Guanghui Li <[email protected]> noticed that -svypoisson- produced
different point estimates when comparing the results of a model fit using -if-
and the -subpop()- option:
> I need to estimate a svypossion model and I am only interested in a
> subpopulation.
> I was curious to see how the "subpop" command would affect my results since
> Stata Manual emphasized the importance of using "subpop" instead of "in".
> Therefore, I estimated two models, one with the "subpop" command, the other
> with the "if" command.
> I was expecting to see different standard errors. However, I got different
> coefficients in addition to different variances.
> Based on what I read from the manual, I got the impression that only
> variances would be affected by using the "if" command.
> Could the coefficients be different?
Bryan Sayer <[email protected]> replied that the point estimates should be
the same.
We have looked into this and discovered a small bug. It has been fixed and
should be available in the next ado update.
--Jeff
[email protected]
*
* 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/
