Hi all,

I'm working with a three-level cross-classified model with repeated measures (i.e. multiple observations over time per individual, nested in states, nested in two groups) using the -mixed command (using Stata 16.0). While I have dealt with multi-level models in the past, I am wondering whether it is possible to specify a finite population correction in such models. I have attempted to specify finite population corrections using the -svyset command but whenever I run the model I am shown the following error:

mixed is not supported by svy with vce(linearized)

Any advice on how finite population can be specified in these kind of (cross-classified) models?

Thanks in advance!


Best
Jon