MANOVA Discussion

1.  a) How are ANOVA and MANOVA alike?  b) How are they different?

2.  What should you keep in mind when selecting DVs for a MANOVA?

3.  Why is it desirable to include more than one DV when comparing treatments or groups?

4.  How are the multiple DVs dealt with in a multivariate situation?

5.  Why is it said that a MANOVA may sometimes be more powerful than separate ANOVAs?

6.  Recall the meaning of experiment-wise or overall Type I error. Suppose you had a single IV with three levels and three DVs. How would MANOVA greatly reduce the overall Type I error rate compared to conducting three univariate ANOVAs?

7.  What is the most commonly used test statistic for MANOVA?

8.  The first hypothesis test in MANOVA tests whether all groups are equal on the combination of DVs. If this multivariate hypothesis is rejected, what is the next step in the analysis?