Advice on multilevel modelling

[Ujjwal]

Dear all, I am planning to conduct a multilevel analysis - where individuals are in urban-rural groups which in turn are within countries. I have a dataset which contains individual level observations across countries but, individual ids are not repeated, i.e., no within person variation. My confusion here, can I conduct a three level (individual, urban-rural, country) analysis in this case? If not, is it going to be a two-level model (urban-rural and country)? Please suggest. Thank you.

[Clyde Schechter]

I do not recommend having an urban-rural level, as there are only two groups at that level, and they are, in fact, exhaustive of the real-world possibilities on that dimension. There is no individual level in this model as you have only one observation per individual--or, otherwise put, the "individual level" is just the residual level. So I would do this as:

multi_level_model_command outcome_var explanatory_var(s) i.urban_rural || country: