Continuing this thread, Jon Heron <[email protected]> now asks:
> in attempt to gain a better understanding, I am trying to verify the
> relationship between alpha in the mean-dispersion nbreg model and delta in
> the constant-dispersion nbreg model.
> This works fine in a pair of models containing just a constant however
> breaks down when I add a 2 level factor.
> How is the single value of delta calculated from the two delta_i from the
> two groups? It doesn't appear to be the arithmetic nor geometric mean.
I don't think that one can easily transform between both parameterizations of
-nbreg- post-estimation. They are, in fact, distinct models with differing
likelihood functions that, when optimized, usually yield different
log-likelihoods.
For more details on -nbreg- parameterizations, you can look at
http://www.stata.com/support/faqs/stat/nbreg1.html
and feel free to email me privately (or post to the list) if you have any
remaining questions.
--Bobby
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
