Re: st: Adjustment to likehood value due to dependence of data observations

[Maarten Buis <maartenlbuis@gmail.com>]

On Thu, Sep 22, 2011 at 5:08 PM, Abdul Q Memon wrote:
> 1. As GEE takes a long time to run (quarter of million observations) so I
> decided to run GLM which is quite fast and compared the likelihood values
> for model selection.

The comment made by Nick still applies. You should not expect that to
work nor should you expect there to be a "correction term". If the
model is wrong, than you should estimate a different model. If that is
inconvenient, e.g. because it takes long, than you'll just have to
learn to live with that inconvenience.

> 2. After running the GEE which is suitable for model (Panel data) it seems
> there is still some trend in residuals (Plot of residuals and fitted
> values possibly becuase of homoscadicity). Am i right in using robust
> after gee command??

Again, the comment made by Nick still applies. You seem to have
misspecified the time part of your model. The solution is to correct
that. Typically you would do that with the -corr()- option in -xtgee-.

Hope this helps,
Maarten

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


http://www.maartenbuis.nl
--------------------------
*
*   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/