Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: kclass estimators, xtivreg2


From   "Alejandro Quijada" <aquijadab@hotmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   RE: st: RE: kclass estimators, xtivreg2
Date   Tue, 06 Mar 2007 08:08:09 -0400

I'll follow your advise,

Thanks a lot,


From: "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
Reply-To: statalist@hsphsun2.harvard.edu
To: <statalist@hsphsun2.harvard.edu>
Subject: st: RE: kclass estimators, xtivreg2
Date: Mon, 5 Mar 2007 23:00:29 -0000

Alejandro,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of
> Alejandro Quijada
> Sent: 28 February 2007 13:32
> To: statalist@hsphsun2.harvard.edu
> Subject: st: kclass estimators, xtivreg2
>
> Dear statalist members,
>
> Since I have a panel dataset with a problem of errors in
> variables (EV), I wanted to use the class of estimators
> proposed by Verbeek and Nijman (1993) and Deaton (1985) which
> try to correct this kind of problem. Adopting the notation of
> Verbeek and Nijman (1993), the estimator is given by:
>
> B(alpha) = (mxy - alpha*sigma)/(Mxx - alpha*omega)
>
> Where Omega and sigma represent the structure on the
> measurement errors. The correction parameter alpha for
> Verbeek and Nijman is aproximately T-1/T, while Deaton'
> correction parameter would be a constant alpha=1.
>
> I think xtivreg2 is the indicated stata command to implement
> these two estimators, since it has the kclass option where
> you can put the correction parameter like fuller's option
> does. So, translating these two parameters into xtivreg2's
> language, I think they would be kclass=1 + (T-1/T) for
> Verbeek and Nijman and kclass=2 for Deaton. Is that right? if
> not, could you please guide me? I use stata 8.2.

It's hard to tell without more detail (usually on Statalist you are encouraged to provide fuller references and discussion - you can't count on everybody knowing this literature). But you are right that xtivreg2 supports k-class estimation such as Fuller's LIML, Nagar's bias-corrected IV estimator, etc. If so, then you can get the results you want using the -kclass- option ... but it's up to you to make sure that the -kclass- parameter is correctly specified.

One alternative if you're not sure is to replicate some published results that use these estimators.

HTH.

Cheers,
Mark

> Thank you very much for your help
>
> Best regards,
>
> Susana
>
> _________________________________________________________________
> Charla con tus amigos en línea mediante MSN Messenger:
> http://messenger.latam.msn.com/
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
_________________________________________________________________
Charla con tus amigos en línea mediante MSN Messenger: http://messenger.latam.msn.com/

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* 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   |   What's new   |   Site index