Thanks so much Alfonso, and I hope the convention went well!

Your answer helped me clarify the question I had in mind: I do want a single contrast that would identify all of the between-group difference and the association with the emotion covariate, as you suggested. There's one thing that is still not clear to me: the contrast that you suggested [1 -1 0 0 0 0 0 0; 0 1 -1 0 0 0 0 0; 0 0 0 1 0 0 0 0] is supposed to show significant results for all the significant results of [1 -1 0 0 0 0 0 0; 1 -1 0 0 0 0 0] AND all the significant results of [0 0 0 1 0 0 0 0], right? To simplify, if I had 2 significant results for the two-line contrast, and 3 significant results for the one-line  contrast, than the three-line contrast should give me all 5 (with the same p-values)?  I might be doing something wrong, but this is not the results that I'm getting (despite using an uncorrected p).


Thanks a lot for all of your help!!

L.