100%!
My additional recommendation for studies like this would be to also do it with cluster robust errors and worry a lot if they're giving you a much more conservative result than HLM. Different versions of this problem can crop up quite easily.
My additional recommendation for studies like this would be to also do it with cluster robust errors and worry a lot if they're giving you a much more conservative result than HLM. Different versions of this problem can crop up quite easily.
Reposted from
Mattan S. Ben-Shachar 🎗️🇮🇱🇺🇦
#statsmeme
Comments