st: Average marginal effects for a multiply imputed compley survey data set

[Elke.Luedemann@campus.lmu.de]

Hello,


I am using a multiply imputed complex survey dataset to fit a multinomial
logit model in Stata 11. The dependent variable has three possible
outcomes, and I have both continuous and discrete independent variables.

I calculate individual average marginal effects after mlogit for each
of the imputed data sets separately using the following code:

svyet psuid [pw=mywgt], strata(stratavar)
forvalues j=1/3{
quietly: svy: mlogit depvar var1 i.var2 var3 i.var4
margins, vce(unconditional) dydx(*) predict(outcome(`j')) post
eststo modela_`j'
}


My question is: how can the estimates of the average marginal effects for
several continuous and/or discrete independent variables (as well as their
standard errors) be combined across the different imputed datasets?

Similar questions have been asked to the Statalist before, but I was
wondering whether either the new Stata 11 module MI (i.e., the mi estimate
command) or the user-written commands mim/mim2 are capable of estimating
these statistics.

I have tried to use the "mim2, category(fit):" command, but this does not
seem to work because "statsby" doesn't allow pweights.

(Remark: I am not interested in estimating odds ratios, which would be
simpler to combine.)

Any suggestions would be appreciated.


Best,

Elke




*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/