Hey all, I have a time-invariant independent variable and a continuous dependent variable. My dataset is unbalanced panel data. I already know that I cannot use the fixed effects model as my time-invariant variable would be removed. Consequently, I decided to employ a Hausman-Taylor model in order to still control for fixed effects as I assume that they are crucial for my analysis. However, I actually wanted to analyze mediation and moderation effects as well. Can I do that with a Hausman-Taylor model? If yes, how and if no, what alternatives do I have? Pooled OLS? GLS? Thanks for your help!!