I've calculated Prevalence Ratios (IRR through Poisson with robust error variance) on Stata 14.2 for a weighted dataset with 25,000 subjects and 10 variables (all variables binary or ordinal), with commands like:
svy linearized : poisson outcome exposure covariates, irr
I now want to do multiple imputation to assess the robustness of findings against missing data, which most variables in the dataset have. After imputing variables with missing data, I used the basic command:
mi estimate : poisson outcome exposure covariates
to get a coefficient output which I can convert to an effect size with a 'display exp' command. But when I specify weighted data with 'svy linearized', I get the error message "subcommand mi svy is unrecognized". I've tried other combinations of wording in the command, but none work. Can Stata give me multiple imputation outputs on my weighted data?
Thanks very much for any help. My Stata skills are limited!
Related Posts with Multiple imputation with weighted data
Machine Learning setupHow can I break down my loaded dataset into training set and test set, and develop random forest on …
Replacing missing variable with other observations that satisfy certain conditionsHi all, I am working on a cross-country dataset. The dataset is created when I merge the bilateral t…
Results InterpretationHi everyone, Can anyone help me interpret these results, specifically the F-test below. What does t…
Collapse different columns differentlyIn a panel database of companies (bisnode), the revenue information is not always for the full year.…
Why do i have large z test statistics when i ran translog modelFirst i ran frontier model and then i did translog method here i have obtained large wald chisquare …
Subscribe to:
Post Comments (Atom)
0 Response to Multiple imputation with weighted data
Post a Comment