Hello everyone,
I am a beginner with Stata so please bear with me.

I'm having a hard time understanding the effects of my interaction term on my main explanatory variables. For context, one of my hypotheses assumes the relationship between social media use (IV) and an individual's inclination to participate in non-traditional political participation (DV created by indexing boycott, protest, and petition) will be stronger for citizens situated on the left of the political spectrum. To determine this, I created an interaction term combining Social Media and Ideology by running Social*Ideology.

1. When I ran regress with my interaction term, my interaction term becomes statistically significant, the original Social Media variable becomes statistically insignificant (and the coefficient becomes negative) and the original Ideology variable stays statistically significant. How can I interpret this? Am I even approaching this the right way?

2. When I add the interaction term, I now fail the Omitted Variable Test when running estat ovtest. Is there any way to remedy this, or is this something that's expected to happen with an interaction term?


The first model is without the interaction term, while the second model is with the interaction term.
Array




Thank you so much in advance, any help is extremely appreciated!