In short,
the first.
If you reject exogeneity, you should do something along the lines of
-ivprobit- to try to get consistent estimates. OTOH, it is always the
case that that even after instrumenting, the endogeneity problem
persists in the model--IV estimates are still biased, and have much
higher variance (i.e. aren't very efficient) than their inconsistent
counterparts...
On 12/19/06, Alejandro Delafuente
<[email protected]> wrote:
Hello statalisters, I am running some maximum likelihood probits with one
endogenous regressor (ivprobit). Stata reports the result of a Wald test of
exogeneity at the bottom and I want to make sure that am giving the right
interpretation: a rejection of the null hypothesis of exogeneity (rho=0) would
mean that the error terms in the structural equation (probit) and the reduced-
form equation for the endogenous variable (instrumented regression) are
correlated and therefore instrumenting the endogenous variable was the
APPROPIATE DECISION. Or does it mean that even after instrumenting the
endogeneity problem persists in my model?
Any help would be appreciated.
Alejandro
--
Alejandro de la Fuente
Department of International Development/QEH
University of Oxford, Mansfield Road, Oxford OX1 3TB
Tel: 01865 281836
*
* 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/
*
* 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/
