Dear Statalist,
we want to estimate a three-level random intercept model (only random
coefficients for the constant on level 2 and level 3). Dependent variable
is a count, negative binomial is most appropriate model because of its
correction for overdispersion.
So far, we have a working command to estimate a poisson model within Stata:
gllamm emiprown verb ageleg amotiv mot_gel, i(id intno) family(poisson)
link(log) adapt
what would be the command for a negative binomial, or more important where
can we quote/look up the syntax for this specific problem. Specs: so far
only fixed effects on the lowest level are included, but we plan to add
some coeffs on lev2 and lev3. mot_gel is an interactin term (mot_gel=
ageleg * amotiv); id's are nested within intno's (no crossclassification).
thanks,
Christian
======================================================
Christian Hunkler
student research assistant
"Aspirations, Reference Groups and Educational Decisions"
Project Program "Rationality Concepts, Decision Making
and Economic Modeling" (SFB504)
University of Mannheim, Germany
Phone: +49(0)621-181-3433 (O)
+49(0)6202-702516 (H)
Fax: +49(0)621-181-3451
E-Mail: [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/
