Statalist


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

Re: st: Sequential regression run for imputed data


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Sequential regression run for imputed data
Date   Thu, 30 Apr 2009 08:03:14 +0000 (GMT)

--- On Wed, 29/4/09, Frank Gallo wrote:
> I performed an ice run (m=5). My desire is to perform a
> sequential regression analysis (hireg) so that I can 
> document not only the change in R2 after adding a
> variable but also the change in coefficients. However,
> using the hireg option with mim is not permissible. Is
> my only option to run hireg on a single imputed data set?
> However, this would eliminate the variability in the
> imputed data? I would greatly appreciate any thoughts and
> suggestions. Thank you.

Unfortunately, the fact that -mim- can't do this is not a
programming problem, but a statistical problem: Multiple
Imputation is all about inference on the level of single
parameters in isolation, it doesn't work easily with
statistics on the model level like R2 or tests of 
multiple paramters together. Some work has been done in
this area, but my impression is that this literature
has not yet fully matured. If you still want to do 
something like -hireg- you could try to estimate the
separate models and compare them using the methods
discussed in this presentation by Rose Medeiros given
at the last 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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index