Re: Assessing Negative Binomial Model Performance

["Donald Catanzaro, PhD" <dgcatanzaro@gmail.com>]

Hi All,

I have used the MASS library to fit a negative binomial generalized 
linear model and am now trying to assess its' performance by using first 
a call of anova.negbin(ly.nb) and use sequential likelihood ratio test 
to look at the coefficients.  Someone advised that I use the Wald test 
as well on these coefficients
I have two questions:

Is this appropriate ?  If so,  how would I implement it ? 

I searched around in the list and did not find a lot of discussion about 
the Wald test with respect to negative binomial models, just a few 
descriptions of how the t-ratios reported by glm.nb are not 
t-distributed and to use the anova dropping technique (which I already 
am doing).
Thanks in advance !

--

-Don 

Don Catanzaro, PhD                  Landscape Ecologist
dgcatanzaro@gmail.com               16144 Sigmond Lane
479-751-3616                        Lowell, AR 72745