Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: multiple imputation using STATA


From   Todd Wagner <twagner@stanford.edu>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: multiple imputation using STATA
Date   Tue, 03 Dec 2002 09:42:41 -0800

Beth,

I have used Amelia created by Gary King and am familiar with the programs created by Joe Schafer (Penn State). In my experience, which is a couple years old now, I would guess that this would add a least a week to your analysis. Poisson or neg bin regression should be relatively easy to run with the MIEST command. But, as you point out, the real limitation is that you have to run any specification test on the individual datasets. There is no easy way to summarize these tests; I reported the worst fitting model.

All the best,

Todd




At 07:33 AM 12/3/2002 -0500, you wrote:

Hi all.  I am a fairly new STATA user who is determinedly trying to finish my
dissertation.  I have uploaded the hot deck add on and at this point (ever
changing) plan to do hot deck multiple imputation based upon the advice of
the statistician on my committee.  In trying to think about the ramifications
of using this method, being that I am also a novice to multiple imputation, I
had some questions after reading about the disadvantages.  I am concerned
about the time intensiveness of testing models for each dataset separately
and recombining the model results into one summary model.  In addition to the
time intensiveness, I wondered if any of you out knew more about summary
methods that might be available in STATA?  I have read that summary methods
are generally available in most statistical software packages for linear and
logistic regression models but that work is still in progress to provide
statistically appropriate summarization methods for other models.  Has anyone
out there used these summary methods in STATA that might have some words of
wisdom? I will be using Poisson regression in my analysis and I wondered if
these summary methods were available for this type of regression modelling?

Thanks in advance for any help,
Beth
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index