I am trying to perform a repeated measures ANOVA using the GLM.
I have a couple of within-subject variables (confidence in different abilities before and after a training program), and a between-subject variable (previous experience in the field, a binary yes/no variable). I am trying to test weather having previous experience effected elevation in confidence in abilities following the training, in other words - did more experienced trainees acquired les from the training (an interaction effect between the within- and between- variables).
I found I get different results when using the GLM procedure separately for each of the abilities (the interactions are not significant), and when combining them in a single GLM procedure (the p value is lower). Why is that happening? What is the difference between the two ways?
------------------------------
Roi Eliashar
------------------------------
#SPSSStatistics