Re: st: RE: Different coefficient magnitudes in ols and 2sls estimation

[Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>]

I think, in this situation the Hausman test would be the formal
procedure that tests the divergence between the two estimators. Under
the null of Y2 being exogenous, both the OLS and 2SLS estimator are
consistent and tend to the same probability limit, else the OLS is
inconsistent and the probability limits of the two estimators diverge.
The Hausman test is a formal tests of this convergence under the null.

T

2010/4/22 Nick Cox <n.j.cox@durham.ac.uk>:
> No, I meant whether your judgement "significantly smaller" reflected a
> difference confirmed as such by an appropriate formal procedure. I don't
> know what it would be, but this is not it.
>
> Nick
> n.j.cox@durham.ac.uk
>
> Sukesh Patro
> Sent: 22 April 2010 19:11
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: RE: Different coefficient magnitudes in ols and 2sls
> estimation
>
> Sorry Nick. I left out the t-stats. The OLS t-stat is about 3.00 and
> the 2SLS t-stat is about 2.50. Is this what you meant? I only left
> them out because they were part of the original question.
>
> Thanks again, SSKP
>
>
> On Thu, Apr 22, 2010 at 1:07 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
>> So, no formal significance test in either case. Thanks for the
>> clarification.
>>
>> Nick
>> n.j.cox@durham.ac.uk
>
> *
> *   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/
>



-- 
To every ω-consistent recursive class κ of formulae there correspond
recursive class signs r, such that neither v Gen r nor Neg(v Gen r)
belongs to Flg(κ) (where v is the free variable of r).

*
*   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/