Chaper 7

Question: When I estimate the APIM using Multilevel Modeling, do I need to run the analysis using partner outcome at the dependent variable. Another version of this question, how can two equations be estimated in one equation?

Answer: When you use Multilevel Modeling because you have a pairwise data set, both the person's and partner's data are treated as outcomes in the one analysis. Also, when you use Multilevel Modeling with indistinguishable dyads, there are two equations, but because members are indistinguishable, you have constrained the actor and partner effects to be equal across equations.