Prashanth Mahagaonkar <[email protected]>:
Use -ivreg2- with the -first- option to run a linear probability model
and report that output together with your MLE -ivprobit-. There is no
paper describing the equivalent tests for probit models, to my
knowledge, but the quality of the instruments is about first-stage
results (from a linear model) not the second stage.
On Wed, May 21, 2008 at 11:13 AM, Prashanth Mahagaonkar
<[email protected]> wrote:
> hi
> I need to perform Anderson-Rubin Wald test of weak instruments,
> Kleibergen-Paap test of underidentification, Stock-Yogo weak ID test after
> using IVPROBIT. I understand that the two-step method (based on the minimum
> chi2 distribution) allows for one such test, but I prefer using the simple
> ivprobit (which follows a MLE method). can anybody help me regarding this?
*
* 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/
