BUSI3007

BUSINESS RESEARCH METHODS

LECTURE 9: ANALYSIS OF VARIANCE (ANOVA)

REFERENCE

Statistical techniques in Business and Economics, Lind, Marchal

and Wathen, 16th edition

Chapter 12 (pg. 380 - 394)

2
OVERVIEW
The F Distribution
Comparing Two Population Variances
The ANOVA Test

3
THE F DISTRIBUTION
The distribution is:
Used to test a hypothesis of equal population variances.
Used to simultaneously test a hypothesis that several populati
on means are equal. The simultaneous comparison of several
population means is called analysis of variance (ANOVA).

4
CHARACTERISTICS OF
A F - DISTRIBUTION
1. There is a family of F Distributio
ns. A particular member of the fa
mily is determined by two param
eters: the degrees of freedom in t
he numerator and the degrees of
freedom in the denominator.
2. The F distribution is continuous
3. F value cannot be negative.
4 The F distribution is positively sk
ewed.
5. It is asymptotic. As F the cur
ve approaches the X-axis but nev
er touches it.

5
COMPARING TWO POP
ULATION VARIANCES
The F distribution is used to test the hypothesis that the variance
of one normal population equals the variance of another normal p
opulation.

Example:
The mean rate of return on two types of common stock may be th
e same, but there may be more variation in the rate of return in on
e than the other. A sample of 10 technology and 10 utility stocks s
hows the same mean rate of return, but there is likely more variati
on in the Internet stocks.

6
TEST FOR EQUAL VAR
IANCES

H 0 : 12 22
H1 : 12 22
To conduct the test, we select a random sample of n1 observation
s from one population, and a random sample of n2 observations fr
om the second population. The test statistics is defined as follows:

s12
F 2
s2

7
TEST FOR EQUAL VAR
IANCES
The term s12 and s22 are the respective sample variances.

If the null hypothesis is true, the test statistic follows the F distribut
ion with n1 - 1 and n2 - 1 degrees of freedom.

In order to reduce the size of the table of critical values, the larger
sample variance is placed in the numerator. The tabled F ratio is a
lways larger than 1.00. The right-tail critical value is the only one r
equired.

8
TEST FOR EQUAL VAR
IANCES
Lammers Limos offers limousine servi
ce from the city hall in Toledo, Ohio, to
Metro Airport in Detroit. The president
of the company, is considering two rou
tes. One is via U.S. 25 and the other vi
a I-75. He wants to study the time it ta
kes to drive to the airport using each r
oute and then compare the results. He
collected the following sample data, w
hich is reported in minutes.
Using the .10 significance level, is ther
e a difference in the variation in the dri
ving times for the two routes?

9
TEST FOR EQUAL VAR
IANCES

10
TEST FOR EQUAL VAR
IANCES
Step 1: The hypotheses are:
H0: 12 = 22
H1: 12 22

Step 2: The significance level is .10.

Step 3: The test statistic is the F distribution.

11
TEST FOR EQUAL VAR
IANCES
Step 4: State the decision rule.
This is a two-tailed test, divide by 2: 0.1/2 = 0.05.
There are n1 - 1 = 7 1 = 6 degrees of freedom in the numerator.
There are n2 - 1 = 8 1 = 7 degrees of freedom in the denominato
r.
Based on these information, the critical value = 3.87.
Reject H0 if F > 3.87.

12
TEST FOR EQUAL VAR
IANCES

13
TEST FOR EQUAL VAR
IANCES
Step 5: Compute the value of F and make a decision.

The decision is to reject the null hypothesis, because the compute

d F value (4.23) is larger than the critical value (3.87).

Step 6: Interpret the result.

The data indicates that there is a difference in the variation of the
travel times along the two routes.

14
TEST FOR EQUAL VAR
IANCES
A stock broker at Critical Securities reported that the mean rate of
return on the sample of 10 oil stocks was 12.6 percent with a stan
dard deviation of 3.9 percent. The mean rate of return on a sampl
e of 8 utility stocks was 10.9 percent with a standard deviation of
3.5 percent. At the 0.05 significance level, can we conclude that th
ere is more variation in the oil stocks?

15
TEST FOR EQUAL VAR
IANCES
Step 1: The hypotheses are:
H0: 12 22
H1: 12 > 22

Step 2: The significance level is 0.05.

Step 3: The test statistic is the F distribution.

16
TEST FOR EQUAL VAR
IANCES
Step 4: State the decision rule.
This is a one-tailed test, = 0.05.
There are n1 - 1 = 10 1 = 9 degrees of freedom in the numerato
r.
There are n2 - 1 = 8 1 = 7 degrees of freedom in the denominato
r.
Based on these information, the critical value = 3.68.
Reject H0 if F > 3.68.

17
TEST FOR EQUAL VAR
IANCES
Step 5: Compute the value of F and make a decision

s12 0.039
2
F 2 1.24
s2 0.035 2

The decision is not to reject the null hypothesis, because the com
puted F value (1.24) is less than the critical value (3.68).

Step 6: Interpret the result.

Based on the sample evidence, we conclude that oil stocks return
s do not have more variation.

18
COMPARING MEANS OF TH
REE OR MORE POPULATIO
NS
The F distribution is also used in the analysis of variance (ANOV
A) technique in which we compare three or more population mean
s to determine whether they could be equal.
Assumptions:
1. The populations follow the normal distribution.
2. The populations have equal standard deviations .
3. The populations are independent.

19
COMPARING MEANS OF TH
REE OR MORE POPULATIO
NS
The Null Hypothesis is that the population means are all the sam
e. The Alternative Hypothesis is that at least one of the means is
different.
The Test Statistic is the F distribution.
The Decision rule is to reject the null hypothesis if F (computed) i
s greater than F (table) with numerator and denominator degrees
of freedom.
Hypothesis Setup and Decision Rule:
H0: 1 = 2 == k
H1: The means are not all equal

Reject H0 if F > Critical Value

20
ANALYSIS OF VARIAN
CE
If there are k populations being sampled, the numerator degrees
of freedom is k 1.

If there are a total of n observations the denominator degrees of fr

eedom is n k.

The test statistic is computed by:

21
ANALYSIS OF VARIAN
CE
Grand Mean:

where n is the total number of observations.

X is each sample observation.

22
ANALYSIS OF VARIAN
CE
Total Variation (SS Total)
The sum of the squared differences between each observation an
d the grand mean

23
ANALYSIS OF VARIAN
CE
Treatment Variation (SST)
The sum of the squared differences between each treatment mea
n and the grand mean

Random Variation (SSE)

The sum of the squared differences between each observation an
d its treatment mean.

where is the sample mean for treatment c.

24
ANALYSIS OF VARIAN
CE

25
ANALYSIS OF VARIAN
CE
A real estate developer is considering investing in a shopping mail
on the outskirts of Atlanta, Georgia. Three parcels of land are bein
g evaluated. Of particular importance is the income in the area su
rrounding the proposed mail. A random sample of four families is
selected near each proposed mail. Following are the sample resul
ts. At the 0.05 significance level, can the developer conclude ther
e is a difference in the mean income?

Southwyck Area Franklin Park Old Orchard

($000) ($000) ($000)
64 74 75
68 71 80
70 69 76
60 70 78

26
ANALYSIS OF VARIAN
CE
Step 1: State the null and alternate hypotheses.
H0: 1 = 2 = 3
H1: The means are not all equal

Step 2: State the level of significance.

The .05 significance level is stated in the problem.

Step 3: Find the appropriate test statistic.

Use the F statistic

27
ANALYSIS OF VARIAN
CE
Step 4: State the decision rule.
= 0.05

Numerator degrees of freedom = k 1 = 3 1 = 2

Denominator degrees of freedom = n k = 12 3 = 9

Based on the above information, the critical value = 4.26

Reject H0 if F > 4.26

28
ANALYSIS OF VARIAN
CE
Step 5: Compute the value of F and make a decision

Compute the grand mean.

29
ANALYSIS OF VARIAN
CE
Compute total variation (SS Total)

30
ANALYSIS OF VARIAN
CE
Compute random variation (SSE)

31
ANALYSIS OF VARIAN
CE

32
ANALYSIS OF VARIAN
CE

Since F > 4.26, reject H0.

Step 6: Interpret the result.

Based on the sample evidence, we can conclude that the mean in
comes are not all equal.

33
THE ANOVA TABLE
It is convenient to summarize the calculations of the F statistic in
an ANOVA table. The format is as follows:

ANOVA Table
Source of Sum of Squares Degrees of Mean Square F
Variation Freedom
Treatments SST k1 MST MST/MSE
Error SSE nk MSE

Total SS Total n1

34
THE ANOVA TABLE
Substituting the information from the previous example in the AN
OVA table.

ANOVA Table
Source of Sum of Squares Degrees of Mean Square F
Variation Freedom
Treatments 276.5 2 138.25 14.18
Error 87.75 9 9.75

Total 364.25 11

35
ANALYSIS OF VARIAN
CE
Three assemble lines are used to produce a certain component fo
r an airliner. To examine the production rate, a random sample of
six hourly periods is chosen for each assembly line and the numb
er of components produced during these periods for each line is r
ecorded. The output from a statistical software package is:

Summary

Groups Count Sum Average Variance

Line A 6 250 41.66667 0.266667

Line B 6 260 43.33333 0.666667

Line C 6 249 41.5 0.7

36
 ANALYSIS OF VARIAN
CE
ANOVA Table
Source of Sum of Squares Degrees of Mean Square F
Variation Freedom
Between Groups 12.33333 2 6.166667 11.32653
Within Groups 8.166667 15 0.544444

Total 20.5 17

Use a 0.01 level of significance to test if there is a difference in th

e mean production of the three assembly lines.

37
ANALYSIS OF VARIAN
CE
Step 1: State the null and alternate hypotheses.
H0: 1 = 2 = 3
H1: The mean productions are not all equal

Step 2: State the level of significance.

The .01 significance level is stated in the problem.

Step 3: Find the appropriate test statistic.

Use the F statistic

38
ANALYSIS OF VARIAN
CE
Step 4: State the decision rule.

Numerator degrees of freedom = k 1 = 2

Denominator degrees of freedom = n k = 15
Based on the above information, the critical value = 6.36
Reject H0 if F > 6.36

39
ANALYSIS OF VARIAN
CE
Step 5: Compute the value of F and make a decision
From the ANOVA table, F = 11.32653
Since F > 6.36, reject H0

Step 6: Interpret the result.

Based on the sample evidence, we can conclude that the mean p
roductions are not all equal.

40