Dear all,
I have a model with several IVs, DVs and moderating variables. In one of the relationships, I find a significant path coefficients (beta) but the latent variable correlation is low and non-significant. Other than multicollinearity, what could explain these contradicting results?
Thanks,
Shamel