I've used Hayes' Process Model 3 Moderated-Moderation to look at the effects of gender (M,F) and learning context (University or college) at levels of Math Efficacy (-1,0,+1) on Math GPA.

So now ive plotted the data for visualizing conditional effects. But Im also interested in the plot for gender and learning context on Math GPA without levels of Math Efficacy (eg with the 3 lines for ME colapsed).

Would that be interpretable? If so, is it like having covaried Math Efficacy? How would one talk about it?

I've looked. The pattern is similar to that using the raw data, but cleaner.