Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: RE: Re: About weakiv
From
"Schaffer, Mark E" <[email protected]>
To
"[email protected]" <[email protected]>
Subject
st: RE: Re: About weakiv
Date
Sat, 21 Dec 2013 12:59:45 +0000
Tak Wai,
The AR stat implemented by -weakiv- is just the standard AR stat, either with or without a finite-sample adjustment. No need to refer to "the performance of AR in weakiv" - just "the performance of AR" is good enough!
There are papers around that have done Monte Carlo studies of AR and related stats. The one that comes to mind is the 2005 working paper version of Chernozhukov and Hansen (2008), cited in the -ivreg2- help file and available here:
http://faculty.chicagobooth.edu/christian.hansen/research/ch_weakiv_mar07.pdf
Their simulations include specifications where the VCV used in weak-instrument-robust inference is heteroskedastic-robust, which is the case you're interested in.