|Date: ||Tue, 4 Jul 2006 05:14:14 -0400|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||Jeff Miller <millerjm@UFL.EDU>|
|Subject: ||Negative binomial hurdle|
|Content-Type: ||text/plain; charset="us-ascii"|
Does anyone know how to run a negative binomial hurdle model?
Since the likelihoods are additive between parts, I would think it to be
just a logit for the zeros and a negative binomial for the positives;
However, these results are diverging from those of other software and really
aren't that sensible.
I'm trying it in NLMIXED.
Has anyone tried this?