Sonia,
I also can't replicate the problem. I'm a coauthor of ivreg2, so can
you contact me off-list with a detailed log file of the output of
ivreg2 and ivreg, and we'll see if we can locate what is going on?
Cheers,
Mark
Subject: RE: st: ivreg versus ivreg2 and pweight
Date sent: Wed, 21 Dec 2005 15:52:01 -0500
From: "Sonia Laszlo, Prof." <[email protected]>
To: <[email protected]>
Send reply to: [email protected]
> Thanks Scott,
>
> I just got the current version of ivreg2 and had already just updated
> Stata. I am still getting the same problem as I did before. I am still
> confused why this might be happening.
>
> Sonia
>
> ________________________________________________________________________
> ___
>
> Do you have the current version of -ivreg2- (2.1.14) and an updated
> Stata?
> Using an example data set I cannot replicate this problem.
>
>
> Scott
>
> . clear
>
> . use http://fmwww.bc.edu/ec-p/data/hayashi/griliches76.dta
> (Wages of Very Young Men, Zvi Griliches, J.Pol.Ec. 1976)
>
> . qui ivreg2 lw s expr tenure rns smsa (iq=med kww mrt) [pw= age]
>
> . estimates store ivreg2_pw
>
> . qui ivreg lw s expr tenure rns smsa (iq=med kww mrt) [pw= age]
>
> . estimates store ivreg_pw
>
> . qui ivreg2 lw s expr tenure rns smsa (iq=med kww mrt)
>
> . estimates store ivreg2
>
> . qui ivreg lw s expr tenure rns smsa (iq=med kww mrt)
>
> . estimates store ivreg
>
> . estout ivreg2_pw ivreg2 ivreg_pw ivreg, style(fixed) ///
> > stats(N) drop(iq s expr tenure rns smsa ) collabels(,none)
>
> ivreg2_pw ivreg2 ivreg_pw ivreg
> _cons 3.865665 3.84138 3.865665 3.84138
> N 758 758 758 758
>
>
>
>
> . which ivreg2
> c:\ado\plus\i\ivreg2.ado
> *! ivreg2 2.1.14 4Dec2005
> *! authors cfb & mes
> *! cloned from official ivreg version 5.0.9 19Dec2001
> *! see end of file for version comments
>
> . update query
> (contacting http://www.stata.com)
>
> Stata executable
> folder: D:\Stata9\
> name of file: wsestata.exe
> currently installed: 15 Sep 2005
> latest available: 15 Sep 2005
>
> Ado-file updates
> folder: D:\Stata9\ado\updates\
> names of files: (various)
> currently installed: 13 Dec 2005
> latest available: 13 Dec 2005
>
> > Greetings,
> >
> > I am running a 2SLS with IV. When I run my regression using
> > ivreg, it
> > runs the regression over the correct number of observations (call
> > it n).
> > When I run it using ivreg2, it runs the regression over n-1
> > observations:
> >
> > ivreg lhours (lw = z1 z2 z3) x1 x2 x3 x4 x5 ===> N=n
> > ivreg2 lhours (lw = z1 z2 z3) x1 x2 x3 x4 x5 ===> N=n-1
> >
> > In both cases I use weights ([pweight]). This seems to be causing
> the
> > problem. This seems to also be a problem using [aweight]. However,
> N=n
> > for both ivreg and ivreg2 when I do not include weights. Is there a
> > reason for this discrepancy using weights? Or am I overlooking
> > somethingobvious here?
> >
> > Thank you
> >
> > Sonia Laszlo
> >
> >
> >
> > _____________________________
> > Sonia Laszlo
> > Department of Economics,
> > McGill University
> > 855 Sherbrooke Street West
> > Montreal, Quebec, H3A 2T7
> > Tel: 514-398-1924
> > Fax: 514-398-4938
> >
> >
> >
> >
>
> *
> * 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/
Prof. Mark E. Schaffer
Director
Centre for Economic Reform and Transformation
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS UK
44-131-451-3494 direct
44-131-451-3296 fax
44-131-451-4202 CERT administrator
http://www.sml.hw.ac.uk/cert
*
* 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/
