[R-sig-ME] extracting coefficient standard errors

[Kari Ruohonen]

There is se.coef() method in the arm package. A special case is
se.fixef().

regards, Kari


On Tue, 2008-12-30 at 14:31 -0500, Ben Bolker wrote:
> doh.
>   I looked at all the fancy stuff in the summary method
> and overlooked the obvious (well, sort of obvious)
> 
>   cheers
>     Ben
> 
> Doran, Harold wrote:
> > sqrt(diag(vcov(mm)))
> > 
> > 
> > -----Original Message-----
> > From: r-sig-mixed-models-bounces at r-project.org on behalf of Ben Bolker
> > Sent: Tue 12/30/2008 11:44 AM
> > To: R Mixed Models
> > Subject: [R-sig-ME] extracting coefficient standard errors
> >  
> > 
> >   Does anyone happen to know a more elegant/robust way
> > to extract the (approximate) standard errors of fixed
> > effect parameters than
> > 
> > vcov(mm)@factors$correlation at sd
> > 
> >   ?
> > 
> 
>