Can anyone suggest why performing negative binomial regression and GLM with log as link function and negative binomial error distribution I obtain very different results on the same data? The actual difference is about the error of the coefficient (much smaller in GLM).
I can't say why, but I've noticed I get different results using, for example, xtnbreg and xtgee with family(nb), even though xtnbreg is supposed to be a "convenience" command to do the same thing. I'm using 11.1, by the way.
I've found the answer to my question by myself.
In negative binomial regression STATA estimates the parameter alpha, that is simply the inverse of the k parameter of negative binomial distribution, well known by parasitologists. The k parameter is inversely related to aggregation and can be estimated with the additional module nbfit with STATA. The K parameter requested by GLM is exactely the alpha parameter of negative bin regression (and it is equal to 1/k); its default value is 1, so if alpha is not 1, the results will be different.
Thank you; that explains some of the anomalies. I'll look at the assumptions behind the xtnbreg and xtgee, though I thought they were comparable.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software