Dear Daniel,
One can indeed compute 2SLS estimates by two OLS regressions (hence the
name), though the usual OLS standard errors from the second stage will then
require correction, so some care is needed.
Regards,
John
--------------------------------
John Fox
Department of Sociology
McMaster University
Hamilton, Ontario
Canada L8S 4M4
905-525-9140x23604
http://socserv.mcmaster.ca/jfox
--------------------------------
> -----Original Message-----
> From: Daniel_J_Celta@fpl.com [mailto:Daniel_J_Celta@fpl.com]
> Sent: Tuesday, July 25, 2006 7:35 PM
> To: John Fox
> Cc: 'Neil Hepburn'; s-news@lists.biostat.wustl.edu;
> s-news-owner@lists.biostat.wustl.edu
> Subject: Re: [S] Two stage least squares
>
> Dear John,
> I have to disagree with your first statement.
>
> Ordinary Least Squares (OLS), and Two Stage Least Squares are
> closely related (2SLS).
>
> 2SLS is a method that consist on running OLS in two stages.
>
> First :
> Run OLS on the reduced form equations for each endogenous
> variable that appear as explanatory variables in the
> structural equations in the system
>
> Second:
> Substitute the reduced-form Y's (instrumental variables) for
> the Ys that appear on the right side (only) of the structural
> equations, and then estimate these revised structural
> equations with OLS.
>
> For a person with no econometrics background it may sound
> strange but OLS is based on a least squares approach but
> additional tests are added such as Durbin Watson, etc.
>
> Although I am not familiar with the R sem library, I do agree
> that 2SLS should be easy to implement from scratch.
> Thanks
>
>
>
> Daniel J Celta Ph.D.
> Project Valuation - Special Valuations
> 561 691 7653
>
>
>
>
>
> "John Fox"
>
>
> <jfox@mcmaster.ca> To:
> Daniel_J_Celta@fpl.com, "'Neil Hepburn'"
> <nhepburn@ualberta.ca>
> Sent by: cc:
> s-news@lists.biostat.wustl.edu
>
> s-news-owner@lists.biosta
> Subject: Re: [S] Two stage least squares
>
> t.wustl.edu
>
>
>
>
>
>
>
>
> 07/22/2006 10:06 AM
>
>
>
>
>
>
>
>
>
>
>
>
> Dear Daniel and Neil,
>
> I'm not sure what the OLS function in Finmetrics does, but
> the name doesn't suggest 2SLS.
>
> Although I haven't tried it, you could probably port the tsls
> function in the R sem package to S-PLUS with little effort.
> Indeed, programming 2SLS from scratch shouldn't be too hard,
> since the calculations are very simple.
>
> I hope this helps,
> John
>
> --------------------------------
> John Fox
> Department of Sociology
> McMaster University
> Hamilton, Ontario
> Canada L8S 4M4
> 905-525-9140x23604
> http://socserv.mcmaster.ca/jfox
> --------------------------------
>
> > -----Original Message-----
> > From: s-news-owner@lists.biostat.wustl.edu
> > [mailto:s-news-owner@lists.biostat.wustl.edu] On Behalf Of
> > Daniel_J_Celta@fpl.com
> > Sent: Wednesday, July 19, 2006 8:35 PM
> > To: Neil Hepburn
> > Cc: s-news@lists.biostat.wustl.edu
> > Subject: Re: [S] Two stage least squares
> >
> > Neil
> > Try the Finmetrics library. There is an OLS function that
> may do the
> > trick.
> >
> > Daniel J Celta
> > Project Valuation - Special Valuations
> > 561 691 7653
> >
> >
> >
> >
> >
> > "Neil Hepburn"
> >
> >
> > <nhepburn@ualberta.ca> To:
> > s-news@lists.biostat.wustl.edu
> >
> > Sent by: cc:
> >
> >
> > s-news-owner@lists.biosta
> > Subject: [S] Two stage least squares
> >
> > t.wustl.edu
> >
> >
> >
> >
> >
> >
> >
> >
> > 07/19/2006 08:40 PM
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > I'm looking for a library to do two stage least squares in S-Plus.
> > I've found the sem library for R but have not found anything for
> > S-Plus. Anyone got any ideas?
> >
> > Cheers,
> > Neil
> >
> >
> > =============================================
> > Neil Hepburn, Economics Instructor
> > Department of Social Sciences, Augustana Faculty University
> of Alberta
> > Camrose, Alberta
> >
> > Email nhepburn@ualberta.ca
> > URL http://www.augustana.ca/programs/profs/nhepburn/
> >
> >
> > --------------------------------------------------------------------
> > This message was distributed by
> > s-news@lists.biostat.wustl.edu. To unsubscribe send e-mail to
> > s-news-request@lists.biostat.wustl.edu with the BODY of the
> message:
> > unsubscribe s-news
>
> --------------------------------------------------------------------
> This message was distributed by
> s-news@lists.biostat.wustl.edu. To unsubscribe send e-mail
> to s-news-request@lists.biostat.wustl.edu with the BODY of
> the message: unsubscribe s-news
>
>
>
>
>
|