Dear Stata experts,

I'm trying to test several GLM models. I have two independent variables (X and Y). X is binary and Y is continuous. The dependent variable (Z) is binary. I also have two moderators (M and N).

According to my understanding, if I only have one independent variable (X), the models are like this:

Model 1 (Empty model): Z = M + N + control variables
Model 2 (Main effect): Z = X + M + N + control variables
Model 3 (Moderating effect M): Z = X + X*M + control variables
Model 4 (Moderating effect N): Z = X + X*N + control variables
Model 5 (Full model): Z = X + X*M + X*N + control variables

Now since I have two IVs X and Y, and I want to test the moderating effects of M and N on both X and Y. How should I add Y, Y*M, and Y*N in the models? Should I just repeat Model 1 through Model 5 by replacing X with Y? Or should I put X and Y together in the five models?

Thank you very much for helping!