Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: ivreg coefficients not as they should be?


From   "Gustavo Sanchez" <gsanchez@stata.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: ivreg coefficients not as they should be?
Date   Tue, 6 Nov 2007 09:34:15 -0600

Ognjen <ognjen.obucina@upf.edu> asked,

>I am actually in the process of getting familiar with the ivreg command and
>I am facing what I think I can call a problem. Namely, I am working with
some
>sample dataset (for the purpose of exercise) and I tried to reproduce ivreg
>coefficients doing two-stage regression. Alas, the ordeer of magnitude of
>coefficients is same, but they are far from being identical. Is this
strange?

Without looking at the actual procedure that Ognjen is implementing I cannot
determine with certainty what might be the cause of the problem, but I would
suggest that Ognjen makes sure that he is including all the exogenous
variables in the first stage regression. Omitting some of the exogenous
variables in the first step would lead to biased coefficient estimates in
the second stage regression. If Ognjen still has problems to reproduce the
results from -ivreg-, he could send me (privately) his dataset and the code
that he is running, and I would look for the potential cause of the
difference in the results that he is reporting.

-- Gustavo
gsanchez@stata.com



*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index