On Jun 14, 2014, at 1:15 AM, Dan Hughes wrote:
> Dear users,
>
> Can anyone help with a rather simplistic question about the interpretation
> of output using the Boot ( ) function in the Car package?
>
> I am trying to bootstrap a simple univariate linear regression in order to
> look at the bootstrap regression coefficients e.g.
>
> mod1 <- lm (y~x)
> mod1.b <- Boot(mod1, R=999)
> summary (mod1.b)
>
> This returns output like this:
>
> R original bootBias bootSE bootMed
> (Intercept) 999 5.365528 -8.0960e-05 0.0501306 5.365241
> x 999 0.032211 7.3683e-05 0.0029743 0.032158
>
> I'm having trouble finding out what "bootMed" means. If
'original' is the
> non-bootstrapped regression coefficient, is bootMed the bootstrapped
> regression coefficient?
>
> I presume this is a frighteningly simple question for regular car package
> users, but I can't seem to find an explanation what this means.
I'm not a regular or even an intermittent user of `car` but offhand I would
suspect it is the median.
Do read the Posting Guide,
and the help page for Boot and the tutorial that is linked from that help page,
and please learn to post in plain text.
>
> Best wishes
>
> Dan Hughes
>
> [[alternative HTML version deleted]]
--
David Winsemius
Alameda, CA, USA