Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st:poisson vs zip vs nbreg vs zinb


From   "Scott Merryman" <smerryman@kc.rr.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st:poisson vs zip vs nbreg vs zinb
Date   Wed, 13 Nov 2002 20:04:25 -0600

----- Original Message -----
From: "C Johnson" <cjj1@gpu.srv.ualberta.ca>
To: <statalist@hsphsun2.harvard.edu>
Sent: Wednesday, November 13, 2002 12:06 PM
Subject: st:poisson vs zip vs nbreg vs zinb
<snip>
>
> Those summary stats suggested that ZINB would be most appropriate.
> However, after fitting the ZINB model a number of the "inflate"
> coefficients and SEs are very large, but the Vuong test suggests ZINB over
> NBREG.  The faq http://www.stata.com/support/faqs/stat/nbreg.html notes
> similar problems with coefficients and SEs, but provides little
> explanation.  Both ZIP and NBREG provide results that are much more
> well-behaved.  Given my data and the fact that ZINB does not fit well, it
> would appear that ZIP is the best choice.
>
> Any suggestions on why the coefficients for the ZINB model exploded?  Any
> advice/confirmation for selection of a count model is also appreciated.
>
In the choice between ZIP and NBREG, is there a theoretical justification for
having an additional data generating process for the zeros?  If there is no
compelling case for having two DGP you might go with the negative binomial model
because it is more parsimonious.

Scott




*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index