I have a SEM model with Three exogenous and eight endogenous variables. some of the effect size becomes negative. is that showing that exoogenous variable in fact is a reduced variable that reduced the power of prediction model ? how can I interpret it?
Im appreciat if some one have idea.