Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Kit Baum <baum@bc.edu> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: RE: ivreg2 and xtoverid error |
Date | Sun, 4 Apr 2010 09:43:20 -0400 |
On Apr 4, 2010, at 2:33 AM, John wrote: > Also, concerning the power issue, on one hand, with more instruments the > model has more ways to go wrong so ceteris paribus, power to detect > misspecification goes up with more degrees of freedom, correct? On the > other hand, with weak instruments the power of the test is reduced. I > guess a simulation would be needed to settle this. Your intuition is well taken, and you might think that if you had hundreds of instruments, if for no other reason than brute force the Sargan-Hansen test would be able to reject. But aficionados of xtabond/xtabond2 know well the problem that with several hundred Arellano-Bond generated instruments, Sargan test pvals tend to be reported as 1.000. This may well be, along the lines of your intuition, the result of many of them not pulling much weight in the auxiliary regression, but in this context it is well known that Sargan tests often tell you very little. Kit Baum | Boston College Economics & DIW Berlin | http://ideas.repec.org/e/pba1.html An Introduction to Stata Programming | http://www.stata-press.com/books/isp.html An Introduction to Modern Econometrics Using Stata | http://www.stata-press.com/books/imeus.html * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/