# st: RE: statalist-digest V4 #3109

 From "Stephen P. Jenkins" To Subject st: RE: statalist-digest V4 #3109 Date Wed, 2 Jul 2008 09:40:47 +0100

```------------------------------

Date: Tue, 01 Jul 2008 10:11:07 +0200
From: daniela vuri <daniela.vuri@uniroma2.it>
Subject: st: mvprobit and overidentification test

Dear Statalist users,
I wonder if there is any way to compute an overidentification test
after
using the "mvprobit" command.
I have a system of 3 equations, where the dependent variables are
binary (0 or 1); moreover, 2 dependent variables enter as covariates
in
the third equation. In other words:

X1=delta1+lambda1*Z1+epsilon1
X1=delta2+lambda2*Z2+epsilon2
Y=alpha+beta1*X1+beta2*X2+epsilon2

The system of equations is estimated using the command "mvprobit".
The variables Z1 and Z2 are used to as instruments to identify the
model. Since Z1 and Z2 contain more than one variable, i would like to
perform an overidentification test.
Any suggestion on how to do it in this case?
===============================

>>>>>>>>>>>>

How about a LR test of the full model (A) against the model which
excludes the "extra" instruments (B)?

Over-identification tests assume that both A and B are identified; in
your case, the null hypothesis is that A is 'over-identified' relative
to B.

Stephen
-------------------------------------------------------------
Professor Stephen P. Jenkins <stephenj@essex.ac.uk>
Director, Institute for Social and Economic Research
University of Essex, Colchester CO4 3SQ, U.K.
Tel: +44 1206 873374.  Fax: +44 1206 873151.
http://www.iser.essex.ac.uk
Survival Analysis using Stata:
http://www.iser.essex.ac.uk/teaching/degree/stephenj/ec968/

Learn about the UK's new household panel survey, the United Kingdom
Household Longitudinal Study: http://www.iser.essex.ac.uk/ukhls/

*
*   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/
```