Hi Jean --
Thanks for the reply. I have a number of data sets that require more than
30 points to obtain estimates that seem to converge using gllamm with the
adapt option, so I'm a bit concerned with the 30 limit. Do you know if
xtlogit does a better job of producing estimates than gllamm with fewer
quadrature points specified? If not, hopefully the 30 limited will be
bumped up quite a bit sometime in the future.
"Bryan W. Griffin" <[email protected]> asked:
> I don't have Stata 9. I understand that in Stata 9 xtlogit
> estimation has been changed to adaptive quadrature as the
> default. In Stata 8 maximum number of quadrature points allowed was
> 30. Has this limit been changed, and if yes, what is the new maximum
> number of quadrature points allowed?
It is still 30. The adaptive quadrature routine is much more flexible
in fitting the likelihood than standard Gauss-Hermite quadrature is
for the same number of quadrature points. See [XT] quadchk for
some examples.
--Jean Marie
[email protected]
___________________________________________________________________
Bryan W. Griffin
Curriculum, Foundations, & Reading
P.O. Box 8144
Georgia Southern University
Statesboro, GA 30460-8144
Phone: 912-681-0488
E-Mail: [email protected]
WWW: http://coe.georgiasouthern.edu/foundations/bwgriffin/
*
* 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/
