Sigmasquare hat=1/N* sum of square (yi-(k0hat+k1hatZi+B1hat*Qihat))
when it fact, we want
Sigmasquare hat=1/N* sum of square (yi-(k0hat+k1hatZi+B1hat*Qi))
So to speak, we need Q1 rather than Q1hat for correct standard errors.
Just want to know how this correction can be implemented at STATA.
One say that it is possible to correct the variance but generally tedious.
Looks so easy but not to me.
Any suggestion would make me easy
Thanks
WT
Quoting Stas Kolenikov <[email protected]>:
> There's been a discussion on the list about a week ago about two-stage
> estimation and correction of the variance estimates. Please look it
> up.
>
> On 8/30/06, [email protected] <[email protected]> wrote:
> > I attempted OLS estimation when one of the regressors has been estimated
> from a
> > first stage procedure(generated regressor)
> >
> > Y=c0+ c1X1+c2x2+...+rq(hat)+e : q(hat) is a generated regressor for my
> case
> >
> > from q=a0+a1z1+a2z2
> >
> > According to the reference(Wooldridge), an adjustment is needed for the
> > asymptotic variance of all OLS estimators.
> > I just want to know how implemented this adjustment at STATA. Besides, I
> want to
> > know whether rreg(robust regression) is applied to OLS with generated
> regressor
> >
> > Any comments and suggestion are welcomed.
> >
> > Thanks
> >
> > WT
> > *
> > * 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/
> >
>
>
> --
> Stas Kolenikov
> http://stas.kolenikov.name
> *
> * 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/
