Analysis of Variance

Chapter 12

Copyright © 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of McGraw-Hill Education.
Learning Objectives
LO12-1 Apply the F distribution to test a hypothesis that
two population variances are equal.
LO12-2 Use ANOVA to test a hypothesis that three or
more population means are equal.
LO12-3 Use confidence intervals to test and interpret
differences between pairs of population means.
LO12-4 Use a blocking variable in a two-way ANOVA to
test a hypothesis that three or more population
means are equal.
LO12-5 Perform a two-way ANOVA with interaction and
describe the results.

12-2
 LO12-1 Apply the F distribution to test a hypothesis
that two population variances are equal.

Testing the Hypothesis of Two Equal

Population Variances: The F Distribution
 The distribution is named to honor Sir Ronald Fisher, one of the
founders of modern-day statistics.

 The distribution is:

 Used to test a hypothesis of equal population variances.
 Used to simultaneously test a hypothesis that several population means
are equal. The simultaneous comparison of several population means is
called analysis of variance (ANOVA).

12-3
 LO12-1

Characteristics of a F-
Distribution
 There is a “family” of F- distributions.
A particular member of the family is
determined by two parameters: the
degrees of freedom in the numerator
and the degrees of freedom in the
denominator.
 The F-distribution is continuous.
 A F-value cannot be negative.
 The F-distribution is positively
skewed.
 It is asymptotic. As F  , the curve
approaches the X-axis but never
touches it.

12-4
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances
H0: σ12 = σ22
H1: σ12 ≠ σ22
The F-distribution is used to test the hypothesis that the variance of one normal
population equals the variance of another normal population.

Examples:
 Two Barth shearing machines are set to produce steel bars of the same length. The
bars, therefore, should have the same mean length. We want to ensure that in
addition to having the same mean length they also have similar variation.
 The mean rate of return on two types of common stock may be the same, but there
may be more variation in the rate of return in one than the other. A sample of 10
technology and 10 utility stocks shows the same mean rate of return, but there is likely
more variation in the technology stocks.
 A study by the marketing department for a large newspaper found that men and
women spend about the same amount of time per day reading the paper. However,
the same report indicated there was nearly twice as much variation in time spent per
day among the men than the women.
12-5
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example
Lammers Limos offers limousine service
from the city hall in Toledo, Ohio, to Metro
Airport in Detroit. The president of the
company, is considering two routes. One
is via U.S. 25 and the other via I-75. He
wants to study the time it takes to drive to
the airport using each route and then
compare the results. He collected the
following sample data, which is reported
in minutes.

Using the .10 significance level, is there a

difference in the variation in the driving
times for the two routes?

12-6
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example
Computing the sample means and variances:

12-7
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example
Step 1: State the null and alternate hypotheses.
H0: σ12 = σ22
H1: σ12 ≠ σ22

Step 2: Select a level of significance.

The requested significance level is .10.

Step 3: Select the test statistic.

The appropriate test statistic to test a hypothesis
of equal variances is the F statistic.
12-8
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example
Step 4: State the decision rule. The F-statistic is computed as
the ratio of two variances. The numerator is always the
larger of the two variances with its corresponding degrees
of freedom.

Reject H0 if F > F/2,v1,v2

F > F.10/2,7-1,8-1
F > F.05,6,7

12-9
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example
Step 5: Compute the value of F and make a decision.

The decision is to reject the null hypothesis because the computed 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.

12-10
 LO12-1

Testing the Hypothesis of Two Equal

Population Variances – Example

12-11
 LO12-2 Use ANOVA to test a hypothesis that
three or more population means are equal.

Testing the Hypothesis of Three or More

Equal Population Means
The F-distribution is also used for testing whether
two or more sample means came from the same
or equal populations.

Assumptions:
 The sampled populations follow the
normal distribution.
 The populations have equal
standard deviations.
 The samples are randomly selected
and are independent.

12-12
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means
 The null hypothesis is when the population means are all the
same.
 The alternative hypothesis is when at least one of the means is
different.
 The test statistic is the F distribution.
 The decision rule is whether to reject the null hypothesis if F
(computed) is greater than F (table) with numerator and
denominator degrees of freedom.
 Hypothesis Setup and Decision Rule:

H0: µ1 = µ2 =…= µk
H1: The means are not all equal.
Reject H0 if F > F,k-1,n-k

12-13
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means - Illustrated

Joyce Kuhlman manages a

regional financial center. She
wishes to compare the
productivity, as measured by the
number of customers served,
among three employees. Four
days are randomly selected and
the number of customers served
by each employee is recorded.

1412-
12-14
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
Recently a group of four major carriers
joined in hiring Brunner Marketing
Research, Inc., to survey recent
passengers regarding their level of
satisfaction with a recent flight. The
survey included questions on
ticketing, boarding, in-flight service,
baggage handling, pilot
communication, and so forth.

Twenty-five questions offered a range

of possible answers: excellent, good,
fair, or poor. A response of excellent Is there a difference in the mean
was given a score of 4, good a 3, fair satisfaction level among the four airlines?
a 2, and poor a 1. These responses Use the .01 significance level.
were then totaled, so the total score
was an indication of the satisfaction
with the flight. Brunner Marketing
Research, Inc., randomly selected and
surveyed passengers from the four
airlines.

12-15
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example

Computing the “treatment” and grand means:

12-16
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
Step 1: State the null and alternate hypotheses.

H0: µN = µW = µP = µB
H1: The means 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.

Because we are comparing means of more than
two groups, use the F statistic.

12-17
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
 Step 4: Formulate a decision rule. The F-statistic will be
used to formulate the decision rule. The F-statistic is a ratio
of two variances, each divided by their degrees of freedom.
These are called mean squares. For this ANOVA, we will
divide the treatment mean square by the error mean square.
Therefore, we need the degrees of freedom for treatments
and error to find the F value for the decision rule.

The degrees of freedom in the numerator:

(Number of treatments – 1) = (k - 1) = 4 - 1 = 3

The degrees of freedom in the denominator:

(Total number of observations – Number of treatments) = (n –
k) = (22 - 4) = 18
12-18
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
From the F-table with the .01 level of significance, the critical
value of F with 3 numerator and 18 denominator degrees of
freedom is 5.09.
Numerator degrees of freedom
Reject H0 if F > F,k-1,n-k

or
Denominator degrees of

Reject H0 if F > 5.09

freedom

12-19
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
Step 5: Compute the value of F and make a decision.
Creating the Analysis of Variance table:

12-20
 LO12-2

Creating the ANOVA Table: Computing

SS Total and SSE
Computing the total sum of Computing the error sum of
squares: squares:

12-21
 LO12-2

Creating the ANOVA Table: Treatment Sum

of Squares, SST, and the ANOVA Table

Step 5 (continued): Compute the value of F and make a decision. The

computed value of F is 8.99, which is greater than the critical value of
5.09, so the null hypothesis is rejected.

12-22
 LO12-2

Testing the Hypothesis of Three or More

Equal Population Means – Example
Step 6: Interpret the result. The population means are
not all equal. The mean scores are not the
same for the four airlines; at this point we can
only conclude there is a difference in at least
one pair of treatment means. We cannot
determine which of the airlines’ satisfaction
mean scores differ.

12-23
 LO12-2

Testing the Hypothesis of Three or

More Equal Population Means – Excel

12-24
 LO12-3 Use confidence intervals to test and interpret
differences between pairs of population means.

Testing for Differences Between Pairs of

Population Means
When we reject the null hypothesis that the means are equal, we
may want to know which treatment means differ. One of the simplest
procedures is to use confidence intervals for the difference between
two means.

12-25
 LO12-3

Testing for Differences Between Pairs of

Population Means – Example
From the previous example, develop a 95% confidence interval for
the difference in the mean customer satisfaction between Northern
and Branson airlines. Can we conclude that there is a difference
between the two airlines’ ratings?

12-26
 LO12-3

Testing for Differences Between Pairs of

Population Means – Minitab Output
The differences in mean satisfaction ratings between each pair of
airlines can be obtained directly from Minitab using the one-way
ANOVA analysis and Fisher’s method to compare means. The Minitab
output follows. If zero is in the interval, the means are not different.

12-27
 LO12-4 Use a blocking variable in a two-way ANOVA to test a
hypothesis that three or more population means are equal.

Two-way ANOVA: Testing the Hypothesis of

Three or More Equal Population Means Using
a Blocking Variable
 For the two-factor ANOVA, we test whether there is a
significant difference between the treatment effect
and whether there is a difference in the blocking
effect.
 The two-way ANOVA table now includes a “blocks”
source of variation in addition to treatment, error, and
total. Notice there are “b” blocks.

12-28
 LO12-2

Two-Way ANOVA: Testing the Hypothesis of Three or

More Equal Population Means using a Blocking Variable
WARTA, the Warren Area Regional
Transit Authority, is expanding bus
service from the suburb of Starbrick
into the central business district of
Warren. There are four routes being
considered from Starbrick to
downtown Warren: (1) via U.S. 6, (2)
via the West End, (3) via the Hickory
Street Bridge, and (4) via Route 59.

WARTA conducted several tests to determine whether there was a difference in the mean
travel times along the four routes. Because there will be many different drivers, the test
was set up so each driver drove along each of the four routes. The next slide shows the
travel time, in minutes, for each driver-route combination. At the .05 significance level, is
there a difference in the mean travel time along the four routes? If we remove the effect of
the drivers, is there a difference in the mean travel time?

This is a two-way ANOVA. The routes are the treatments and the drivers are the blocks.
12-29
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of

Three or More Equal Population Means
Using a Blocking Variable – Example
Observed Sample Data:

12-30
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of

Three or More Equal Population Means Using
a Blocking Variable – Example
Step 1: State the null and alternate hypotheses.

H0: µu = µw = µh = µr
H1: Not all treatment means are the same.

Step 2: State the level of significance.

The .05 significance level is stated in the problem.
Step 3: Find the appropriate test statistic.
Because we are comparing means of more than two groups,
use the F-statistic.

12-31
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of

Three or More Equal Population Means
Using a Blocking Variable – Example
 Step 4: Formulate a decision rule. The F-statistic will be
used to formulate the decision rule. The F-statistic is a ratio
of two variances, each divided by their degrees of freedom.
These are called mean squares. For the two-way ANOVA,
we will divide the treatment mean square by the error mean
square. Therefore, we need the degrees of freedom for
treatments and error to find the F value for the decision rule.

The degrees of freedom in the numerator:

(Number of treatments – 1) = (k – 1) = 4 – 1 = 3

The degrees of freedom in the denominator:

(Number of treatments – 1)(Number of blocks – 1) =
(k – 1)(b - 1) = (4 – 1)(5 – 1) = (3)(4) = 12
12-32
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of

Three or More Equal Population Means
Using a Blocking Variable – Example
From the F-table with the .05 level of significance, the critical
F value with 3 numerator and 12 denominator degrees
of freedom is 3.49.
Numerator degrees of freedom

Step 4: State the decision rule.

Reject H0 if F > F,v1,v2
Denominator degrees of freedom

F > F.05,k-1,(k – 1)(b – 1)

F > F.05,4-1,(4 – 1)(5 – 1)
F > F.05,3,12
F > 3.49

12-33
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of Three or

More Equal Population Means Using a Blocking Variable
Calculating the Block Sum of Squares

SSB  k( x b  x G ) 2

12-34
 LO12-4

Two-Way ANOVA: Testing the Hypothesis of Three or More

Equal Population Means – Example Using a Blocking Variable

Step 5: Compute the value of F and make a decision. The computed value of F is
7.93, which is greater than the critical value of 3.49, so the null hypothesis is rejected.

Step 6: Interpret the Result. At least one pair of mean route times are different.
12-35
 LO12-4
Two-Way ANOVA: Testing the Hypothesis of Three or
More Equal Population Means Using a Blocking
Variable – Excel Example

Using Excel to perform the calculations, we conclude:

(1) The mean time is not the same for all drivers.
(2) The mean times for the routes are not all the same. 12-36
 LO12-5 Perform a two-way ANOVA with
interaction and describe the results.

Two-Way ANOVA with Interaction

INTERACTION The effect of one factor on a
response variable differs depending on the value of
another factor.
 In the previous example, we studied the separate or
independent effects of two variables, routes into the city and
drivers, on mean travel time.
 There is another effect that may influence travel time. This is
called an interaction or combined effect of route and driver on
travel time. For example, is it possible that one of the drivers
is especially good driving one or more of the routes?
 To measure interaction effects, it is necessary to have at least
two observations in each cell.

12-37
 LO12-5

Two-Way ANOVA with Interaction

 When we use a two-way ANOVA to study interaction, we call the two
variables factors instead of treatments and blocks.
 Interaction occurs if the combination of two factors has some
combined effect on the variable under study, in addition to each
factor alone.
 The variable being studied is referred to as the response variable.
 One way to study interaction is by plotting factor means in a graph
called an interaction plot.

12-38
 LO12-5

Graphical Observation of Interaction

Between Driver and Route
Our graphical observations show us that interaction
effects are possible. For example, Deans (green
line) is not the fastest driver on all the routes. So,
the travel times depend on both the driver and the
route driven.
The next step is to conduct statistical tests of
hypothesis to further investigate the possible
interaction effects. In summary, our study of travel
times has several questions:

 Is there a significant interaction between routes

and drivers?

 Are the travel times for the drivers the same?

 Are the travel times for the routes the same?

Of the three questions, we are most interested in

the test for interactions. To put it another way, does
a particular route/driver combination result in
significantly faster (or slower) driving times?
Also, the results of the hypothesis test for
interaction affect the way we analyze the route and
driver questions.

12-39
 LO12-5

Two-way ANOVA with Replication –

Example
Suppose the WARTA
blocking experiment
discussed earlier is
repeated by measuring
two more travel times
for each driver and
route combination with
the data shown in the
worksheet.

12-40
 LO12-5

The two-way ANOVA with interaction now

has three sets of hypotheses to test:
1. H0: There is no interaction between drivers and routes.
H1: There is interaction between drivers and routes.

2. H0: The driver means are the same.

H1: The driver means are not the same.

3. H0: The route means are the same.

H1: The route means are not the same.

12-41
 LO12-5

Two-way ANOVA Table with

Interaction

The interaction of driver and route is significant, the

p-value is less than .05.
12-42
 LO12-5

Analysis of Interaction: One-way ANOVAs

for Each Route
H0: For each route, the mean driver times are equal.
H1: For each route, at least one pair of driver times are not equal.

12-43