TRUE OR FALSE

19. The two-sample t-test assumes heteroscedasticity.

20. In a regression study, you set the values of the confounding variable, and you control or randomize all of the possible independent variables.

21. In a 2-way ANOVA, when the interaction term is significant, it is advised to look at each factor separately, using a 1 way-ANOVA.

22. In a repeated measures ANOVA, one of main effects is usually uninteresting and the test of its null hypothesis may not be reported.

23. ANOVA depends on only one computation for degree of freedom to obtain the F- value.

24. Because of the paired design of the data, the null hypothesis of a paired t-test is usually expressed in terms of the mean difference.