1. An advantage of the permutation test over the t test for a difference in means is that
A. it gives accurate P-values even when the distributions are not normal.
B. it tests a more general hypothesis.
C. Both choices are true.

2. Permutation tests cannot be used for
A. tests when we believe the two populations differ under the null hypothesis.
B. matched pairs designs where we believe only random differences occur.
C. testing relationships where the null hypothesis says no relationship exists.

3. Which is more accurate?
A. a t distribution confidence interval
B. a bootstrap BCa interval
C. a bootstrap percentile interval