1)

Cohen's d and R-squared are similar in that they

a. are both measures of effect size

b. are both expressed as a percentage of variance

c. are both used in ANOVA

d. are both calculated the exact same way

2)

The magnitude of effect is

a. The proportion of variance in the DV due to error

b. The proportion of variance in the IV due to error

c. The proportion of variance in the IV due to the treatment

d. The proportion of variance in the DV due to the IV

3)

Assume you conducted a study consisting of 5 groups--each group having 9 participants. (45 participants total)

The Means of the groups are as follows:

Group A: 5.00

Group B: 7.00

Group C: 8.00

Group D: 9.00

Group E: 10.00

You created the following source table from the omnibus ANOVA:

SSbetween: 365.00

SSwithin: 1040.00

SS total: 1405.60

dfbetween: 4

dfwithin: 40

dftotal:44

MSbetween: 91.40

MSwithin:26.00

F: 3.52

P: <.05

Now, assume that you want all possible comparisons between the 5 groups and you elected to run a post hoc Tukey.Why would you decide to run a Tukey rather than just conduct a bunch of t-tests to test the different means between the groups?

a. Because I made a priori hypotheses

b. Because I am making more than 3 comparisons

c. To keep the familywise error rate at .05

d. All of the above

4)

What would the q value be for the Tukey test between group A and group E?

a. 0.72

b. 1.73

c. 2.94

d. 5.86

5)

What would be the critical value you should use for the above Tukey test?Assume you are conducting the test at the .05 level.(q table is available under files on Canvas)

a. 4.04

b. 4.76

c. 4.41

d. 3.79

6)

According to your Tukey results is group A significantly different than group E?

a. Yes

b. No

7)

In the formula, (M-GM)2/dfBetween, "GM" is the

a. Group Mean

b. General Multiple

c. Grand Mean

d. Grand Median

8)

In an analysis of variance, if the null hypothesis is true, then

a. there is less variance among means of samples than if the null hypothesis were not true.

b. fewer participants are needed than with a t test for independent means.

c. difference scores are computed as in a t test for dependent means.

d. a minimum of 4 groups must be involved in the experiment.

9)

When would you use a Tukey test?

a. Following a significant t-test

b. When your F value is low to find out why

c. When the omnibus test is significant

d. When you want to determine effect size for an ANOVA

10)

In the question above about Groups A through E what would be the effect size? (round to two decimal places)

SSbetween: 365.00

SSwithin: 1040.00

SS total: 1405.60

dfbetween: 4

dfwithin: 40

dftotal:44

MSbetween: 91.40

MSwithin:26.00

F: 3.52

P: <.05

11)

Assume that you know the Between groups Mean Square for an ANOVA problem comparing 3 groups is 100.What would the SS between be?