Hi all,
Following Austin's excellent feedback, I am trying to use an
instrumented probit model instead of ivreg2 to examine the
determinants of happiness levels. I have Stata Version 8.2.
Can someone please clarify the following?
1. The command ivprob works well with binary probit estimation. But
are there any tests like overid for ivreg2 that I can use with ivprob
that can tell me whether my model is overidentfied?
2. While I can, at a push, reduce my dependent variable to binary
form, ideally I would like to have at least 3 ordered options. For
which oprobit is the best. But is it possible to instrument with
oprobit, and if so are there tests for overidentification that one may
use?
Thanks in advance,
Ramani
*
* 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/
