Re: st: which -cmp- option to use for poisson model with count data?

["Laura R." <laura.roh@googlemail.com>]

Dear Nick,

I really appreciate your help.

(a) information
dependent variable:
- experts
- in this sample: 0 to 5
- 0 should be in the analysis (not "missing")
- in other samples, it can be more than 5, but naturally not less than 0
- one independent variable can be endogenous, so I thought about
estimating this variable with a probit, then the experts-regression,
in a framework like -cmp- where the error terms can be correlated and
which is not restricted to a contiuous variable in the main regression
(like Heckman selection would be)

(b) understanding
"> "Unless the counts are high, count data can't be realistically
modeled as the outcome of a single underlying process consisting of a
linear functional plus a normally distributed error." "
Yes, I certainly got that.

"> That was his advice about using -cmp-. He's the author and an
expert. If you want to go against his advice, that's your call, but in
the only example you have given, your counts have a maximum of 5."
Of course, he is the expert, and no, I don't want to go against his
advice. That's why I am looking for another solution on how to analyse
my data correctly. So far, options might be using -gllamm-, -ssm-,
binomial, maybe ordered probit within -cmp-.

"Whether your count data can be treated as ordered probit is something
on which experts have different tastes and judgements. Counts that can
go 0,...,5 could be treated as graded variables 0 < 1 < 2 < 3 < 4 < 5.
I can't comment on the example you refer to, as I have not studied
it."
O.k.

>
"> On terminology: I wouldn't describe a counted variable as a
> categorical variable, although counted variables do certainly appear
> in categorical data analysis texts."
Sure, you are right.

Best regards
Laura
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/