[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: -nbreg- not working properly?

From   Peter Siminski <>
Subject   Re: st: -nbreg- not working properly?
Date   Fri, 10 Aug 2007 09:18:35 +1000

OK thanks David.


>From   "David M. Drukker" <>
Subject   Re: st: -nbreg- not working properly?
Date   Wed, 8 Aug 2007 08:47:45 -0500 (CDT)

Peter Siminski <> asked why -nbreg- (with
defaults) does not produce the same results as -glm- with family(nbinomial).

The manual entry for -glm- (note 4 on page 499 of [R] A-H) notes that

``Although -glm- can be used to fit [negative binomial models],
using Stata's maximum likelihood -nbreg- is probably better. In the
GLM approach, you specify -family( nbinomial #_k) and then search
for the #_k that results in the deviance-based dispersion being 1.
-nbreg-, on the other hand, finds the maximum likelihood estimate of
#_k and reports a confidence interval for it.

I hope that this helps.


Peter Siminski
PhD Student
Social Policy Research Centre (SPRC)
Building G2
University of New South Wales
Kensington NSW 2052
Ph: 0425223257
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index