IV: State (NJ, HI, CA, FL) (reference =CA) Moderator: # of hours exercising DV: BMI
In this case, the IV is categorical and I have 3 dummy variables. As such, I created 3 interaction terms. In using hierarchical regression for the moderation, would I put all of the interaction terms into one model? For example would it be...
step 1: exercise step 2: exerciseNJ, exerciseHI, exercise*FL
If this is the case, I am unsure of how to interpret the F change since it includes all of the interaction terms. If it is significant, what does that mean? In this example, I found F change to be significant and had some significant betas.
However, when I ran the analysis separately for each dummy variable, nothing was significant. Namely, I ran it this way too...
step 1: exercise step 2: exercise*NJ
I ran it this way separately for each of the dummy variables since I could theoretically understand what the F change means. In this case, there were no sig F change or betas.
Can you please advise on how to approach this moderation with dummy variables?