Multilevel hurdle model

[RonD McDowell]

I wondered if anyone had experience fitting multilevel hurdle models in STATA? I see from Long and Freese's text for non-clustered data they recommend fitting the logit and truncated poisson/neg binomial models separately, then using suest to get the correct SE. However, the suest command does not work after fitting the separate equivalent multilevel models.

[Jeph Herrin]

I did this using MCMC, see

http://www.statalist.org/forums/foru...random-effects

for the code.

hth,
Jeph

[RonD McDowell]

Thanks for that link, that's great!