Saturday, June 26, 2021

Interpretation of significances with moderation effect

I am running a panel regression with the following problem:

In the first regression X has a significant impact on Y.
Then I introduce the interaction term XZ to test for a moderation. X becomes insignificant in this case despite XZ being insignificant.

How do I interpret this? It might be obvious but I dont really get it. How does the interaction term cause X to become insignificant despite XZ being insignificant itself?

No comments:

Post a Comment