[Clyde Schechter]

Seeing your data example, I really don't understand what you are trying to do. It is clear that personid is not at all nested within Region. Sampling personids within Regions is the only kind of stratified sampling I can wrap my mind around. But since the same personid can appear in multiple regions, that can't be right. So I can't figure out what you mean.

What would a solution to your problem look like?

[Jefferson Pereira]

Clyde Schechter ,
One of my model hypotheses is that there is mobility of individuals between regions. So, if I first create a subsample with just these individuals who moved and then do a stratified sample of individuals within the regions, considering the individuals who did not perform mobility, would it be possible? Does it make sense?
At the end of the sample I would have individuals who move between regions, plus a random and stratified percentage of those who do not move within regions.,

[Clyde Schechter]

So, if I first create a subsample with just these individuals who moved and then do a stratified sample of individuals within the regions, considering the individuals who did not perform mobility, would it be possible? Does it make sense?

I don't want to say unequivocally that it makes no sense, because I may be missing something. But I am still unable to make sense of it myself. I hope somebody else will grasp your meaning and help.

[Jefferson Pereira]

Many thanks for your time, Clyde Schechter !
You've been very attentive to my issue, that's make me very grateful!