Austin Frakt--
You should provide the reference for Green--do you mean Greene:
http://www.stata.com/bookstore/ea.html
and if so, what page number?
The model you described sounded like one where you have a set of
proportions y_ij that sum to 1 for each individual j, as in a y that
measures proportion of total assets in stocks, bonds, etc. Then I
think you mean to estimate a fractional logit as in Papke and
Wooldridge (1996), see
http://www.stata.com/support/faqs/stat/logit.html
for the Stata command syntax, but with a seemingly unrelated
regression structure imposed on top, to account for the adding-up
constraint, and I thought Brian Poi's article and programs described
at
http://www.stata.com/statalist/archive/2004-01/msg00816.html
might be relevant.
On 3/8/07, [email protected] <[email protected]> wrote:
Thanks to Nick Cox and Austin Nichols (in private communication) for
suggestions on my post about -clogit-like estimation on proportions
data. It isn't clear to me that any of the Stata approaches are the
same as Green's / LIMDEPs. I don't know if I have the expertise to
confirm that or to appreciate the differences but I will look into it.
--Austin Frakt
*
* 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/
