Graphing Interaction Effect

[Martin Gabriel]

I have a linear regression model for which I want to account for an interaction effect.
The basic model is

regress fp g u i

With the interaction effects:

regress fp c.g##schanc c.i##schanc c.u##schanc

What I'm looking for, is a way to graph the effect of schanc on the basic model. My question is, whether it is possible to create a graph for the whole model, akin to one looking only at one variable, e.g.:

regress fp c.u##schanc
margins, at(u=(1.3(.1)9.1) schanc=(0 1))
marginsplot, recast(line) addplot(scatter fp u)

Attached is the respective output.

I hope this isn't to elementary of a question, I've looked for a similar question via the search function but didn't find anything that answered mine. Thank you.

[Jesse Tielens]

Hi Martin,

I think what you're looking for is essentially overlaying two models in a marginsplot:
-One with the interaction effect
-A second model without the interaction effect.

Essentially, what you need to do is run those two regression models -> Save the results -> Graph them in a single marginsplot.

There's an SSC plugin for stata which allows you to do this. It's discussed in this thread: https://www.statalist.org/forums/for...wo-marginsplot

Does that help you?

[Martin Gabriel]

Hi Jesse,
my question was probably wrong, but your answer solved what i was asking for anyways - overlaying two regression models is exactly what I needed! The plugin helped.
Thanks very much!

[Jesse Tielens]

That's nice to hear