Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: overidentification test after cmp

From   Conor Hughes <>
Subject   Re: st: overidentification test after cmp
Date   Wed, 7 Jul 2010 10:06:16 +0700

Hi Sharon,

Personally I've never used the --cmp-- command, so I can't really help
you on that point.  I did just want to warn you, in case this is a
paper you're hoping to publish or something, to be careful about your
wording.  There is no test that can prove that your IV's are
exogenous.  Overidentification can help you make your case, but it
doesn't at all prove that your instruments satisfy the exclusion
criteria.  Really the only way to proceed is with a reasoned
explanation about why you believe the instruments are exogenous.  I
don't know what your variables are exactly, but to claim that fem_work
and male_educ affect the number of children only through their effect
on fem_educ seems a bit tenuous.

I apologize that I couldn't answer your Stata question, but hopefully
this helps with your analysis and interpretation of the results in the


On Tue, Jul 6, 2010 at 11:40 PM, xueliansharon <> wrote:
> Dear all,
> I wonder how to do the overidentification test after cmp command.  Here,
> kids is the dependent variable, fem_educ is the endogenous variable, and
> fem_work, male_educ are the IVs for fem_educ. I employ cmp command to
> realize the IV ordered probit procedure since kids is an ordered variable.
> But then I want to test whether the IVs are correctly excluded from the
> model and whether they are correlated to the error term, that is, I want to
> do overidentification test. So how to realize this test?
> my codes:
> cmp setup
> webuse laborsup, clear
> cmp (kids = fem_educ) (fem_educ = fem_work male_educ), ind($cmp_oprobit
> $cmp_cont) tech(dfp)
> Thanks,
> Sharon
> --
> View this message in context:
> Sent from the Statalist mailing list archive at
> *
> *   For searches and help try:
> *
> *
> *

Conor Hughes
Mathematics and Economics
University of Chicago 2011

*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index