Re: st: table & chi-square from multiply imputed survey data (-ice-, -mim-) [repost]

["Michael I. Lichter" <mlichter@buffalo.edu>]

Thank you Stas and Maarten for the thoughtful and useful responses. 
Thanks in particular to Stas for the pointer to his library, and Maarten 
for the pointer to Rose Medeiros' presentation (though I can't say I 
know what to do with it). I also appreciate the endorsement of -mlogit- 
(in some form). (I would try Stas' suggested bootstrap, mlogit, & repeat 
method if I knew how to implement it.)

I find it interesting that multiple imputation as it exists now does not 
deal with seemingly-simple problems like producing tables and chi-square 
statistics, but works well for apparently-complex problems like 
producing coefficients and standard errors for multivariate models. 
Perhaps there's something here for an ambitious grad student ... :)

FWIW, when I try mlogit (or svy mlogit) with mim -- e.g., "xi: mim: 
mlogit B i.A" I get a "conformability error". I assume that the error is 
coming from mim rather than mlogit because I can run mlogit without mim 
on the whole dataset and on each replicate individually without 
problems. This could easily be a result of the particular variables I'm 
looking at, but it would be nice to know how to avoid "conformability 
error" in mim.

Thanks again.

Michael
*
*   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/