Hi everybody
I have panel data for the entire population. I run my regression and use the margins-command afterward. Should I use the vce(unconditional)-option in this case to get the correct standard errors? When looking into the manual, it seems like vce(unconditional) is only for survey data, but I am not entirely sure as I also use cluster(id) and [aw = cem_weights].
0 Response to Margins and standard errors with population data - vce(unconditional) or vce(delta)?
Post a Comment