I see it in the latest windows snapshot (built today -- I've never used cvs).
Here's the result from the model with X = {const black smsa76 south76 ed76} and
Z = {const black smsa66 south66 nearc4b}:
Hausman test -
Null hypothesis: OLS estimates are consistent
Asymptotic test statistic: Chi-square(3) = 9.92548
with p-value = 0.0192104
Weak instrument test -
Cragg-Donald minimum eigenvalue = 1.13201
Critical values for TSLS bias relative to OLS:
bias 5% 10% 20% 30%
value 0.00 0.00 0.00 0.00
Relative bias is probably less than 5%
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces(a)lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Wednesday, April 21, 2010 2:42 PM
To: Gretl list
Subject: Re: [Gretl-users] tsls tests
On Wed, 21 Apr 2010, Summers, Peter wrote:
> I should've thought of that. You're right of course. If I run a
> model dropping experience altogether, and instrument for ed76,
> smsa76 and south76, the Hausman test has df = 3.
>
> I'm still confused about the 0.0 critical values in the weak
> instruments test though...
Are you seeing those with the current snapshot from CVS?
Allin
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
