I am running some mixed effects models on my data (N = 6,000 approximately). I am try to figue out how to report effect size estimates.
The first model is a two-level main effects model, for which I have calculated pseudo R-square as a measure of effect size. The second model is more complicated and has 5 cross-level interaction terms (magnitude X 5 personality variables separately) along with the same controls and main effects. What measure of effect size can I report in this case and how can it be calculated? Can you please guide me towards relevant literature etc.?
Would you suggest simpler models with just 1 interaction term and then reporting an R2 estimate? Or is it possible to run an anova between models to test the significance?
I use the LME4 package in R, for my data analyses.