1. Suppose that the F ratio you have calculated for a one-way ANOVA is 23. Which of the following can you conclude?
a. A calculation error has probably been made (F values this high don’t happen in real life)
b. The null hypothesis can be rejected at .05
c. The group means must have been spread far apart
d. There must have been very little within-group variability
e. The same size must have been large

2. A study compares five groups with 10 observations in each group. An F statistic of 3.75 is reported. What are the degrees of freedom for this F statistic?
a. 4 and 45 b. 5 and 10
c. 4 and 46 d. 5 and 50

3. A study compares three population means. Three independent samples with 15 observations each are taken. The SSWithin = 1246 and the SSTotal=1600. What is the value of the F statistic?
a. 1.11 b. 3.32
c. 4.98 d. 5.97

4. In a study comparing four groups with six observations in each group, the MSWithin = 487 and the MSBetween= 3456. What proportion of the variance in the dependent variable is explained by the independent variable?
a. 0.484 b. 0.516
c. 0.587 d. 0.939

Use the following to answer questions 5 and 6:
A realtor wishes to assess whether a difference exists between home prices in three subdivisions. Independent samples of homes from each of the three subdivisions are obtained and their prices are recorded. The analysis of variance results for comparing these prices are provided below:
Source Sums of Squares DF Mean Square F
Groups 157.44 2 78.72 4.66
Error 253.50 15 16.90
Total 410.94 17
5. How many homes were sampled in total?
a. 15 b. 17
c. 18 d. 19

6. What can we say about the p value for this F test?
a. p value <0.01 b. 0.1 < p < 0.05
c. 0.01 < p < 0.05 d. 0.05 < p < 0.10




7. Which of the following tends to increase the size of the F ratio in a RM ANOVA?
a. Larger subject-to-subject differences (averaging across treatment levels)
b. Better matching of the scores across treatments
c. Larger subject by treatment interaction
d. Smaller differences between treatment means

8. An RM ANOVA on a certain set of data did not attain statistical significance. What advantage would be derived from performing an independent groups ANOVA on the same data?
a. A larger alpha level would normally be used
b. The numerator of the F ratio would probably be larger
c. The critical F would probably be smaller
d. The error term would probably be larger

9. The interaction for a particular two-way ANOVA is statistically significant. This implies that:
a. Neither of the main effects is significant
b. Both of the main effects are significant
c. The lines on the graph of cell means will intersect
d. The lines on a graph of cell means will not be parallel

10. When there is no interaction among the population means in a two-way design, the numerator of the F ratio for interaction in the data from an experiment expected to be:
a. About the same as MSw
b. Zero
c. About the same as the sum of the numerators for the two main effects
d. Dependent on the number of cells in the design

11. Supposes that F ratio you have calculated for a one-way ANOVA is .04. Which of the following can you conclude?
a. A calculation error has probably been made
b. The null hypothesis can be rejected because F<.05
c. There must have been a great deal of within group variability
d. The null hypothesis cannot be rejected
e. No conclusion can be drawn without knowing the degrees of freedom

12. How many cells are there and what is the value of df within for the following ANOVA designs? (assume n=6 subjects per cell)
a. 2X5 (1pt)
b. 6X3 (1pt)
c. 4X4 (1pt)





Part II (computational):
1. The following summary table corresponds to a hypothetical experiment in which subjects must complete a mental task at five levels of difficulty and with three different levels of reward. Much of the table has deliberately been left blank, but all of the missing entries can be found by using the information in the previous sentence and the information in the table:
source SS Df MS F
difficulty 100 ? ? ?
reward 150 ? ? ?
interaction ? ? 5 ?
Within ? 90 ?
Total 1190
a. Fill in all the missing blanks in the table
b. State the null and alternative hypotheses
c. State your conclusions

��

Optional Information:
Level/Year: Sophomore
Subject: Psychology