Hello,

I'm wondering what happens when I both use a multilevel model and cluster my standard errors for cross sectional data.

Ordinarily, I understand that these two techniques serve the same purpose, but seeing as clustering is an option in STATA's multilevel modelling, I am now wondering if I should cluster. I'm using the -melogit- command.

I have a fairly simple model of firms nested in countries nested in two global regions.

Thank you