The pair of plots below shows two artificial data sets, A and B.

a. Which data set will have the larger mean absolute pairwise difference?
A. Larger for A than for B
B. Larger for B than for A
C. Same size for both
D. Can€™t tell
b. Suppose your null hypothesis is that there is no association between response value and group number. You use simulation to fi nd a p-value for each data set based on its mean absolute pairwise difference. Which data set will have the larger p-value?
A. Larger for A than for B
B. Larger for B than for A
C. Same size for both
D. Can€™t tell
c. Suppose instead you use the F-statistic and theory-based test. Which data set will have the larger mean square for groups (between groups)?
A. Larger for A than for B
B. Larger for B than for A
C. Same size for both
D. Can€™t tell
d. Which data set will have the larger mean-square error (within groups)?
A. Larger for A than for B
B. Larger for B than for A
C. Same size for both
D. Can€™t tell
e. Based on the F-statistic, which data set will have the larger p-value?

A. Larger for A than for B
B. Larger for B than for A
C. Same size for both
D. Can€™t tell

Transcribed Image Text:

0. 6. 10 A 4 6. 8. 10 B 2. sdno1) sdno1)