could have been followed up with the coefficient of
contingency.
SIMULATION E
A researcher wished to test the hypothesis that older men sleep less than younger men
do. Random samples of 30-year-old men, 50-year-old men, and 70-year-old men were
selected. Each subject was brought to a sleep laboratory and measured as to how
many hours of sleep per night occurred.
This is a post facto study; the independent variable (age) was a subject variable, not
manipulated (Age can, of course, never be a manipulated independent variable). Even
38
if the results prove to be significant, great care must be taken in interpreting them. If it
is found that the older men sleep less; it may be they did so as young men too. As
youngsters, these men may have been more apt to rise early as a result of the differing
cultural patterns that typified their younger days. (This is actually cross-sectional
research).
1. Scale of Measurement? The dependent variable (hours of sleep) provides at least
interval data, with an underlying distribution that is probably normal.
2. Hypothesis? The hypothesis being tested is one of difference that different age
groups have different sleep habits.
samples? These sample groups are independent of one another. The fact that a
given man is selected for the 30-year-old group has no bearing on who is being
selected in the 50- or 70-year-old groups.
4. How many sets of measurements? There are three sets, one for each age group.
SOLUTION
Use the one-way ANOVA, the F ratio. If F proves to be significant, proceed with
Tukey's HSD.
SIMULATION F
A researcher wanted to test the hypothesis that racial prejudice is a function of
personal authoritarianism. A random sample of college students was selected and
measured on the F Scale, an index of personal rigidity and authoritarianism. All
subjects were then given the A-S (for Anti-Semitic) scale, a measure of prejudice
toward Jews.
ANALYZING THE METHODOLOGY,
This is a post. facto study; the independent variable (authoritarianism) is a subject
variable, not manipulated. Even if the hypothesis is validated, there will be no way to
tell whether authoritarianism affects prejudice or prejudice affects authoritarianism. It
is even possible that a third variable, such as a family's child-rearing practices,
produces both authoritarianism and prejudice.
1. Scale of Measurement? Both the F and A-S scales are considered to be interval
measures distributed normally in the population.
2. Hypothesis? The hypothesis in this case is one of association. (One group is being
measured on two different response dimensions.) We can never test for
differences between completely unrelated measures.
39

4. How many sets of measurements? There are two sets, one for F scale and one for
A-S scores. .
SOLUTION
Use the Pearson r. If found to be significant, the r could be followed by a regression
equation, with which specific A-S scores could be predicted from given F scores.
SIMULATION G
A researcher working for a large corporation wished to test the hypothesis that the
company's toothpaste, containing fluoride, reduces dental caries. A random sample of
18-ycar-olds was selected, and all subjects were checked for caries. A dentist then
filled the cavities for all subjects having them. For the next three years, all subjects
received free monthly supplies of the fluoride toothpaste, Finally, at age 21, the
subjects were again checked for dental caries. The researcher then compared the
number of persons with caries found in the first dental checkup with the number of
persons with caries found at age 21.
This is experimental research, before-after design. The independent variable
(toothpaste) is manipulated (rather than being assigned on the basis of whether the
subjects, on their own, were using it). This is however, a shaky design, because
maturation is a variable and may confound the independent variable. Perhaps 18-yearolds, as a group, are more apt to have caries (the cavity-prone years) than are 21-yearolds. Perhaps, there might have been a significant result even without the introduction
of the independent variable. It would have been better to have had a separate control
group, checked for caries at both ages, but given, instead of the fluoride brand,
toothpaste that looked the same but did not contain fluoride.
1. Scale of Measurement? These are nominal data. The subjects were categorized as
a group, at each age as to whether caries were detected. We do not know whether
any subject had more or fewer caries than another, only whether any caries were
present. Frequencies of occurrence for each age were compared. Had the subjects
been rank ordered in terms of amount and/or severity of caries, the data would
have been ordinal. Perhaps even interval measures could have been designed,
where the subjects received scaled scores based on the number and severity of
caries found.
2. Hypothesis? This is the hypothesis of difference the assumption being that after
using the fluoride toothpaste the frequency of caries within the group will
diminish.
40

samples? The groups are as correlated as possible, since the same group was used
as its own control. The samples are always correlated in a before-after design.
4. How many sets of measurements? There are two sets, one taken at age 18 and
other at age 21.
SOLUTION
Use the McNemar test, which is a chi-square based on the change scores.
SIMULATION H
A researcher wished to find out whether the perception of a person's height depends
on that person's perceived status. A random sample of army inductees was selected
and equally divided into four groups. An actor gave a short address to each group
separately, extolling the joys of army life. For the first group, the actor was dressed as
a private; for the second, as a sergeant; for the third, as a captain; and finally, for the
fourth group, as a colonel. The inductees were asked to fill out a questionnaire
evaluating the speech. Among the questions was one asking for an estimate of the
lecturer's height.
This is experimental research, after-only design.. The independent variable (perceived
status) was manipulated by having the same actor wear different uniforms. Using the
same actor was a good idea, since otherwise, differences in personal characteristics
might have confounded the independent variable.
1. Scale of Measurement? The dependent variable (estimated height) provided least
interval data in this study.
2. Hypothesis? The researcher is looking for differences in the height estimates by
the four groups.
samples? The four groups are independent of one another. The selection of one
soldier had no bearing on whether another was or was not selected. This is an
after-only experimental design, in which sample groups must be independent of
one another.
4. How many sets of measurements? There are four sets, one for each treatment
condition.
SOLUTION
Use the One-way ANOVA, the F-ratio, If F is significant, follow it up with Tukey's
HSD.
41
SIMULATION I
A researcher suspects that both meaningfulness and length of presentation affect word
retention, as scored on a standardized test. Four groups of fifth-grade students were
randomly selected and then assigned to different treatment conditions. Group A was
given low-meaningful words and a 2-second presentation time. Group B was given a
list of high-meaningful words and also a 2-second presentation time. Group C was
given low-meaningful words and a five-second presentation time, whereas Group D
received a high-meaningful list and a 5-second presentation time. The results of their
retention scores were analyzed.
1. Scale of measurement? Standardized retention scores come in at least interval
form.
2. Hypothesis? The researcher is attempting to discover if there is a difference in the
retention scores.
samples? Since each of the four groups was randomly selected and assigned
separately to the various conditions, the samples are independent.
4. How many sets of measurements? There are four sets of measures, one set for
each group.
SOLUTION
Use the factorial ANOVA and be especially careful to look for the possibility of a
significant interaction.
SIMULATION J
A researcher for an electronics corporation wished to establish whether, other things
being equal, the tonal quality of a hi-fi set is judged to be better as the size of the
speaker enclosure is increased. A random sample of subjects was selected and asked
to listen to the same CD played on "different sound systems." Actually, the amplifier,
the size and quality of the speaker and baffle, and so on, remained the same. Only the
size and quality of the speaker enclosure was allowed to vary. Three enclosure sizes
were used-small, medium and large. The subjects were asked to rank order their
preferences, from 1 (best) to 3 (worst). The order in which the subjects were presented
with the various speaker sizes was counterbalanced, so that some subjects had the
large enclosure first, others the small enclosure, and so on.
This is experimental research, repeated-measure design (before-after-after). The
independent variable (enclosure size) was manipulated by the experimenter. If
significant result are obtained, cause and effect inferences can be made.
42

1. Scale of measurement? The dependent variable (judgment of tonal quality) is
ordinal form, that is, the subjects rank ordering of the three listening conditions.
2. Hypothesis? The study tests the hypothesis of difference that different
judgments of sound quality will occur as enclosure size is changed.
samples? The groups in this study are definitely correlated, as the same group is
used in all three treatment conditions. Groups are always correlated in repeatedmeasure designs.
4. How many sets of measurements? There are three sets, one for each treatment
condition.
SOLUTION
Use the Friedman ANOVA by ranks. Compare the ranks 1, 2 and 3 of each subject
under the three listening conditions.
SIMULATION K
A researcher is interested in whether coaching can have any effect on Math SAT
scores. A group of 100 high school seniors was randomly selected from a large
metropolitan school district. The group was then randomly divided into two subgroups. One group was given three months of daily coaching in those math skills
deemed important to the SAT, while the other group spent the same amount of time
each day watching reruns of the TV show Happy Days. At the end of the threemonth period, all students took the SAT and their math scores were compared.
1. Scale of measurement? Standardized retention scores come in at 1east interval
form.
2. Hypothesis? The researcher is testing the hypothesis of difference.
samples? The two groups are independent, since the selection of one student being
random had no influence on the selection of other students.
4. How many sets of measurements? There are two sets of measures, the SAT scores
from each of the two groups.
SOLUTION
Use the independent t test.
43
SIMULATION L
.
A study was designed to test whether presenting one side or both sides of an argument
is more effective in changing attitudes. Perhaps presenting just the pro side would be
more effective because an audience might not be fully aware of the anti side. Or
perhaps to appear impartial and to avoid having members of the audience go over to
the anti side and therefore tune out the pro message, it would be more effective to at
least present some of the anti arguments.
A large random sample was selected, and the subjects were assigned to one of two
conditions. Group A heard only the pro side of the issue, whereas Group B heard the
entire pro side plus a few minutes of anti arguments. Both presentations was made by
the same person. A questionnaire tapping attitudes toward the issue, was then filled
out by each subject.
This is experimental research. The independent variable (one-sided versus two-sided
presentations) was manipulated by the researcher. As no matching occurred and no
attitude testing was done prior to the presentation, this was an after-only design. If the
results prove significant, causal inferences can be drawn.
1. Scale of Measurement? The questionnaire was scored as interval data and the
assumption of a normal distribution was made.
2. Hypothesis? As in all experimental research, the hypothesis of difference was
tested. Presumably, differences in attitudes between the two groups can be
attributed to the independent variable.
samples? As is true of all after-only experimental designs, the groups were
independently selected.
4. How many sets of measurements? There are two sets of scores to be compared.
SOLUTION
Use the independent t. As no prediction regarding direction was even suggested,
check the t as a two-tail test.
SIMULATION M
A researcher wished to test the hypothesis that male business majors earn more in
later life than do either male liberal arts or education majors. A random sample of
alumni was selected from the university files from each of the three subject major
categories. To attempt to control for length of experience on the job, all subjects were
selected from the same graduating class-the class that graduated 10 years ago. All the
44
selected alumni were contacted and asked to indicate their yearly incomes. The men
were promised that the information would be held in strict confidence and would not
be given to the chairman of the upcoming alumni fund drive. Because a few of the
subjects reported enormously high incomes, the resulting distribution became so
skewed that it was decided to rank order the incomes.
This is post facto research. The independent variable, college major, was a subject
variable, not manipulated.
1. Scale of Measurement? Although income is an interval measurement, skewed
distribution forced a rank ordering of the data, thus creating a series of ordinal
measures.
2. Hypothesis? The researcher was testing for differences among the income ranks of
the three groups
samples? The groups are independent. The assignment of alumni into subject
major categories is strictly independent. The selection of one person from the
"education" category did not demand or preclude another subject being selected
from the "liberal arts" category.
4. How many sets of measurements? There are three sets. one for each of the subject
major categories.
SOLUTION
Use the Kruskal-Wallis H test for three or more independent groups and ordinal data.
SIMULATION N
A researcher wished to increase the predictability of student pilot scores on the FAAs
written general aviation exam. Dependable relationships were found to exist between
number of ground school and FAA exam scores and also between IQ and exam
scores. Finally, a small, but significant relationship was found to exist between IQ and
number of hours of ground school. (Note that for the private pilot's license. the
number of hours of ground school is not fixed by the FAA. A few student pilots put in
many hours and a few study on their own and never attend at all.)
This is post facto research. Although the pilots did experience different conditions
(attending ground school or not) this was their choice, not the choice of the
experimenter. Also, IQ can never be a manipulated variable. Thus. the two
independent variables (ground school and IQ) were subject variables, not
manipulated.
45
1. Scale of Measurement? The three measures (hours of ground school, lQ, and
scores on the FAA exam) all yielded at least interval scores. The distributions all
appear to be normal.
2. Hypothesis? The researcher has attempted to test for associations among the three
measures.
3. If the hypothesis of difference is tested, are the groups dependent or correlated
4. How many sets of measurements? There are three sets-hours of ground school,
FAA examination scores, and IQ.
SOLUTION
Use the Multiple R. The three separate values of the Pearson r (between ground
school and the exam scores, between IQ and the exam scores, and between IQ and
ground school) should all be used together to determine whether their combinations
increase the predictive efficiency. If the value of the Multiple R is larger than the
separate correlation with the exam scores, solve the multiple regression equation.
SIMULATION O
A firearms manufacturer hired a researcher to establish whether a new handgun
increases accuracy. A group of law enforcement agents was randomly selected and
brought to the firing range. First, all subjects used the same traditional service
revolver, and their error scores (in inches from the bull's-eye) were determined. Then
they all fired again, using the new weapon, and their error scores were again
determined.
This is experimental research, the independent variable (type of weapon) was
manipulated by the experimenter. Since the same subjects are used in both treatment
conditions, the design is before-after. This is not the best design for the study, because
it is possible that scores might improve the second time as a result of practice. This
could act to confound the independent variable. It would have been better to set up a
separate group that used the old weapon twice, another group that used the new
weapon first and then the old weapon, and another group that used the new weapon
twice.
1. Scale of measurement? The dependent variable (error measured on the basis of
inches from the bull's-eye) provides at least interval data. The researcher claimed
a normal distribution for these error scores. Had the distribution not been normal,
a different statistical test should have been used,
2. Hypothesis? As in all experimental research, the hypothesis of difference
(between error scores) was tested.
46

samples? The groups are correlated; the same group is used as its own control.
4. How many sets of measurement? There are two sets, one for each treatment
condition.
SOLUTION
Use the paired t-ratio, probably as a one-tail test since the manufacturer undoubtedly
has some reason for believing in the superiority of the new weapon. (If the results had
gone the other way, they possibly would have been filed away in the back of the
drawer.) Had the distribution of error scores been skewed (many officers hitting the
bull's-eye, but a few missing the target altogether), then the scores should have been
rank ordered and the Wilcoxon T test performed.
SIMULATION P
An investigator wished to establish whether a dependable re1ationship exists between
height at age 3 and height at age 21. A random sample of 3-year-olds was selected,
and height measures were taken on each. The researcher then patiently waited 18
years and measured the subjects again.
This is post facto research, the independent variable (height at age 3) is a subject
variable, not manipulated. This is also called longitudinal research, since the same
subjects are followed through the years and are used again. (A less patient researcher
could have obtained adult heights and then checked personal records for the infant
heights.)
1. Scale of measurement? The data are in at least interval form, and the distributions
for each age level are probably normal.
2. Hypothesis? This is strictly the hypothesis of association. (Testing the hypothesis
of difference in this situation that is, that 21-year-olds are significantly taller
than 3-year-olds would hardly add much to the book of knowledge.)
samples? not applicable; the hypothesis of association is being tested.
4. How many sets of measurements? There are two sets of measurements, one taken
at age 3 and the other at age 21.
SOLUTION
Use the Pearson r. If it is significant, set up the regression equation of Y on X. Thus
predictions of adult height can be made from height at age 3.
47
SIMULATION Q
Some researchers have suspected that, because of academic and other frustrations,
adolescents labeled as LD (learning disabled) would have more symptoms of
depression and even possibly higher levels of suicidal ideation than would non-LD
adolescents. Two groups of 16-year-old students, one labeled LD and the other nonLD (50 male adolescents in each group), were selected on the basis of a certain school
district's records. All students were then given the SIQ-JR (Reynolds Suicide
Ideational Questionnaire), and the results were as follows:
LD
Range (0 to 90)
M
=
15.21
SD
=
17.32
Mdn
=
7.92
Sk
=
+2.40
Non-LD
Range (0 to 90)
M
=
SD
=
Mdn
=
Sk
=
12.33
16.28
6.34
+2.45
This is post facto research, as the IV in this study, LD versus non-LD, is clearly a
subject variable. If the results prove to be significant, predictions, but not direct causeand-effect, become more viable.
1. Scale of measurement? The measurements are in interval form, but since the mean
is so much higher than the median and the standard deviation is large, relative to
the mean, the distributions are significantly skewed to the right. Notice that with
means of only 15 and] 2 no negative scores ( the ranges were 0 to 90) the
distribution must be skewed to the high side. If these distribution were to approach
normality, the range of scores would have been from about -37 to +61 for the nonLD and -37 to +67 for the LD. With severe skews of this sort, reported as +2.40
and +2.45, respectively, the interval data should be converted to ordinal.
2. Hypothesis? The researcher is testing for differences in suicidal ideation between
the two groups.
samples? The groups are independent.
4. How many sets of measurements? The researcher was comparing two sets of
measurements.
SOLUTION
Use. the Mann Whitney U test for detecting differences between two sets of ordinal
scores.
48
SIMULATION R
A cultural anthropologist became interested in discovering whether differences in the
age of menarche (the age when young woman have their first menstrual cycle) are a
function of climate. Two groups of young women were selected one from a northern
climate (Norway) and one from a southern climate (Italy). The subjects were matched
according to both height and weight, and their ages at menarche were compared. The
age distribution was found to be skewed.
This is a post facto research; the independent variable (climate) was a subject
variable, not manipulated. Thus, even if significance is established? no positive causal
statement can be made. The subjects obviously differ on a host of variables (diet,
genetic background, medical care, etc.) other than climate.
1. Scale of Measurements? Although age is at least an interval measure, the lack of
normality in the underlying distribution forces a conversion of the age into ordinal
data.
2. Hypothesis? The researcher is testing the hypothesis of difference that age at
menarche differs as a function of climate.
samples? The groups are correlated, having been matched on both height and
weight.
4. How many sets of measurements? There are two sets of measurements, one taken
in Norway and the other in Italy.
SOLUTION
Use the Wilcoxon T test.
SIMULATION S
A researcher wanted to find out whether IQ is a function of family size. The
speculation was that among families with fewer children, each children receives more
parental attention and intellectual stimulation and should therefore have a higher IQ
than would a child reared in a larger family. A large random sample of two-child
families was selected as well as a similar sample of six-child families. The IQ of all
children were measured, and the two sample groups were compared.
This is post facto research. The independent variable (family size) was a subject
variable, not manipulated. (natural forces or their own decision, not the decision of the
experimenter, determined which families had small or large numbers of children.)
Thus, even if significance is established, the causal factor remains in the realm of
speculation. Could it be, instead, that lower-IQ parents have more children.
49
1. Scale of Measurements? IQ scores are considered to be interval measures, and the

underlying distribution to be fairly normal.
2. Hypothesis? The researcher is looking for IQ differences among children from
small and large families.
3. If the hypothesis of difference is tested? Are the groups independent or correlated
samples? The groups are independent. The selection of a given family depended
on its size, not on whether or not some other family had been selected.
4. How many sets of measurements? There are two sets of IQ scores, one taken from
large families and one taken from small families.
SOLUTION
Use the independent t. If the score for each child is to be used separately, use the
equation for unequal values of N (there are three times as many IQ scores in the sixchild families). If the children's IQ scores are to be averaged within each family, then
equal values of N can be maintained.
SIMULATION T
A study on rats was conducted to find out whether environmental inputs could affect
brain growth. A group of 16 laboratory rat twins was selected at birth and randomly
assigned to the two groups. The rats were thus paired off on the basis of genetic
inputs, one rat from each twin-pair in the experimental group and the other twin in the
control group. The experimental group was then raised in a stimulating environment,
in a cage equipped with ladders, running wheels, and other "rat toys". These animals
were also let out of their cages for 30 minutes each day and allowed to explore new
territory. They were also trained on a number of learning tasks and in general received
a rich and varied array of stimulus homogeneity. These rats lived alone in dimly lit
cages, were rarely handled, and were never allowed to explore areas outside their
cages. All animals received exactly the same diet. After 90 days, all the animals were
sacrificed and their brains analyzed morphologically and chemically. The weight of
each rat cortex was then recorded in milligrams, and the two groups were compared
on the basis of cortical weight.
This is experimental research, with the independent variable, environmental inputs,
manipulated by the researcher. If the results prove to be significant, cause and effect
statements become possible.
1. Scale of measurement? The data are in at least interval form.
2. Hypothesis? The researcher is testing the hypothesis of difference.
samples? The groups were deliberately correlated by the researcher, since all the
rats were paired off on the basis of twin-ship.
50
4. How many sets of measurements? There were two sets of measurements, one set
from each group.
SOLUTION
Use the paired t.
2. Research Situations and Solutions by: Levin, J. Elementary Statistics for
Social Sciences, 2nd Edition, 1990
Each statistical procedure has a set of assumptions for its appropriate application. In
selecting among procedures, any researcher must therefore consider a number of
factors such as:
1. Whether the researcher seeks to test for statistically significant differences, degree
of association, or both;
2. Whether the researcher had achieved the nominal, ordinal or interval level of
measurement of the variables being studied;
3. Whether the variables being studied are normally distributed in the population
from which they were drawn; and
4. Whether the researcher is investigating independent samples or the same sample
measured more than once.
Table 1. locates each statistical procedure with respect to some important
assumptions that must be considered for its appropriate application. Looking at the
columns of the table, we face the first major decision related to the selection of a
statistical procedure: Do we wish to determine whether or not a relationship exists?
The test of significance is designed to determine whether an obtained sample
difference reflects a true population difference. Or, do we seek instead to establish the
strength of the relationship between two variables. This is a question of correlation,
which can be addressed by means of the statistical procedures. A researcher who
decides to employ a test of significance rather than a correlation procedure must also
be aware of whether he is studying independent samples or the same sample measured
more than once.
The rows in Table 1 directs our attention to the level at which variables are measured.
If we have achieved the interval of measurement, we may well consider employing a
parametric procedure such as t, F, or r. If however, we have achieved either the
nominal or ordinal level of measurement, the choice is limited to several non
parametric alternatives.
RESEARCH SITUATION NO. 1
A researcher conducted an experiment to determine the effect of a lecturer's age on
student preferences to hear him lecture. In a regular classroom situation, 20 students
were told that the administration wished to know their preferences regarding a
forthcoming visiting lecturer series. In particular, they were asked to evaluate a
professor who might be visiting the campus. The professor was described to all
students in the same way with one exception: One-half of the students were told the
professor was 65 years old; one half were told the professor was 25 years old. All
51
students were then asked to indicate their willingness to attend the professor's lecture
(higher scores indicate greater willingness). The following results were obtained:
X1
(Scores of students
told professor was 25 years old)
65
38
52
71
69
72
55
78
56
80
X2
(Scores of students
told professor was 65 years old)
78
42
77
50
65
70
55
51
33
59
Which statistical procedure would you apply to determine whether there is a

significant difference between these groups of students with respect to their
willingness to attend the lecture?
RESEARCH SOLUTION NO.1
Research situation 1 represents a comparison between the scores of two independent
samples of students. The t ratio is employed in order to make comparisons between
two means when interval data have been obtained. The median test is a non
parametric alternative, which can be applied when we suspect that the scores are not
normally distributed in the population or that the interval level of measurement has
not been achieved.
RESEARCH SITUATION NO.2.
student preferences to hear him lecture. In a regular classroom situation, 30 students
were told that the administration wished to know their preferences regarding a
forthcoming visiting lecturer series. In particular, they were asked to evaluate a
professor who might be visiting the campus. The professor was described to all
students in the same way with one exception: one third of the students were told that
the professor was 75 years old; one third were told the professor was 50 years old; and
one third were to1d the professor was 25 years old. All students were then asked to
indicate their willingness to attend the professor's lecture (higher scores indicate
greater willingness). The following results were obtained.
52
X1
X2
X3
(Scores of students told (Scores of students told (Scores of students told
professor 25 yrs)
professor 25 yrs)
professor 25 yrs)
65
63
67
38
42
42
52
60
77
71
55
32
69
43
52
72
36
34
55
69
45
78
57
38
56
67
39
80
79
46
significant difference between these groups of students with respect to their
willingness to attend the lecture?
RESEARCH SOLUTION 2
Research situation 2 represents a comparison of the scores of three independent
samples of students. The F ratio (analysis of variance is employed in order to make
comparisons between three or more independent means when interval data have been
obtained.
Kruskal-Wallis one-way analysis of variance is a non parametric alternative, which
can be applied when we have reason to suspect that the scores are not normally
distributed in the population or when the interval level of measurement has not been
achieved.
To investigate the relationship between spelling and reading ability, a researcher gave
spelling and reading examinations to a group of 20 students who had been selected at
random from a large population of undergraduates, the following results were
obtained (higher scores indicate greater ability):
Student
A
B
C
D
E
F
G
H
X
(Spelling Scores)
52
90
63
81
93
51
48
99
53
Y
(Reading Scores)
56
81
75
72
50
45
39
87
I
J
K
L
M
N
O
P
Q
R
S
T
85
57
60
77
96
62
28
43
88
72
75
69
59
56
69
78
69
57
35
47
73
76
63
79
Which statistical procedure would you apply to determine the degree of

association between spelling and reading ability?
RESEARCH SOLUTION NO. 3
Research situation is a correlation problem, since it asks for the degree of association
between X (spelling ability) and Y (reading ability). Pearson r can be employed to
detect a straight-line correlation between X and Y variables when both of these
variables have been measured at the interval level. If X (spelling ability) and Y
(reading ability) are not normally distributed in the population, we should consider
applying a non parametric alternative such as Spearman's rank-order correlation
coefficient.
To investigate the validity of a particular reading test, researchers gave the reading
test to a sample of 20 students whose ability to read had been previously ranked by
their teacher. The test score and teacher's rank for each student are listed below:
Student
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
X
(Reading Scores)
28
50
92
85
76
69
42
53
80
91
73
74
14
29
86
54
Y
(Teachers Rank)
18
17
1
6
5
10
11
12
3
2
4
9
20
19
7
P
Q
R
S
T
73
39
80
91
72
8
16
13
15
14

association between reading scores and teacher's ranking?
RESEARCH SOLUTION NO.4
Research situation 4 is a correlation problem, asking for the degree of association
between X (reading scores) and Y(teacher's evaluation of reading ability).
Spearman's rank-order correlation coefficient can be employed to detect a
straight-line relationship between X and Y variables, when both of these variables
have been ordered or ranked.
Pearson r cannot be employed, since it requires interval level measurement of X and
Y. In the present case, reading scores (X) must be ranked from 1 to 20 before rankorder is applied.
To investigate regional differences in helpfulness towards strangers, a researcher
dropped 400 keys (all of which had been stamped and tagged with a return address)
around mailboxes in the northeastern, southern, mid western and western regions of
the United States. The number of keys returned by region (as an indicator of
helpfulness) is indicated below:
Returned
Not returned
Total
Region
Northeast
f
55
45
100
South
f
69
31
100
Midwest
f
82
18
100
West
f
61
39
100
Which statistical procedure would you apply to determine whether these

regional differences are statistically significant?
Research Situation 5 represents a comparison between the frequencies (returned
versus not returned) found in four groups (northeast, south, mid west and west) The
chi-square test of significance is used to make comparisons between two or more
samples. Only nominal data are required. Present results can be cast in the form of a
2x4 table, representing 2 rows and 4 columns. Notice that the degree of association
between returns rate (x) and region (y) can be measured by means of the Contingency
coefficient (C) or Cramer's V.
55

To examine the relationship between authoritarianism and prejudice, a researcher
administered measures of authoritarianism (the F Scale) and prejudice (a checklist of
negative adjectives to be assigned to black Americans) to a national sample of 950
adult Americans. The following results were obtained: Among 500 authoritarian
respondents, 350 were prejudiced and 150 were tolerant. Among 450 non
authoritarian respondents, 125 were prejudiced and 325 were tolerant.
Which statistical procedure would you apply to study the degrees of association
between authoritarianism and prejudice?
Research situation 6 is a correlation problem, which asks for the degree of association
between X (authoritarianism) and y (prejudice). The phi coefficient is a measure of
association, which can be employed when frequency or nominal data can be cast in
the form of a 2x2 table (2 rows by 2 columns)
Example:
Level of Prejudice
Prejudiced
Tolerant
Level of Authoritarian
Authoritarian
Non-authoritarian
350
120
150
325
N = 950

To investigate the relationship between year in school and grade point average,
researchers examined the academic records of 186 college students who were selected
on a random basis from the undergraduate population of a certain university. The
researchers obtained the following results:
Year in School
Grade
Point
1st
Average
f
A- or better
6
B- to B+
10
C- to C+
23
C or worse
15
Total
54
2nd
f
5
16
20
7
48
3rd
f
7
19
15
6
47
4th
f
10
18
7
2
37
Which statistical procedure would you apply to determine the degrees of

association between grade point average and year in school?
between X (grade-point average) and Y (year in school). Goodman's and Kruskal's
56
gamma coefficient is employed to detect a straight line relationship between X and

Y, when both variables have been ranked and a large number of ties have occurred. In
the present problem, a grade-point average has been ranked from A to D or worse and
year in school has been ranked from 1st to 4th. Both of these crude ordinal measures
have generated numerous tied ranks (for example, 54 students were in their first year
of school; 48 students were in their second year, and so on.) The contingency
coefficient (C) or Cramer's V represents an alternative to gamma, which assumes
only nominal-level data.

To investigate the influence of frustration on prejudice, 10 subjects were asked to
assign negative adjectives such as lazy, dirty and immoral to describe the members of
a minority group (a measure of prejudice). All subjects described the minority group
both before and after they had taken a series of lengthy and difficult examinations (the
frustrating situation). The following results were obtained (higher scores represent
greater prejudice).
Subject
A
B
C
D
E
F
G
H
I
J
X1
Prejudice Scores Before
Taking the Frustrating
Exam
22
39
25
40
36
27
44
31
52
48
X2
Prejudice Scores After
Taking the Frustrating
Exam
26
45
24
43
36
29
47
30
52
59

statistically significant difference in prejudice before and after the
administration of the frustrating examinations?
Research situation 8 represents a before-after comparison of a single sample measured
at two different points in time. The t ratio can be employed to compare two means
from a single sample arranged in a before-after panel design. Friedman's two-way
analysis of variance is a non parametric alternative, which can be applied to the
before-after situation when we have reason to suspect that the scores are not normally
distributed in the population or when we have not achieved the interval level of
measurement.
57

To investigate the relationship between a respondents actual occupational status and
his subjective social class (that is a respondent's own social class identification). A
total of 677 individuals were asked to indicate their occupation and the social class in
which they belonged. Among 190 respondents with upper status occupations
(professional-technical-managerial), 56 identified themselves as upper class, 122 as
middle class and 12 as lower class; among 221 respondents with middle-status
occupations (sales-clerical skilled labor), 42 identified themselves as upper class, 163
as middle class and 16 as lower class, among 266 with lower-status occupations
(semi-and unskilled labor), 15 identified themselves as upper class, 202 as middle
class and 49 as lower class.
association between occupational status and subjective social class?
between X (occupational status) and Y (subjective social class). Gamma is especially
well suited to the problem of detecting a straight-line relationship between X and Y,
when both variables can be ranked and1llarge number of ties have occurred. In the
present situation, occupational status and subjective social class have been ordered
from "upper to middle to lower" generating a very large number of tied ranks (for
example, 221 respondents had middle -status occupations). In order to obtain the
gamma coefficient, the data must be rearranged in the form of a frequency table as
follows:
Subjective
Class (Y)
Upper
Middle
Lower
Total
Social Occupational Status (X)

Upper
Middle
f
f
56
42
122
163
12
16
190
221
Lower
f
15
202
49
266
The contingency coefficient (C) and Cramer's V are alternatives to gamma which
assume only nominal data.
To investigate the influence of college major on the starting salary of college
graduates, researchers interviewed recent college graduates on their first jobs who had
majored in engineering, liberal arts, or business administration. The results obtained
for these 21 respondents are as follows:
Example:
Starting Salaries
(in Dollars)
58
Engineering
Liberal Arts
Business Administration
10,500
12,300
14,000
9,500
9,000
8,500
7,500
.
7,000
9,500
10,000
11,000
8,500
7,500
7,000
7,500
9,000
8,000
9,300
10,500
10,000
7,000

significant difference between these groups of respondents with respect to their
starting salaries?
Research situation 10 represents a comparison of the scores of three independent
samples of respondents. The F ratio is used to make comparisons between three or
more independent means when interval data have been obtained. KruskalWal1is
one-way analysis of variance is a non-parametric alternative, which can be employed
when we suspect that the scores may not be normally distributed in the population or
when the interval level of measurement has not been achieved.
RESEARCH SITUATION NO. l1
To investigate the influence of college major on the starting salary of college
graduates, researchers interviewed recent college graduates on their first jobs who had
majored in either liberal arts or business. The results obtained for these 16
respondents are the following:
Starting Salaries
(in Dollars)
Liberal Arts
7,000
9,500
10,000
11,000
8,500
7,500
7,000
Business
7,500
9,000
8,000
9,300
10,500
10,000
7,000
8,000
9,300

significant difference between liberal arts majors and business majors with
respect to starting salaries?
59

Research situation 11 represents a comparison between scores of two independent
samples of respondents. The t ratio is employed in order to compare two means when
interval data have been obtained. The median test is a nonparametric alternative,
which can be applied when we cannot assume that the scores are normally distributed
in the population or when the interval level of measurement has not been achieved.
student willingness to hear him lecture. In a regular classroom situation, 130 students
were told that the administration wished to know their preferences regarding a forthcoming visiting lecturer series. In particular, they were asked to evaluate a professor
who might be visiting the campus. The professor was described to all students in the
same way with one exception: One half of the students were told the professor was 65
years old; one half were told the professor was 25 years old. All students were then
asked to indicate their willingness to attend the professor's lecture with the following
results: Among those students told that the professor was 65, 22 expressed their
willingness to attend his lecture and 43 expressed their unwillingness; among the
students told that the professor was 25, 38 expressed their willingness to attend his
lecture and 27 expressed their unwillingness. Which statistical procedure would you
apply to determine whether there is a significant difference between these groups of
students with respect to their willingness to attend the professor's lecture
Research situation 12 represents a comparison of the frequencies (willingness versus
unwillingness) in two groups of students (those told by the professor was 25). The
chi-square test of significance is used to make comparisons between two or more
samples, when nominal or frequency data have been obtained. Present results can be
cast in the form of the following 2x2 table, representing 2 rows and 2 columns:
Willingness
attend
Willing
Unwilling
to
Experimental Condition
Students told professor was 65
f
22
43
Students told professor was 25

f
38
27
N = 130
STATISTICAL REQUIREMENTS
1. Chi-Square
1. A comparison between 2 or more independent samples. We have at least 2 x 2
table. At least two samples of respondent must be obtained.
2. Nominal data- only frequencies are required.
3. Random sampling- done on samples at random from a particular population.
60
4. The expected cell frequencies should not be too small-exactly how large fe
must depend on the nature of the problem (from 2 x 2 problem- no expected
frequency should be smaller than 10.
2. Phi Coefficient
1. Nominal data - only frequency data are required.
2. A 2x2 table - the data must be capable of being cast in the form of a 2x2 table
(2 rows by 2 columns). It is inappropriate to use tables larger than 2x2 in
which several groups or categories are being compared.
3. Random Sampling - in order to test the significance of the phi-coefficient,
sample numbers must have been drawn on a random basis from a larger
population.
3. Contingency Coefficients
1. Nominal data- only frequency data are required. These data may be cast in the
form of 2 x 2 table or larger.
2. Random Sampling- for purpose of testing the significance, all sample numbers
must have been taken from a larger population.
4. Median Test
1. A comparison between two or more independent medians- the median test is
employed to make comparisons between 2 or more medians for independent
samples.
2. Ordinal data - perform a median test, we assume at least the ordinal level of
measurements. Nominal data cannot be used.
3. Random sampling - we should have drawn on samples on a random basis from
a given population.
5. KruskalWallis One-way ANOVA
1. A comparison of 3 or more independent samples one-way ANOVA cannot be
applied to test differences within a single sample of respondents measured
more than once.
2. Ordinal data - only data capable of being ranked are required.
3. Each sample must contain at least 6 cases where there are more than 5
respondents in each group, the significance of H can be determine by means of
the appropriate chi-square value.
6. Friedman's Two-way ANOVA
1. A comparison of a single sample measured under 2 or more conditions. It
cannot be applied to test differences between independent samples but
assumes that the same sample of respondents has been measured at least twice
(or that the numbers of 2 or more samples have been matched on relevant
variables).
2. Ordinal data - only data capable of being ranked are required.
3. The number of respondents must not be too small. The exact minimum
requirement for N depends on the number of conditions (K) to which
respondents are to be exposed. Example: N must equal or exceed 10 when K =
3 whereas N must equal or exceed 5 when K = 4
7. Spearman's Rank-Order Correlation Coefficient
61
1. A straight-line correlation - the rank order coefficient denotes a straight line

relationship between x and y.
2. Ordinal data - both x and y variables must be ranked or ordered.
3. Random sampling - sample numbers must have been taken at random from a
larger population.
8. Goodman's and Kruskal's Gamma
1. A straight-line correlation - gamma denotes straight-line relationship between
x and y.
2. Ordinal data - both x and y variables must be ranked or ordered.
3. Random sampling - to test the null hypothesis (G = 0), sample numbers must
have been taken on a random basis for some specified population.
9. Z Score - t Ratio
1. A comparison between two means - the z scores or t ratio are employed in
order to make comparison between two means from independent samples or
from a single sample arranged in a before-after panel design.
2. Interval data - the assumption is that we have scores at the interval level of
measurement. Therefore, we cannot use the z score or t-ratio for ranked data or
data that can only be categorized at the nominal level of measurement.
3. Random sampling- we should have drawn our samples at a random basis from
a population of scores.
4. A nominal data - the t ratio for small sample requires that the sample
characteristics we have measured be normally distributed in the underlying
population.
10. F-Ratio
1. A comparison between 3 or more independent means - the F-ratio is usually
employed to make comparisons between 3 or more means for independent
samples.
2. Interval data. to conduct an analysis of variance, we assume that we have
achieved the interval level of measurement.
3. Random sampling - we should have taken our samples at random from a given
population of scores.
4. A normal distribution - we assume the single characteristics we measure to be
normally distributed in the underlying population.
11. Pearson R
1. A straight-line relationship - it is only useful for detecting a straight-line
correlation between x and y.
2. Interval data - both x and y variables must be measured at the interval level so
that scores may be assigned to the respondents.
3. Random sampling - sample numbers must have been drawn at random from a
specified population otherwise a test of significance cannot be applied.
4. Normally distributed characteristics - testing the significance requires both x
and y variables to be normally distributed in the population. In small samples,
failure to meet the requirements of normally distributed characteristics may
someday impair the validity of Pearson r. However, the requirement is of
minimum. importance where the sample size equals or exceed 30 cases.
62
LEVEL OF MEASUREMENTS
1. The Nominal Level
It simply involves the process of naming or labeling, that is of placing cases of
categories and counting their frequency of occurrence.
Example:
Attitude Towards Puerto Ricans
Frequency
1 = prejudiced
2 = unprejudiced
Total
5
5
10
Other nominal-level measures in social research are sex (male vs. female), welfare
status (recipient vs. non-recipient), political party (Republican, Democratic,
Independent and Socialist); social character (inner-directed, other-directed and
tradition-directed); mode of adaptation (conformity, innovation, ritualism, retreatism
and rebellion); time orientation (present, past and future); and urbanization (urban,
rural and suburban), to mention a few.
In dealing with nominal data, we must keep in mind that every case must be placed in
one, and only one category. The reader should note that nominal data are not graded,
ranked or scaled for qualities such as better or worse, higher or lower, more or less.
Nominal data are merely labeled, sometimes by name (male vs. female) but always
for the purpose of grouping the cases into separate categories to indicate sameness or
differences with respect to a given quality or characteristics.
2. The Ordinal Level
When the researcher goes beyond this level of measurement and seeks to order his
cases in terms of the degree they have any given characteristic, he is working at the
original/level of measurement.
Example:
To classify individuals with respect to socio-economic status as lower class, middle
class or upper class. He might rank them according to their degree of prejudice against
Puerto Ricans.
Student
Joyce
Mary
Bill
Jill
Kathy
Jack
Phillip
Steve
Patricia
Pete
Rank
1 - most prejudiced
2 - second
3 - third
4 - fourth
5 - fifth
6 - sixth
7 - seventh
8 - eighth
9 - ninth
10 - tenth
63
The ordinal level of measurement yields information about the ordering of categories,
but does not indicate the "magnitude of the differences between numbers. It is not
possible to determine how much more prejudiced Joyce is than Mary or how much
less prejudiced Roberta is than Patricia or Steve. This is because the intervals between
the points or ranks on an ordinal scale are not known or meaningful. Therefore it is
not possible to assign scores to cases located at points along the scale.
3. The Interval Level
By contrast, it tells us about the ordering of categories and also indicates the exact
distance them. Interval measures use constant units of measurement. Example:
(dollars or cents, Fahrenheit or centigrade, yards or feet, minutes or seconds), yields
equal intervals between points on the scale.
Example:
Student
Joyce
Mary
Bill
Jill
Kathy
Jack
Phillip
Steve
Patricia
Roberta
Score*
98
96
95
94
22
21
20
15
11
6
* higher scores indicate greater prejudice against Puerto Ricans. It indicates distances
separating one from another. Depending on the purpose for which the study is
designed. such information might be important to determine, but is not available at the
ordinal level of measurement. .
Characteristics of Choosing the Correct Statistical Test
Characteristics
1. Nominal data, hypothesis of difference and
independent selection (two or more measures)
2. Nominal data, hypothesis of difference and
correlated selection (two measures)
3. Nominal data, hypothesis of association and any
number or cell
4. Ordinal data, hypothesis of difference, independent
selection and two measures
5. Ordinal data, hypothesis of difference, independent
selection and more than two measures
6. Ordinal data, hypothesis of difference, correlated
selection, and two measures
7. Ordinal data, hypothesis of difference, correlated
selection and more than two measures
8. Ordinal data, hypothesis of association
64
Statistics
Chi-Square
McNemar Test
Coefficient of
Contingency
Mann-Whitney U Test
Kruskal-Wallis H Test
Wilcoxon T Test
Friedman ANOVA
ranks
Spearman r s
by
9.
10.
11.
12.
13.
14.
Interval data, hypothesis of difference, independent

selection, and two measures
Interval data, hypothesis of difference, independent
selection, two or more measures more than one
independent variable
Interval data, hypothesis of difference, correlated
selection, two or more measures
Interval data, hypothesis of difference, correlated
selection, two or more measures
Interval data, hypothesis of association, two
measures with prediction
Interval data, hypothesis of association, more than
two measures with prediction holding one or more
factors out
Independent t Test
One-Way ANOVA
Ratio
Factorial ANOVA
Paired t Ratio
Within Subjects F Ratio

Pearson r
Regression of Y on X
Multiple R
Multiple
Regression
Partial Correlation
Definitions of Terms in Statistics

1. Chi-Square - a statistical test of significance used to determine whether or not
frequency differences have occurred on the basis of chance. It requires that the
data be in nominal form, or the actual number of cases that fall into two or more
discrete categories. It is a non parametric test.
2. McNemar Test - a technique developed by the statistician Quinn McNemar, that
uses chi-square for the analysis of nominal data from correlated samples.
3. Coefficient of Contingency - a test of correlation on nominal data sorted into any
number of independent cells.
4. Mann Whitney U Test - A test on ordinal data of the hypothesis of difference
between two independently selected random samples. The U test is the ordinal
counterpart of the independent t-test.
5. Kruskal-Wallis H Test - a test of the hypothesis of difference on ordinal data
among at least three independently selected random samples. The H test is the
ordinal counterpart of the one-way ANOV A.
6. Wilcoxon T - Test-a test on ordinal data of the hypothesis of difference between
two sample groups when the selection are correlated (as in the matched-subjects
design). The Wilcoxon T is the ordinal counter part of the paired t.
7. Friedman ANOVA by Ranks - a test of the hypothesis of difference on ordinal
data when either the sample group have been matched or a single sample has been
repeatedly measured. The Friedman ANOVA is the ordinal counterpart of the
within-subjects F.
8. Spearman rs - correlation coefficient-devised by Charles E. Spearman for use with
rank-ordered (ordinal) data. sometimes called the Spearman p (rho).
9. Independent t- Test-it assumes that the two samples are of equal size.
65
10. One Way ANOVA or F-Ratio - statistical test of significance developed by Sir
Ronald Fisher. It is also called the F-ratio, or ANOV A for analysis of variance.
The test is designed to established whether a significant (nonchance) difference
exists among several sample means. Statistically, it is the ratio of the variance
occurring between the sample means to the variance occurring within sample
groups. A large F ratio that is when the variance between is larger than the
variance within-usually indicates a nonchance or significant difference.
11. Factorial ANOVA-as opposed to a one-way ANOV A, it allows for the analysis of
data when more than one independent variable is involved. Results can be
analyzed In the basis of the main effects of each independent variable or on the
basis of the possible interaction among the independent variables. Data to be
analyzed should be in at least interval form.
l2. Paired t-Ratio-statistical test of the hypothesis of difference between two sample
means where the sample selection is not independent. The paired t (also called
correlated t) requires interval data and is typically used when the design has been
before-after or matched-subjects.
13. Within-subjects F-Ratio - statistical test of the hypothesis of difference among
several sample means, where sample selection is not independent. It is used when
samples are correlated, as in repeated-measure designs and the data are in at least
interval form.
14. Pearson r - statistical technique introduced by Karl Pearson for showing the
degree of linear relationship between two variables. Also called the product!11oment correlation coefficient, it is used to test the hypothesis of association,
that is whether there is a relationship between two sets of measurements.
15. Regression of Y on X-the point where the regression line crosses the Y axis. The
regression crosses the ordinate, that is we must establish the value of Y when X is
equal to zero. this is called the point of intercept or the Y intercept.
16. Multiple R-a single numerical value that quantifies the correlation among three or
more variables. the equation for a three variable multiple R.
17. Multiple Regression - technique using the Multiple R for making predictions of
one variable given measures on two or more others. It requires the calculation of
the intercept (a) and also at least two slopes (b1 and b2). For the three-variable
situation.
18. Partial Correlation-correlation technique that allows for the ruling out of the
possible effects of one or more variables on the relationship among the remaining
variables. In the three-variable situation, the partial correlation rules out the
influence of the third variable on the correlation between the remaining two
variables. The equation for partialing out the influence of the third variables.
19. Nominal Data - data (measurements) in which numbers are used to label discrete,
mutually exclusive categories, nose-counting data, which focuses on the frequency
of occurrence within independent categories.
66
20. Ordinal Data - ranked ordered data, that is derived only from the order of the
numbers, not the differences between them. It provides information regarding less
than or greater than status, but how much greater or less.
21. Interval Data - data (measurements) in which values are assigned such that both
the order of the numbers and the intervals between numbers are known. It
provides information how much greater than or less than.
RESEARCH SIMULATIONS: CHOOSING THE CORRECT STATISTICAL

TEST
NOMINAL DATA
Hypothesis of
Difference
Hypothesis of
Association
Independent
Selection
Correlated
Selection
(Two or More
Measures)
(Two Measure)
Hypothesis of
Difference
McNemar Test
Any Number
of Cells
67
Coefficient of
Contingency
ORDINAL DATA
Hypothesis of
Association
Hypothesis of Difference
Independent
Selection
Two
Measures
Two
Measures
Correlated
Selection
Any Number
of Cells
Two
Measures
Coefficient of
Contingency
McNemar Test
Hypothesis of
Difference
68
INTERVAL DATA
Hypothesis of
Difference
Hypothesis of
Association
Independent
Selection
Two
Measures
Two or More
Measures
IndepenDent
t-Ratio
One-Way
ANOVA
or F-Ratio
Corelated
Selection
Two
Measures
Two or More
Measures
Paired
t-Ratio
Within
Subjects
F-Ratio
More
than one
Independent
Variable
Factorial
ANOVA
Two
Measures
More than Two

Measures
Pearson r
Multiple R
With
Prediction
With
Prediction
Regression
of Y on X
Multiple
Regression
Holding One
Or More
Factors Out
Partial
Correlation
69
KINDS OF CONCEPTUAL FRAMEWORKS
1. One Dimensional Frameworks: One sample of respondents
Input
Process
Output
2. Two Dimensional Framework: Two samples of respondents

Sample No.1
Sample No.2
3. Logical Framework Analysis American Style: 16 cells relationships

Narrative
Summary
Objectively
Verifiable
Indicators
Program
Goals
Project
Purpose
Outputs
Inputs
70
Means of
Verification
Important
Assumptions
4. Logical Framework Analysis - German Style: similar to Decision Trees
Hierarchy of Ideas
First Level Main Problem/ Main Variable
Second Level Sub-Problem/ Sub-Variable
Third Level Sub- sub Problem/ Sub- sub- Variable
5. Path Analysis: time series
Time 1
Time 2
71
Time 3
6. Three some Variable Relationships Frameworks

1. Spurious Association of X1 Y: association between X1 and Y Disappears
X1
X2
2. Chain Relationship, X2 intervenes, X1 indirectly causes Y: Association

between X1 and Y disappears
X1
X2
3. Interaction: Association between X1 and Y varies according to level of X2

X2
X1
72
4. Multiple Causes: Association between X1 and Y does not change.

X2
X1
5. Both Direct and Indirect Effects of X1 on Y: Association between X1 and

Y changes, but does not disappear.
X1
X2
73
APPENDIXES
74
APPENDIX A
PHYSICAL SPECIFICATIONS
A.1
Paper
Plain bond paper, typical for photocopying and printer needs, is acceptable for the
thesis. It is required that tables and figures be arranged consistently in uniform
font and layout, in the format check copy and in the final copies.
The two final copies of the thesis presented to the GSB Office must be of a
uniform format and typed, error-free, on high-quality paper of durability.
Requirements are that the paper be:
- white
- 8.5 inches x 11 inches
- unlined
- of at least 25 percent rag or cotton content, and
- of at least twenty-pound weight.
The use of erasable or recycled paper is not acceptable.

A.2
Copies
The term original refers to the typed thesis prepared with a word processor or
typewriter and printed on paper in the format required by the GSB. The term
photocopy refers to exact duplicates of the original that is prepared by
photographic or photocopy means.
A.3
Corrections
There are to be no errors in the thesis, that is, errors must be corrected so as to be
undetectable. Erasure and ink-removal may be acceptable means of correction,
however, correction tape, correction fluids, and correction strips are not acceptable
means of correction. When extensive corrections are detectable to the thesis
examination committee members before the defense, the student may be required
to present re-typed pages before the Right-to-Defend is issued. After the defense,
pages containing errors must be corrected and replaced prior to when the final
thesis is to be submitted.
A.4
Fonts
The font used should be a common size such as Pice, Elite, Times or Helvetica
and not of an unusual style such as Cursive, Script, Gothic or Italic. The entire
thesis is to be typed in the same typeface, although exceptions can be made, when
necessary, for tables, charts, figures, appended materials and computer-generated
data. The font recommended by the GSB is Times New Roman 12.
75
A.5
Spacing
Headings or titles on opening pages (first pages of acknowledgement contents,
tables listing, chapters, etc.) must begin approximately 2 inches from the top of
the paper. The design and spacing of prefatory pages should follow the format
given in the Sample pages of this handbook in the Appendix. The 2 inches mark
should typically fall somewhere on the line of type, i.e., from the top of the
uppercase letter to the bottom of the letters with tails.
Text throughout the document is to be 1.5 inches. Judicious use of single, triple,
and quadruple spacing to set off headings, tables and illustrative materials is
sometimes allowed within the Graduate School guidelines
A.6
Margins
Because these are ultimately filed as bound volumes in the Universitys Library,
students must maintain margins, which allow ease of binding and reading of the
bound work. Binding pages at the left requires 1.5 inches of the paper to be
reserved for the binders convenience. Thus, text must be typed on the 6 inches
space between the 1.5 inch left and 1 inch right margins. The top and bottom of
each page will also maintain a 1-inch right margin. In setting up any centered
headings, titles or numbers, centering should take place within the 6.0 inch-long
typing line. Text on full pages should fill the space allowed. Some specific fonts
lose readability with right and left margin justification. The student may want to
prepare a template to indicate the acceptable margins for the finished work.
A.7
Numbering
Every page of the thesis is numbered except for the annotation, abstract, title and
approval pages. Do not add periods, hyphens, or any other decorative punctuation
to page numbers.
Correct
1
2
3
Incorrect
1.
2.3.
For ease in typing the number, consider it part of the actual text, appearing as the
first on the last line of the text allowed on each full page.
Chapters and the title of the chapters must be placed in the middle of the page on
separate lines. The page numbers of the chapters need not be shown but they must
be counted. The main title must be placed at the far-left end of the page with
bigger letters. The sub-title must be 8 alphabets from the left end.
After the copyright page will be the first page with a printed number and the
numbers will be in Roman Numerals." i ii iii iv " Roman numerals will be used
up to the Abstract page. Begin using cardinal numbers on the first page of the text
76
of the thesis (from Chapter I till the last page of the thesis). The numbers must be
placed at the right hand side of the page.
A.8
Photographs and Illustrative Materials

Mounting of photographs, drawings, pockets or other illustrative materials onto
the thesis pages must be accomplished to assure permanent adherence. Adhesive
tape and glue are not acceptable as these materials become unsightly and are
impermanent. Materials should be mounted with heat fixed adhesive paper or with
permanent-bonding spray adhesive.
Computerized laser equipment, which produces acceptable photographic images,
is now available. To ensure that the required level of reproduction and final
product is maintained, contact the GSB Office with sample for approval. Please
note that photographic material in theses must be black and white. Color
illustrations will not be allowed
A.9
Lettering and Drawing

Lettering and drawing for charts, figures, tables, and exhibits must be done neatly
in black, permanent ink. Computer-generated text from graphic packages and
software is preferred for lettering and drawing. Hand-lettering or drawing must be
of professional quality. Every effort must be made to prevent potential smudging
and/or bleeding in final library copies.
A.10
Footnotes and bibliography
There are several styles of writing footnotes and bibliography but the Graduate
School of Business has set only one style (APA Style) to make it easier for the
students.
A.11
Proofreading the Thesis

It is highly recommended that the final work is professionally and clearly
edited, proof-read for errors in grammar, spelling, misused words, sentence
construction, etc. This is so for all candidates but particularly so for anyone
using English as a second language. The GSB has a form that needs to be
attached attesting that the thesis has been proofread (The English Dept. has a
list of recognized thesis proof readers). The thesis may also benefit from being
academically proof-read by someone other than the advisor (or potential
examiners).
77
APPENDIX B:
SAMPLES
B.1
Thesis Title Page (Cover)
The Influence of Information Search, Involvement and Loyalty on

Switching Behavior: A Case Study of DTACs Mobile Phone Users
in Bangkok Area
By
Ms. Usanee Kasemkiatsanti
Submitted in Partial Fulfillment of the Requirements for the Degree

of Master of Business Administration
Examination Committee:
1. Prof. Dr. Navin Mathur
(Advisor) ..
2. Dr. Ismail Ali Siad
(Member) ..
3. Dr. Chittipa Ngamkroeckjoti (Member) ..

4. Dr. Theerachote Pongtaveewould (Member)
Examined on:
Approved for Graduation on:
Graduate School of Business
Bangkok Thailand
March, 2005.
78
B.2:
Sample of an Abstract
Abstract
Owing to the rapidly changing environment and intense competition in
Thailands mobile phone industry, an issue of concern is customer switching behavior
among mobile phone service providers. Identification of movement in and out of
mobile phone users in the marketplace leads to the requirement for the mobile phone
service providers, such as DTAC, to develop appropriate customer retention, loyaltybuilding efforts and acquisition programs.
The aim of this research was to study the influence of sources of information
search, involvement and loyalty on switching behavior of DTAC mobile phone users.
In addition, the objectives also were to determine the most influential factors of
sources of information search, involvement and customer loyalty that have
relationships with switching behavior. The data was collected from 400 DTAC mobile
phone users whose age was over 18 years old residing in the Bangkok area only. The
chi-square test was employed to test the hypotheses.
The results of the research showed that there was a difference in switching
behavior of DTAC mobile phone users when determined by sources of information
search. On the contrary, there were no differences in switching behaviors when
determined by involvement and customer loyalty.
The findings provide important implications for DTAC to understand the
relationship between switching behavior and the predicting factors composing sources
of information search, involvement and customer loyalty, which can serve as a
guideline for the company to appropriately develop its marketing strategies, such as
advertising campaigns, brand cognition and pricing strategies.
79
B.3:
Sample of a Table of Contents
TABLE OF CONTENTS
Page No.
Committees Approval Sheet
Abstract
Acknowledgements
Table of Contents
List of Tables
List of Figures
i
ii
iii
iv
v
vi
Chapter 1 Generalities of the Study

1.1 Background of the Study
1.2 Statement of the Problem
1.3 Research Objectives
1.4 Scope of the Research
1.5 Limitations of the Research
1.6 Significance of the Study
1.7 Definition of Terms
1
2
3
4
5
Chapter II Review of Related Literature and Studies

2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
Definition and Features of the Independent Variable

Theories and Studies Related to the Independent Variable
Critical Analysis/Discussion of the Theories Related to the
Independent Variable
Discussion of each of the selected Sub-Variables
Relationship of the Independent Variable to the Dependent Variable
Definition and Features of the Dependent Variable
Theories and Studies Related to the Dependent Variable
Critical Analysis/Discussion of the Theories
Previous Studies
Chapter III Research Frameworks

3.1 Theoretical Framework
3.2 Conceptual Framework
3.3 Research Hypothesis
3.4 Operationalization of the Independent and Dependent Variables
Chapter IV Research Methodology
4.1 Methods of Research Used
4.2 Respondents and Sampling Procedures
4.3 Research Instruments/Questionnaire
4.4 Collection of Data/Gathering Procedures
4.5 Statistical Treatment of Data
80
Chapter V Presentation and Critical Discussion of Results

5.1 Presentation, Analysis and Interpretation of Data
Chapter VI Summary, Conclusions and Recommendations
6.1 Summary of Findings
6.2 Conclusions
6.3 Recommendations
6.4 Future Study
Bibliography
Appendixes
81
B.4
: Sample of Chapter, Topics and Pagination
1 1.5
Chapter I an Introduction
8
1.5
1
.
.
.
2 spaces
Global
Reality
.
.
.
(8 Strokes) Asian Reality
.
.
.
82

A Handbook MBA Thesis PDF

Uploaded by

Copyright:

Available Formats

A Handbook MBA Thesis PDF

Uploaded by

Document Information

Original Title

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

A Handbook MBA Thesis PDF

Uploaded by

Copyright:

Available Formats

A Handbook on Thesis Writing

Patricia Arttachariya, Ph.D.

A Publication of the Graduate School of Business, Assumption University, Bangkok,

Dr. Patricia Arttachariya,

Basics of Technical Writing and Thesis Writing

A Handbook on Thesis Writing

KEY FEATURES OF TECHNICAL WRITING

KEY FEATURES OF THESIS WRITING

be deemed acceptable: a new perspective, a new application, uncover new

The Research Process.

The Research Wheel

Characteristics of a Good Thesis Research Problem

The problem can be stated clearly and concisely. It is tested by writing it as a

A Handbook on Thesis Writing

The following sections make up the front matter of the thesis:

i. Thesis Title Page

A Handbook on Thesis Writing

Introduction of the Study

Statement of the Problem

Scope of the Research

Limitations of the Research

Significance of the Study

Thesis Writing Made Simple

Definition and Features of the Independent Variable

Suppose your Independent Variable is Leadership. First, provide definitions of

Theories Related to the Independent Variable (s)

Critical Analysis/Discussion of the Theories Related to the Independent

Relationship of the Independent Variable (s) to the Dependent

For instance: A Study on Leadership by (Adams, 1997) a study which relates

Introduction & background introduces the research topic and

Methods how the study was conducted

Results summary of the data that was collected

Discussion Discusses the implications and significance of the

A Handbook on Thesis Writing

This chapter discusses the Research Frameworks. It included the theoretical

Two-dimensional framework: permits analysis of the interrelationships between

A hypothesis is a claim or statement either about a value of a single population

Classifying - crowded or not crowded

The following is an example of a section of the Operationalization Table:

Q1: Read available information

from printed media (such as

brochures, pamphlets, point-ofsale display, others information

A Handbook on Thesis Writing

Brief introduction of the chapter what would be the different sections to be

Methods of Research Used

Respondents and Sampling Procedures

Respondents of the Study

Table __ Theoretical Sample Sizes for Different Sizes of Population and a 95

Required Sample for Tolerable Error

Anderson, G. Fundamentals of Educational Research, 1996 p.202.

This section discusses the structured instrument/questionnaire which contain the

4.4 Collection of Data/Gathering Procedures

Statistical Treatment of Data

A Handbook on Thesis Writing

The First Main Independent Variable

The Second Main Independent Variable

The same process as in 5.1 .

The First Main Dependent Variable

The same process as above.