Tuesday, January 03, 2006

Re: st: Bootstrap/Robust Standard Errors

Tinna, The White estimator will attempt to compensate for the heteroskedasticity. Depending on the procedure, you will have different options-- the standard sandwich ( with an added n/(n-k) adjustment), the hc2 or hc3 adjustments for leverage. The bootstrap provides empirical standard errors. The bias correction and acceleration is good if there is some skewness in the residual distribution here. Tilting might be better but Stata doesn't offer it yet. If you have autocorrelation bias in the residuals, you might want to consider the newey procedure (Newey-West regression). It does the White Correction plus one for autocorrelation of the residuals for a specified number of lags. But a bootstrap for time series data is not yet part of the Stata package. Hope this helps, Bob Yaffee

Robert A. Yaffee, Ph.D. Research Professor Shirley M. Ehrenkranz School of Social Work New York University

home address: Apt 19-W 2100 Linwood Ave. Fort Lee, NJ 07024-3171 Phone: 201-242-3824 Fax: 201-242-3825 yaffee@nyu.edu

----- Original Message ----- From: Tinna <statalist@gmail.com> Date: Monday, January 2, 2006 3:25 pm Subject: st: Bootstrap/Robust Standard Errors

> Dear Statalisters, > > A White test has revealed heteroscedastic SEs. > Can I use either -vce(robust)- or -vce(bootstrap)- in that incidence? > My understanding (after an internet search) is that both correct > biases in SEs due to heteroscedasticity in the error structure. Is one > more appropriate then the other? > > Thanks > Tina > > * > * 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/


Tag:


Links to this post:

Create a Link



<< Home

This page is powered by Blogger. Isn't yours?