R help - Jan 2011 - robust standard error of an estimator

Hi,

I have ove the robust standard error of an estimator but I don't know how to
do this.
The code for my regression is the following:
reg<-lm(fsn~lctot)
But then what do I need to do?

-- 
Charlène Lisa Cosandier

	[[alternative HTML version deleted]]

It depends on what you mean by "robust."  Robust to what?

I recommend looking at the sandwich package which gives  
heteroskedasticity and autocorrelation robust variance/covariance  
matrices.  For instance, you could do the following to get your OLS  
estimates with heteroskedasticity consistent SEs

library(sandwich)
library(lmtest)
reg=lm(fsn~lctot)
coeftest(reg, vcov=vcovHC(reg))

Or to get cluster robust SEs, check out this: people.su.se/~ma/ 
clustering.pdf

Hope that helps.

Andrew Miles


On Jan 1, 2011, at 10:09 AM, Charlène Cosandier wrote:
> Hi,
>
> I have ove the robust standard error of an estimator but I don't  
> know how to
> do this.
> The code for my regression is the following:
> reg<-lm(fsn~lctot)
> But then what do I need to do?
>
> -- 
> Charlène Lisa Cosandier
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

	[[alternative HTML version deleted]]

Is the (non-clustered) sandwich estimator really robust to autocorrelation?
Thanks
Frank

-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
-- 
View this message in context:
http://r.789695.n4.nabble.com/robust-standard-error-of-an-estimator-tp3170257p3170363.html
Sent from the R help mailing list archive at Nabble.com.