Hi again and thank you for your thoughtful reply.
Yes, I have a version where each group is designated as 1-4 in the group column, respectively. The analysis is much more powerful ( as you would assume) with the groups designated as 1s, with fixed or random effects.
I was wondering why the .gfeat analysis "works" at the third level with fixed effects, but not random. You are correct though. This approach would definitely confuse the reader and reviewer.
Thanks again for your support and insight.
|