Dear Stata users,
I am working with expenditure data with around 15% missing data, which I want to impute. I used multiple chain method to impute data, my command is here:
After this I want to re-estimate my original multi-level model with the imputed dataset. I am using the following command:
It works and provides me a results output that makes sense, but it does not show the coefficients as risk ratios, despite my "irr" specification in the end. Any idea on why this might be happening? And otherwise, is it fine to proceed with the "cmdok" command in my case?
Thanks a lot!
I am working with expenditure data with around 15% missing data, which I want to impute. I used multiple chain method to impute data, my command is here:
Code:
mi register imputed totalexp_month_C mi impute chained (nbreg) totalexp_month_C = AforA age education incomegen female married age education assetsum , add(10) rseed (53421) savetrace(trace1,replace) force
Code:
mi estimate, cmdok: menbreg totalexp_month_C i.period AforA female_c married_c age_c education_c || SiteCode:, irr
Thanks a lot!