RE: st: RE: regression when endogenous regressor continuous?

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Nirina,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu 
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nirina F
> Sent: 28 June 2007 09:20
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: RE: regression when endogenous regressor continuous?
> 
> Actually the error it gives says: matrix should be rank 122, 
> then I try to reduce the number of iv then it gives me the 
> same error but in lower rank: matrix should be 117.

This is almost certainly a problem with your model and data, not with
-ivprobit-.  It might be a problem of collinearity among variables, but
we can't tell without more details.

--Mark

> What about gmm? is there a one-command for a system of equations?
> Thanks again for your reply,
> Best,
> Nirina
> 
> On 6/27/07, Schaffer, Mark E <M.E.Schaffer@hw.ac.uk> wrote:
> > Nirina,
> >
> > > -----Original Message-----
> > > From: owner-statalist@hsphsun2.harvard.edu
> > > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf 
> Of Nirina F
> > > Sent: 27 June 2007 19:47
> > > To: statalist@hsphsun2.harvard.edu
> > > Subject: Re: st: RE: regression when endogenous regressor 
> continuous?
> > >
> > > Unfortunately ivprobit doesn't work. It gives me error problem.
> >
> > Can you explain what you mean by this?
> >
> > > I would like to know if there is a way to estimate two equations 
> > > ($endogvar= $exogvar) (depvar = $endogvar $exogvar) with 
> one command 
> > > like biprobit but for continuous variables.
> >
> > I don't think there is either a built-in or readily-available 
> > user-written command that does this in one command.
> >
> > If you are willing to estimate in two equations, and you 
> don't need a 
> > system estimator, then
> >
> > oprobit $endogvar $exogvar
> >
> > ivprobit depvar ($endogvar=$exogvar), twostep
> >
> > should give you consistent estimates for both equations.
> >
> > --Mark
> >
> > > Thanks
> > > Nirina
> > >
> > > On 6/27/07, Schaffer, Mark E <M.E.Schaffer@hw.ac.uk> wrote:
> > > > Nirina,
> > > >
> > > > I think your problem can be restated as one where IV
> > > estimation gives
> > > > you consistent estimates, so my instinct is the same as Dirk's.
> > > >
> > > > Probably you should take a look at -ivprobit-.  If your
> > > main interest
> > > > is the depvar equation, then the -twostep- option should,
> > > if I'm not
> > > > mistaken, give you consistent estimates.
> > > >
> > > > Cheers,
> > > > Mark
> > > >
> > > > > -----Original Message-----
> > > > > From: owner-statalist@hsphsun2.harvard.edu
> > > > > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf
> > > Of Nirina F
> > > > > Sent: Wednesday, June 27, 2007 11:51 AM
> > > > > To: statalist@hsphsun2.harvard.edu
> > > > > Subject: Re: st: RE: regression when endogenous regressor
> > > continuous?
> > > > >
> > > > > Hi,
> > > > > actually No, because I don't have instrumental variables .
> > > > > I would appreciate your help.
> > > > > thanks
> > > > >
> > > > > On 6/27/07, Nachbar, Dirk <Dirk.Nachbar@infores.com> wrote:
> > > > > > You would use the ivreg command.
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: owner-statalist@hsphsun2.harvard.edu
> > > > > > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf
> > > Of Nirina
> > > > > > F
> > > > > > Sent: 27 June 2007 11:21
> > > > > > To: statalist@hsphsun2.harvard.edu
> > > > > > Subject: st: regression when endogenous regressor 
> continuous?
> > > > > >
> > > > > > Hello all ,
> > > > > >
> > > > > > If my endogenous variable was a binary (1,0) then I 
> would have 
> > > > > > estimated the following equation with stata
> > > > > >                        biprobit ($endogvar= 
> $exogvar) (depvar 
> > > > > > = $endogvar $exogvar)
> > > > > >
> > > > > > where depvar is my dependent variable which 0 or 1 but Now
> > > > > since the
> > > > > > endogvar is years each individuals smoke (0 or 1 or 2 until
> > > > > 18), how
> > > > > > would I run ($endogvar= $exogvar) (depvar = $endogvar
> > > > > $exogvar) with
> > > > > > STATA.
> > > > > >
> > > > > > Many thanks
> > > > > >
> > > > > > Nirina
> > > > > > *
> > > > > > *   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/
> > > > > >
> > > > > *
> > > > > *   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/
> > > >
> > > *
> > > *   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/
> >
> *
> *   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/