[R] interpretation of conf.type in predict.Design {Design}

[Kingsford Jones]

Hi Dylan,

I haven't looked at the code for predict.Design or predict.lm, but I
think it's safe to assume that "mean" and "confidence" refer to the
same concept, as do "individual" and "prediction".   Here's my logic:

In general, confidence intervals refer to parameter estimates and
prediction intervals to predicted point values.  For the linear model,
the fitted line represents the estimated conditional "mean" of y given
the x-values and we form a "confidence" interval around it.  In this
case, our best "prediction" of any "individual" y-value given the
x-values is also the fitted line.  However, upon repeated sampling the
fitted line will vary less than the observed y values at any given set
of x-values, and this is reflected in the fact that the confidence
interval is narrower than the prediction interval.

hth,

Kingsford Jones

On Wed, Dec 31, 2008 at 2:15 PM, Dylan Beaudette
<dylan.beaudette at gmail.com> wrote:
> Hi,
>
> I am not quite sure how to interpret the differences in output when
> changing conf.type from the default "mean" to "individual". Are these
> analogous to the differences between "confidence" and "prediction"
> intervals, as defined in predict.lm {stats} ?
>
> Thanks in advance.
>
> Dylan
>
> ______________________________________________
> R-help at 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.
>