5 FIN10002 Chapter4 2015 PDF
5 FIN10002 Chapter4 2015 PDF
5 FIN10002 Chapter4 2015 PDF
PROBABILITY
DISTRIBUTION
Z=
4. Standardise values in a normal distribution by applying the formula:
5. Utilise Microsoft Excel to determine any given area under any normal curve.
6. Solve problems that utilise the normal probability distribution.
4.1.
4.1.1.
The general shape of a distribution is often described by a smooth line, which takes no account of the
irregularities in the histogram of the sample data. The mathematical equation for the smooth curve
which describes the distribution of a variable, can often be found. If the area under the smooth curve is
equal to 1, then in this case the curve is called a probability density curve.
4.1.2.
When a histogram of a data distribution is replaced by a density curve it is found that many of these
density curves have the same characteristic bell shape. Since this bell shaped density curve occurs so
often in practice, it has been given the special name of the normal density curve, or normal curve for
short. This distribution was originally proposed by Karl Gauss, an 18th century mathematicianastronomer, who observed that discrepancies among repeated measurements of the same physical
quantity displayed a regular bell-shaped pattern which could be approximated by a special distribution
which we now know as the normal distribution.
When the general shape of a data distribution can be described by a normal density curve, the data is
said to be normally distributed and the variable being measured is a normal random variable. Some
examples of data distributions that are approximately normally distributed include height, weight, IQ
scores, scientific measurements and many test scores. Since many natural phenomena have been found
to follow at least approximately a normal distribution, the normal distribution has an important place in
statistics. It is also an important distribution in the theory of statistics.
Page | 3
=0
=1
4.1.3.
Each normal distribution is unique and is distinguished by its mean and standard deviation .
The normal curve is symmetric and bell shaped, with the centre at the mean of the distribution it
represents and the width determined by the standard deviation. Since the curve is symmetric,
the mean equals the median.
The normal distribution is such that the percentage of values that lie within a given number of
standard deviations from the mean is always the same, regardless of the mean and standard
deviation of the particular normal distribution. In particular, it can be shown mathematically
that:
Approximately 68% of the values in a normal distribution lie within 1 standard
deviation of the mean, ie within
Approximately 95% of the values in a normal distribution lie within 2 standard
deviations of the mean ie 2
Approximately 99.7%(almost all) of the values in a normal distribution lie within 3
standard deviations of the mean 3
Example 4-1
The heights of women are approximately normally distributed with a mean =160cm and a standard
deviation =8cm as shown below.
(a)
Solution
(a)
i.
95%, as 144 is 2 standard deviations below and 176 is 2 standard deviations above the mean
ii. (ii) 68%, as 152 is 1 standard deviation below and 168 is 1 standard deviation above the
mean
iii. 16%, as one tail in (ii) has (100 - 68)/2% of the area
iv. 0.15%, as one tail outside 3 standard deviations has (100 - 99.7)/2% of the area
(b)
4.1.4.
176 cm or taller as this gives the top (100 - 95)/2% of area for 2 standard deviations above
the mean.
Standardisation
Normally distributed data can be converted to a new set of units, showing the number of standard
deviations that each data value lies from the mean. These new values are called standardised values or
Z-values. Meaningful comparisons can then be made, not only between data values within a particular
distribution but also for making comparisons across distributions which have different means and
Page | 5
standard deviations. For any data value X, taken from a normal distribution with mean and standard
deviation , the standardised or Z-value is given by
Z=
The mean of a standard normal distribution is zero and the standard deviation has the value 1. Z
measures the distance, in number of standard deviations, of a value from the mean.
See Figure 4.1.
Data above the mean have positive Z-values.
Data below the mean have negative Z-values.
Example 4-2
The heights of women are approximately normally distributed with a mean =160cm and a standard
deviation = 8 cm. (See Figure 4.3 on page 5). What is the standardised or Z-value for a woman who
is:
(a)
160cm tall ?
(b)
150cm tall ?
(c)
172cm tall ?
Solution:
4.1.5.
(a)
(b)
Z= (150-160)/8 = -1.25
(c)
Z= (172-160)/8 = 1.5
Since each unique normal probability distribution has the same general properties regardless of its mean
and standard deviation, we can standardise all normal curves so that they reduce to the same normal
curve with a mean =0 and standard deviation =1. This unique normal curve is called the standard
normal density curve. Specially prepared standard normal tables are available (see the end of the book)
so that any given area under the standard normal curve can be computed.
The area under the normal curve over a specific interval represents the probability that the normal
random variable lies within that interval.
Example 4-3
What is Pr(Z 2)?
Solution
0.975
Figure 4.4
NOTES:
1.
2.
Always draw a diagram when solving normal distribution problems. This will
assist you in identifying if you have a an answer which is clearly incorrect.
The normal distribution is symmetrical and therefore having calculated Pr (Z2)
0.975, then we also could also say that Pr (Z-2) 0.975 just by considering
the equivalent shaded upper area as shown in Figure 4.5.
0.975
Figure 4.5
4.1.6.
Page | 7
Case 1
Pr(Za)
=NORMSDIST(a)
Area given by
NORMSDIST =
Pr(Za)
Case 2
Pr(Za)
=1- NORMSDIST(a)
Area given
NORMSDIST
by
Area required
for Pr(Za)
Case 3
by
for
Area
given
NORMSDIST
Pr(Zb)
by
for
Area required
Pr(aZb)
for
draw
Step 2:
Using
the
Normsdist
command
in
Excel
Note: because the normal distribution is a continuous probability distribution, Pr(Z=a) = 0 for any
individual value in the distribution. For example, Pr(Z=2) = 0. It follows that in the standard normal
distribution, Pr(Za) always gives the same value as Pr(Z<a). So for instance, Pr(Z<2) would also be
calculated as =normsdist(2), giving the same answer of 0.977 as in the above example.
Page | 9
z
For example, to find Pr(Z2), we are looking for the area under the standard normal curve to the left of
z = 2.00. Therefore, we would go to the 2.0 row in the z column, and we would go to the 0.00 column,
which indicates an area of 0.9772 to the left of z = 2.00.
z
0.00
2.0
0.9772
10
Example 4-4
What is Pr(Z -1.37)?
Solution
Step 2:
Using
the
Normsdist
command
in
Excel
0.07
-1.3
0.0853
Page | 11
Example 4-5
Find Pr(z 2).
Solution
Step 1:
Draw diagram
for
Pr(Z 2).
Step 2:
Using
the
Normsdist
command
in
Excel
12
Step 2: Using
the Normsdist
command
in
Excel
Page | 13
Example 4-7
What is Pr(-1.37 Z 2)?
Solution
Step 1:
Draw diagram
for
Pr(-1.37 Z 2)
14
Example 4-8
Find the z value such that the probability of obtaining a larger z value is 20% (or 0.2).
Solution:
This problem is different from the previous 6 problems, where the z value was specified and the
corresponding probabilities (or areas) were found. In this example the probability (area) has been
specified and it is necessary to find the corresponding z value. Lets call the z-value we are looking for
a. We should still commence by drawing a diagram of the information we have been given. We know
Pr (Z > a) = 0.2.
Step 1:
Draw diagram
for
Pr (Z > a) = 0.2
Step 2:
Determine area
below a
Step 3:
Normsinv
function
Therefore a = 0.842
This means that Pr(z < 0.842) = 0.8 and hence Pr(z 0.842) = 0.2 as
required.
Page | 15
To calculate the required value of z from the standard normal distribution table, we would look for the
required area of 0.8 in the body of the table.
z
0.84
0.01
0.02
0.03
The value of z falls between z = 0.842 and z = 0.843 in this case, with it being closer to z = 0.842 (since
0.7995 is closer to 0.8000 than 0.8023). Hence z 0.842
DO EXERCISE SET 4A ON PAGE 209
16
4.1.7.
We previously saw that the normsdist command calculated probabilities in the standard normal
distribution (which has = 0, = 1). Using Microsoft Excel we can calculate probabilities associated
with any normal distribution (not only the standard normal distribution) by using the normdist function.
To see how the normdist function works, consider the example of calculating the probability that that a
randomly selected value in a normal distribution with mean = 20 and standard deviation = 5 is less
than 13.
Page | 17
Step 2:
Using
Normdist
command
Excel
the
in
18
Step 3:
probability
From the resulting output, we can see Pr(X13) = 0.081
To save time, note that the general structure of the Excel normdist function to obtain the area to the left
of a value in a normal distribution is:
normdist(value, mean, standard deviation, true)
Therefore, in the above example, we could have simply clicked cell A1 and typed
=normdist(13,20,5,true) and pressed ENTER to obtain the above result of 0.081. In future examples,
we will just type in the normdist command in this manner.
(see p.188).
X 13 20
) = Pr(Z-1.4) = 0.0808
Lets return to Example 4-2. The heights of young women are approximately normally distributed with a
mean = 160 cm and a standard deviation = 8 cm as shown below.
-2
-1
Suppose we are interested in the probability that a young woman is between 152cm and 168cm tall.
Page | 19
Step 1:
Draw diagram
for
Pr(152 < X <168)
Step 2:
Normdist function
152 160
168 160
<Z<
)
8
8
= P(-1< Z <1)
= P(Z<1) P(Z < -1)
= 0.8413 0.1587
= 0.6826.
20
Example 4-9
Suppose that the distance a truck belonging to a large freight company travels per year is normally
distributed with a mean of 95,000 km and a standard deviation of 8,000 km. What percentage of trucks
travel between 80,000 and 100,000 km per year?
Solution
Step 1:
Draw diagram
for
Pr(80,000<X <100,000)
Step 2:
Normdist function
Example 4-10
A consultant is investigating the time it takes factory workers in a motorcycle assembly plant to
assemble a particular part after the workers have been trained to perform the task using an individually
based learning approach. The consultant determines that X, the time in seconds to assemble the part for
workers trained with this method is normally distributed with a mean of 75 seconds and a standard
deviation of 5 seconds.
(a)
(b)
What is the probability that a randomly selected factory worker can assemble the part in
under 75 seconds?
(c)
What is the probability that a randomly selected factory worker can assemble the motor
cycle part in less than 70 seconds?
(d)
(e)
What is the probability that a randomly selected factory worker can assemble the part in 67
to 77 seconds?
Page | 21
(f)
What is the interquartile range (in seconds) for factory workers to assemble the part?
Solution
(a)
(b)
Step 1:
Draw diagram
for
Pr(X 75)
Step 2:
Normdist function
22
Step 2:
Normdist function
(d)
Step 1:
Draw diagram
for
Pr(X 75 X 85)
Step 2:
Normdist function
(e)
Step 1:
Draw diagram
for
Pr(X 67 X 77)
Page | 23
Step 2:
Normdist function
The interquartile range is the difference between the 25% and the 75% x-values, represented
by a and b in Figure 4.6.
Figure 4.6
There are two methods we could use to find the interquartile range.
24
Example 4-11
A college that specialises in training people for top level management positions administers various
types of physical and psychological tests. In one such test, objects must be assembled and pictures
arranged in order. Although the maximum possible score for this test is 140, the college has found that
results in the past have followed an approximately normal distribution with a mean of 100 and a
standard deviation of 10.
A company wishes to employ a very bright manager, someone who is in the top 10%. Find the cut off
score for a person in the top 10%, as measured by this test.
Page | 25
Solution
Figure 4.7
The x value corresponding to the top 10% is found using =norminv(.9,100,10) =112.8 as we must use
the probability to the left of the x value, ie 1 .1 = .9.
4.2.
References:
Anderson, D.R, et al Statistics for Business and Economics, Australian edition 1991.
Chapter 6, pages 168 - 179
Berenson, Levine, Krehbiel, Watson, Jayne and Turner, Basic Business Statistics: Concepts and
Applications, 2007, Prentice Hall
McLean, A., Stephens, B., Business Mathematics and Statistics, Longman 1996.
Chapter 21, pages 507 520
Selvanathan, A., Selvanathan, S., Keller, G., Warrack, B., Australian Business Statistics,
Abridged third edition, Nelson 2004.
26
USEFUL FORMULA:
z=
4.3.
Exercises
a. Pr (Z 1)
d. Pr(Z 2.5)
b. Pr (Z 1.5)
c. Pr (Z 2)
3. Given that Z is the standard normal random variable, compute the following probabilities.
a. Pr (Z - 1)
d. Pr(Z -2.5)
b. Pr (Z -1.5)
e. Pr (Z -3)
c. Pr (Z - 2)
4. Given that Z is the standard normal random variable, compute the following probabilities.
a. Pr (0 Z .83)
d. Pr (Z -.23)
b. Pr (-1.57 Z 0)
e. Pr (Z 1.20)
c. Pr (Z .44)
f. Pr (Z -.71)
Page | 27
5. Given that Z is the standard normal random variable, compute the following probabilities.
a. Pr ( - 1.98 Z .49)
b. Pr ( .52 Z 1.22)
c. Pr (-1.75 Z -1.04)
6. Given that Z is the standard normal random variable, find Z for each situation.
a.
c.
e.
g.
i.
b.
d.
f.
h.
j.
7. Given that Z is the standard normal random variable, find Z for each situation.
8. The demand for a new product is assumed to be normally distributed with = 200 and
= 40. Letting X be the number of units demanded, find the following:
a. Pr (X 250)
c. Pr ( 225 X 250)
b. Pr (X 100)
d. Pr (180 X 220)
9. When looking at the frequency distribution of a variable for which you have taken repeated
measurements, for example the daily output of a production line, what characteristics of the frequency
distribution may lead you to the conclusion that the variable is approximately normally distributed?
28
1.
2.
The life of a certain type of washing machine connector is normally distributed with a mean
life of 3.2 years and a standard deviation of 1.2 years. If this type of connector is
guaranteed for one year, what percentage of the original sales will require replacement?
3.
The Rural Bank is reviewing its service charges and interest-paying policies on cheque
accounts. The bank has found that the average daily balance on personal cheque accounts is
$550.00, with a standard deviation of $150.00. In addition, the average daily balances have
been found to be normally distributed.
(a) What percentage of personal cheque account customers carry average daily balances in
excess of $800.00?
(b) What percentage of the banks customers carry average daily balances below $200.00?
(c) What percentage of the banks customers carry average daily balances between $300.00
and $700.00?
(d) The bank is considering paying interest to customers carrying average daily balances in
excess of a certain amount. If the bank does not want to pay interest to more than 5% of
its customers, what is the minimum average daily balance it should be willing to pay
interest on?
The sales of High-Brite Toothpaste are believed to be approximately normally distributed,
with a mean of 10,000 tubes per week and a standard deviation of 1500 tubes per week.
(a) What is the probability that more than 12,000 tubes will be sold in any given week?
(b) In order to have a .95 probability that the company will have sufficient stock to cover
the weekly demand, how many tubes should be produced?
4.
Page | 29
5.
Figures on the residential property market in Melbourne over the past 100 years show the
annual return is normally distributed with mean 10.5 points and deviation 4.2 points.
(Note: the term point here is equivalent to percent, ie you can interpret one point as meaning
one percent.)
(a)
Sketch the distribution of annual return.
(b)
Find the probability that the residential property market will give a return of
(i)
More than 15 points.
(ii)
Less than 5 points.
(iii) Between 10 and 20 points.
(iv)
A negative return.
(c)
If the property market returns are only above the required rate of return for
professional investors 20% of the time, what is the required rate of return for
professional investors?
6.
Records kept by a producer of car tyres suggest that the distribution of the life of the tyres is
approximately normally distributed with a mean of 35,000 kilometres and a standard
deviation of 4000 kilometres.
(a)
What proportion of this companys tyres last
(i)
less than 30,000 kms
(ii)
less than 40,000 kms?
(b)
The companys advertising manager wishes to claim that 90% of our tyres last
longer than A kms. What should A be?
7.
A particular section of piping is mass produced. The distribution of the width of the piping
is approximately normal with a mean of 10 cms and a standard deviation of 0.025 cms.
Piping sections are discarded if the width exceeds 10.045 cms or is less than 9.955cms.
(a)
What proportion of piping sections have to be discarded?
(b)
How would this proportion decline if the standard deviation of the width were
reduced to 0.015 cms?
8.
Government officials estimate that the distribution of the Supplementary Security Income
(SSI) payments to those eligible is approximately normal with a mean of $120 per month
and a standard deviation of $35 per month.
(a)
What proportion of SSI recipients received
(i)
less than $100 per month?
(ii)
either less than $100 per month or more than $200 per month?
(b)
What is the 85th percentile of the distribution of SSI payments?
(c)
Interpret the number found in part (b) above.
30
Page | 31
32
Page | 33
34
Page | 35
36
Page | 37
38
Page | 39
40