CHAPTER 7

Hypothesis Testing
• Purpose of ANOVA
• Post hoc multiple comparison
• Software examples using Excel & SAS
• Interpretation of Excel and SAS outputs
 MULTIPLE COMPARISON TEST (POST-HOC)
If our test of the null hypothesis is rejected, we conclude that not all the means
are equal.
 At least one mean is different from the other means
 The ANOVA test itself provides only statistical evidence of a difference, but not
any statistical evidence as to which mean or means are statistically different.

“At least one of the majors average scores

is different from the others.”
MULTIPLE COMPARISON TEST (POST-HOC)
Multiple comparisons conducts an analysis of all possible pairwise means. For
example, with three majors of the students, BBA, SCI, and MUSIC, if the
ANOVA test was significant, then multiple comparison methods would compare
the three possible pairwise comparisons:



MULTIPLE COMPARISON TEST (POST-HOC)

ANOVA

SSB
SSW
SST
 MULTIPLE COMPARISON TEST (POST-HOC)

Scheffe’s Test
Mean with the same letter are not
significantly different.
Scheffe Grouping Mean n Group
A 69.7143 7 BBA
B 53.1667 6 SCI
C 30.4000 5 MUSIC
 MULTIPLE COMPARISON TEST (POST-HOC)
Scheffe’s Test Conclusion for Scheffe:

Mean with the same letter are not • The average scores of all three populations are
significantly different. significantly different.
Scheffe Grouping Mean n Group
A 69.7143 7 BBA • The average score of music students is significantly
less than the others.
B 53.1667 6 SCI
C 30.4000 5 MUSIC • The average score of BBA students is significantly
higher than the others.
ANOV // POST HOC
MULTIPLE COMPARISON TEST (POST-HOC)
Q1: A computer products company sets up a telephone service to assist
customers by providing technical support. The average wait time during service
hours is 37 minutes. However, a recent hiring effort added technical consultants
to the system, and management believes that the average wait time decreased,
and they want to prove it.

Q2: A computer products company sets up a telephone service to assist

customers by providing technical support. The manager believe that the
average time of weekday should be lower than weekend. Test the manager’s
claim using 5% significance level.

Q3: A computer products company has three operation teams (Team 1, Team 2,
and Team 3) handling customer support tickets. The company wants to
determine if there is a significant difference in the average resolution time
among these teams. Using a significance level of 5%, test whether there are
significant differences in the mean resolution times among the three teams.
 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Step1: State the hypotheses (symbolic forms and the context of problems)
 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Step 2: Run ANOVA

 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Step 3: Interpret the result

P-value = ……………………….

Step 4: Making decision and conclusion

 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Step 5: Should we run post hoc analysis?

 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Scheffe’s Test

Mean with the same letter are not

significantly different.
Scheffe Grouping Mean n Group
34.0909 11 2
33.4000 15 3
26.5714 14 1
 Case 1: Customer service
Q3: A computer products company has three operation teams (Team 1, Team 2, and Team 3) handling customer
support tickets. The company wants to determine if there is a significant difference in the average resolution time among
these teams. Using a significance level of 5%, test whether there are significant differences in the mean resolution times
among the three teams.

Scheffe’s Test Conclusion for Scheffe:

Mean with the same letter are not

significantly different.
Scheffe Grouping Mean n Group
A 34.0909 11 2
A 33.4000 15 3
B 26.5714 14 1
Case 2: Production
Q4: A manufacturing company wants to test to determine whether the average
thickness of a plastic bottle is 2.4 millimeters.

Q5: A manufacturing company wants to test to determine whether the average

thickness of a plastic bottle compared between 2 machines used in the
production (Machine A & Machine B). The production manager believe that the
average thickness from these two machines should be the same. Test the
manager’s claim using 5% significance level

Q6: A manufacturing company produces plastic bottles in three different

operation shifts (Morning shift: 6AM-2PM, Afternoon shift: 2PM-10PM, and
Night shift: 10PM-6AM). The company wants to determine if there is a
significant difference in the average thickness of plastic bottles produced during
these three shifts. Using a significance level of 5%, test whether there are
significant differences in the mean thickness among the three operation shifts.
Case 2: Production
Q6: A manufacturing company produces plastic bottles in three different operation shifts (Morning shift: 6AM-2PM,
Afternoon shift: 2PM-10PM, and Night shift: 10PM-6AM). The company wants to determine if there is a significant
difference in the average thickness of plastic bottles produced during these three shifts. Using a significance level of
5%, test whether there are significant differences in the mean thickness among the three operation shifts.

Step1: State the hypotheses (symbolic forms and the context of problems)
Case 2: Production
Q6: A manufacturing company produces plastic bottles in three different operation shifts (Morning shift: 6AM-2PM,
Afternoon shift: 2PM-10PM, and Night shift: 10PM-6AM). The company wants to determine if there is a significant
difference in the average thickness of plastic bottles produced during these three shifts. Using a significance level of
5%, test whether there are significant differences in the mean thickness among the three operation shifts.

Step 2: Run ANOVA

 Case 2: Production
Q6: A manufacturing company produces plastic bottles in three different operation shifts (Morning shift: 6AM-2PM,
Afternoon shift: 2PM-10PM, and Night shift: 10PM-6AM). The company wants to determine if there is a significant
difference in the average thickness of plastic bottles produced during these three shifts. Using a significance level of
5%, test whether there are significant differences in the mean thickness among the three operation shifts.

Step 3: Interpret the result

P-value = ……………………….

Step 4: Making decision and conclusion

 Case 2: Production
Q6: A manufacturing company produces plastic bottles in three different operation shifts (Morning shift: 6AM-2PM,
Afternoon shift: 2PM-10PM, and Night shift: 10PM-6AM). The company wants to determine if there is a significant
difference in the average thickness of plastic bottles produced during these three shifts. Using a significance level of
5%, test whether there are significant differences in the mean thickness among the three operation shifts.

Step 5: Should we run post hoc analysis?

ANOV // POST HOC
Case 3: Demographic
Q7: A retail store wants to test to determine whether the average age of its
customers is more than 30 years old.

Q8: Based on one sample test, we have checked that the average age of the
customers is around 30 years old. However, the data was collected from two
locations (City vs Suburb area). The manager of this retail store believe that
average age of the customers in city area location should be younger than
customers in suburb area location. Test the manager’s claim.

Q9: A retail store operates in three different locations: Branch 1, Branch 2, and
Branch 3. The company wants to determine if there is a significant difference in
the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the
mean customer ages across the three locations.
Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Step1: State the hypotheses (symbolic forms and the context of problems)
 Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Step 2: Run ANOVA

 Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Step 3: Interpret the result

P-value = ……………………….

Step 4: Making decision and conclusion

 Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Step 5: Should we run post hoc analysis?

 Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Scheffe’s Test

Mean with the same letter are not

significantly different.
Scheffe Grouping Mean n Group
 Case 3: Demographic
Q9: A retail store operates in three different locations: Branch 1, Branch 2, and Branch 3. The company wants to
determine if there is a significant difference in the average age of customers among these three locations. Using a
significance level of 5%, test whether there are significant differences in the mean customer ages across the three
locations.

Scheffe’s Test Conclusion for Scheffe:

Mean with the same letter are not

significantly different.
Scheffe Grouping Mean n Group
A 33.4545 11 3
B 30.2308 13 2
C 29.7714 11 1