[R] p values when using rlm

[Marc Schwartz]

On Sat, 2005-03-26 at 20:36 -0500, John Sorkin wrote:
> R 2.0.1
> Linux
>  
> I am using rlm() to fit a model, e.g. fit1<-rlm(y~x). My model is more
> complex than the one shown.
>  
> When I enter summary(fit1) 
> I get estimates for the model's coefficients along with their SEs, and
> t values, but no p values. The p value column is blank.
> Similarly, when I enter anova(fit1) I get DF, Sum Sq, Mean Sq, but the
> column for F value and Pr(>F) are blank. 
>  
> Any suggestions that will allow me to get the p values for summary, the
> F and p values for anova?
> Thanks,
> John

John,

Perhaps these two posts may be of help to you:

http://tolstoy.newcastle.edu.au/R/help/04/10/5260.html

http://tolstoy.newcastle.edu.au/R/help/04/10/5252.html

I also located via a Google search the f.robftest() function in the
'sfsmisc' package on CRAN:

http://finzi.psych.upenn.edu/R/library/sfsmisc/html/f.robftest.html

but I would have to defer to the respective authors as to the pros and
cons of any of these approaches.

HTH,

Marc Schwartz