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, statalist.org is already up and running.


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

st: Quick help on Sargan/Hansen Stat


From   Pedro Ferreira <pedro.ferreira.cmu@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Quick help on Sargan/Hansen Stat
Date   Thu, 7 Apr 2011 01:22:19 -0400

Hello everyone,

quick question on Sargan/Hansen implementation in STATA. Do we "want"
a small or a large "Prob>chi2"? I swear I have seen all kinds of
interpretations floating around on the web, which is quite disturbing,
and I am not able to find the STATA documentation explaining exactly
what is being tested. Needless to say, I am confused now. Can someone
help me out? My understanding is that we "want" a "large Prob>chi2"
because we want the moment conditions to go to zero. So, if we have
something like "Prob>chi2 = 0.213" then we are ok, meaning that we
will not reject that moments on overidentifying restrictions are close
to zero and this is the best we can do with this type of test. I am
interpreting the Hansen statistic in the same way, we also look for a
"large Prob>chi2" and we trust Hansen over Sargan depending on whether
we have heteroskedasticity. Very much appreciated for your help,

Pedro
*
*   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/


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