Hi everyone,

I'm using logistic regression after multiple imputations (ice). In my model, there is an interaction term. I want to use margins function, but it didn't work. I spent a lot of time googling and found a solution on the UCLA page. https://stats.idre.ucla.edu/stata/fa...-imputed-data/

I followed their instruction and it worked. But the margins function only provide me coefficients, is there any way to have odds ratio instead? In addition, I also would like to a likelihood-ratio chi-square (eg: lrtest M1 M2) but it didn't work. I googled, and I saw some people say it is not supported using mi estimate. Is there any other way to run the test? Below are my codes, thank you so much for your help.

program mimargins, eclass properties(mi)
version 14
logit druguse i.loss##i.transgen i.income_recode i.race i.educa_bi i.sex i.orien i.impage3_80 c.chronic i. relation2,or
margins i.loss##i.transgen,asbalanced post
end

mi estimate, or: logit druguse i.loss##i.transgen i.income_recode i.race i.educa_bi i.sex i.orien i.impage3_80 c.chronic i. relation2
estimates store M2

mi estimate,cmdok: mimargins 1