Moderation in Fixed Effects Analysis

[Giulia Til]

Hi to everyone,

The main technique I found to measure moderation is inserting some interaction terms (Moderator*Independent variable) in a multiple linear regression. However, the moderators I'm studying are PIL and other few variables that are fixed and equal for all my sample.
Therefore, I cannot insert these fixed variables as interaction terms in a multiple linear regression in Stata, because they are omitted for collinearity.
Do you know a method to measure the moderating effect of fixed variables?

I hope you can help me!
Thank you,
Giulia

[Phil Bromiley]

You can include such variables as long as you don't try to include the main effects. That is use x1#x2 not x1##x2.