R help - May 2007 - Bootstrapped standard errors

Dear Friends,

I'm trying to learn to how to get Bootstrapped standard errors for estimated
coefficients from a regression.

For instance suppose I have the following model

logitmodel <- glm (y~X1+X2+X3, family=binomial(link="logit"))

beta <- logitmodel$coef

can somebody please guide me on how to use the package boot to obtain
bootstrapped SE's for the associated betas.


Thanks in advance

Anup

       
---------------------------------
Boardwalk for $500? In 2007? Ha! 

	[[alternative HTML version deleted]]

On Thu, 2007-05-17 at 17:43 -0700, Anup Nandialath
wrote:
> Dear Friends,
> 
> I'm trying to learn to how to get Bootstrapped standard errors for
> estimated coefficients from a regression. 
> 
> For instance suppose I have the following model
> 
> logitmodel <- glm (y~X1+X2+X3, family=binomial(link="logit"))
> 
> beta <- logitmodel$coef
> 
> can somebody please guide me on how to use the package boot to obtain
> bootstrapped SE's for the associated betas.
> 
> 
> Thanks in advance
> 
> Anup
You might want to review John Fox' document on Bootstrapping Regression
Models:

http://cran.r-project.org/doc/contrib/Fox-Companion/appendix-bootstrapping.pdf

HTH,

Marc Schwartz