st: Re: Two step estimation and standard error computation

["Rodrigo A. Alfaro" <raalfaroa@gmail.com>]

Georgios,

I don't have access to the article that you mentioned. My suggestion is
to use the matrices e(b) and e(V) to construct the required std. error.
You could check -_robust- command in Programming manual.

Rodrigo.

----- Original Message ----- 
From: <gmc107@york.ac.uk>
To: <statalist@hsphsun2.harvard.edu>
Sent: Saturday, August 18, 2007 6:41 PM
Subject: st: Two step estimation and standard error computation



> Dear statalisters ,
> I was wondering whether somebody knows how to implement in Stata Vella and
> Verbeek's covariance matrix estimation practical impementation which is
> given in (p.259-260: Appendix A) in their Journal of econometrics (1999
> article:) "2 step estimation of panel data models with censored endogenous
> variables and selection bias".
> The only thing that's readily available in Stata for linear models is the
> hc3 (heteroskedasticity) robust standard errors option but this doesn't
> allow for the endogeneity correction terms and in the event the two terms
> are jointly statistically significant then the standard errors will be
> deflated and significance lebels overstated. Would simply going for say 
> 500
> boostrap replications that would be sufficient or would that be a very 
> na�ve
> approach? I have seen some papers that argue that this should be
> sufficient+avoids the computational perplexities.
> Would be most grateful of any help.
>
> Georgios
>
>
> *
> *   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/