Marginals
On the last slide of the Mixed Designs handout, it says that we can check normality of marginals for RM if df is low, and check if Fmax for the marginals is less than 10 in RM for homogeneity of variance. What does this mean? What are marginals in repeated measures and what do they have to do with normality or homogeneity of variance?
Also, how do we check an Fmax of less than 10 for marginals in repeated measures?

1 Comments:
At 5:55 AM,
Mari Clements said…
Marginals are the means across factors. So, if you are looking at RG factor A, you would average across times 1 and 2.
Similarly, if you are looking at Repeated factor B of time, you would average across all levels of RG factor A.
These variables would be computed in the normal manner, using transform and compute.
Post a Comment
<< Home