Statalist


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

Re: st: xtmepoisson


From   Stas Kolenikov <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: xtmepoisson
Date   Tue, 8 Dec 2009 11:36:27 -0600

On Mon, Dec 7, 2009 at 4:55 PM, David Greenberg <dg4@nyu.edu> wrote:
> Is there a way of adapting the xtmepoisson in version 11 of Stata
> to accommodate zero-inflated mixed Poisson models?

It's not adapting, it is re-writing. How exactly do you expect the
likelihood for such model to look like? -xtmepoisson- has a numeric
integration of the random effects for the likelihood, and
zero-inflation means putting a Poisson model into a mixture model with
probability of the identically zero and Poisson components. Depending
on whether you believe there should or should not be a random effect
in the mixture component selection equation, you get two different
ways of zero-inflating -xtmepoisson-, and if you have more than two
levels, it becomes a total mess. You'd probably be able to estimate
whatever mess you want to get into with -gllamm-, with all the usual
wrapping the head around its syntax and data management.

-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   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