Student 09 posted on Friday, March 30, 2012 - 6:08 am
I would like to model an interaction of two continuous between-group variables xm_1 and xm_2.
Using the define-command, I create a mulitplicative interaction term and insert it on the between-level.
However, both xm_1 and xm_2 are aggregate variables of variables x_1 and x_2 measured at the within-level.
My question is whether I need to model the interaction of x_1 and x_2 on the within-level as well in order to get an appropriate parameter and S.E. for the between-group interaction my theoretical interest focuses on - what would you suggest? Many thanks for your answer!