1) Bandura, Blanchard, and Ritter (1969) conducted an experiment that provided individuals with one of four treatments for their intense fear of snakes. One group worked with a model - a person who handled a 4-foot king snake and encouraged others to imitate her. One group watched a film of adults and children who enjoyed progressively closer contact with a king snake. A third group received desensitization therapy, and a fourth group served as a control, receiving no treatment. The numbers of snake-approach responses after treatment are listed in the table. A higher number of snake-approach responses indicates that the individual was willing to approach a snake more often.
|
Type of Treatment
|
|
Model
|
Film
|
Desensitization
|
Control
|
|
29
|
22
|
21
|
13
|
|
27
|
18
|
17
|
12
|
|
27
|
17
|
16
|
9
|
Perform a hypothesis test to determine if there is a significant difference between the groups in the number of snake-approach responses. Show all necessary components of a full hypothesis test, including a source table.
2) What is the effect size for this result? Is this a large effect?
3) An employer wants to know what types of activities encourage her employees to exercise. The first month, she gave out information on the health benefits of exercise; the second, she sent daily emails encouraging physical activity; and in the third month, she gave health insurance bonuses for amount of exercise. She recorded the number of times 4 of her employees exercised in each of the three months. Can she determine if there is a significant difference in the number of days exercised between the three types of activities? Again, show all necessary components of a hypothesis test, including a source table.
|
Employee
|
Month 1 (Info)
|
Month 2 (Emails)
|
Month 3 (Bonuses)
|
|
1
|
4
|
7
|
29
|
|
2
|
6
|
7
|
26
|
|
3
|
2
|
11
|
24
|
|
4
|
3
|
9
|
21
|
4) Calculate a Tukey's HSD post-hoc test for the difference between Month 2 and Month 3. Is the difference significant? Was a post-hoc test necessary to determine this?
5) A market testing company wants to know if there is a difference in preference between two new types of chocolate bar. They also want to know if there is a gender difference.
a) How many cells would this ANOVA have?
b) What are the independent and dependent variables?
c) Here is the source table for this study. Write a conclusion for these results.
|
Source
|
SS
|
df
|
MS
|
F
|
|
Gender
|
6.24
|
1
|
6.24
|
6.50
|
|
Chocolate bar
|
20.24
|
1
|
20.24
|
21.08
|
|
Gender*Chocolate
|
8.99
|
1
|
8.99
|
9.36
|
|
Within
|
11.49
|
12
|
0.96
|
|
|
Total
|
46.96
|
15
|
|
|
d) What is the effect size for the interaction effect?
6) Find the mistakes this source table.
|
Source
|
SS
|
df
|
MS
|
F
|
|
Between
|
1621
|
2
|
810.5
|
81.318
|
|
Subjects
|
45.825
|
5
|
91.65
|
0.919
|
|
Within
|
99.675
|
10
|
9.967
|
|
|
Total
|
1866.50
|
17
|
109.79
|
|
7) What type of ANOVA is the source table in #6 for?
8) What is the main factor that differentiates whether you should calculate a t statistic or an F statistic?
9) How many types of ANOVA have we covered? What feature or features distinguish between them?
10) Look back to the graphic you created to help you choose which z or t test to use (or use one of the graphics from the answer key). Modify the graphic to include the different types of ANOVA from this unit.