Bèiendé Niú posted on Wednesday, March 29, 2017 - 1:51 am
Dear Professors Muthén,
When I model a direct cross-level effect (yb on x1 (aa)), the effect is positive and highly significant. When I additionally enter a second cross-level effect (yb on x2 (bb)), aa becomes insignificant.
My model is as follows:
CLUSTER = clus; BETWEEN = x1 x2;
ANALYSIS: TYPE = TWOLEVEL;
MODEL: %within% yw BY a b c d e; %between% yb BY a b c d e; yb ON x1 (aa); yb ON x2 (bb);
Is there something wrong in this model or how can the changing effect of aa be explained?