Hi Stata-listers,
I'm trying to get marginal effects in a probit regression with endogenous
variables and intra-cluster correlation. The options I know of (and their
associated shortcomings for this purpose) are:
ivprobit: can get marginal effects only by using "mfx compute, predict(p)
force", which is not advised by the FAQ
divprob: gives correct standard errors, but does not allow these to be
clustered
qvf: 1) does not appear to produce marginal effects; 2) the suitability of
this routine to address endogeneity (rather than measurement error for
which it was designed) is disputed.
Any ideas?
Thank you!
Vivian
Vivian Hoffmann
Applied Economics and Management
Cornell University
1.607.316.7291
http://www.people.cornell.edu/pages/veh4/
*
* 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/