Dear all,

I am running a multilevel model with over 40.000 observations, but when I use the command 'predict if e(sample)' I lose roughly 37.000 of them. Anyone that can tell me how that is possible?

Many thanks in advance,

Kris