Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Multilevel models using multiple imputed data


From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Multilevel models using multiple imputed data
Date   Tue, 21 Apr 2009 07:19:58 +0000 (GMT)

--- On Mon, 20/4/09, Nailing Xia wrote:
> I am estimating a multilevel model using -xtmixed-, and the
> dataset I use is from multiple imputation procedure using 
> -ice-. I use -mim- for estimation, and the output does not
> give the log restricted-likelihood ratio (which is
> typically shown in -xtmixed- output when not using -mim-).
> Actually, I can find the log restricted-likelihood ratio for
> each imputation (i.e., I used 10 imputations, and hence 10
> log restricted-likelihood ratios), but how can I obtain an
> overall log restricted-likelihood ratio? Is there any issues
> for using the average of the log restricted-likelihood ratio
> of each imputation?

Yes, the thing to keep in mind with multiple imputation is 
that it is primarily a way of estimating individual parameters
in isolation, so any joint test involving multiple parameters
is going to be hard. Work has been done in this area, and some
of it has been implemented in Stata, but my impression is that
this is still an area of active research. For more on this 
see the presentation by Rose Medeiros gave at the 2008 Fall
North American Stata Users' Group meeting :
http://ideas.repec.org/p/boc/fsug08/11.html

Hope this helps,
Maarten

-----------------------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://home.fsw.vu.nl/m.buis/
-----------------------------------------





      

*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index