[S] Anova confidence limits

David F. Parkhurst (parkhurs@ophelia.ucs.indiana.edu)
Thu, 16 Jul 1998 09:37:41 -0500


In one-way fixed-effect anova with t treatments and r replicates per
treatment, the expected mean squares for denominator and numerator are
respectively:
(variance of individual observations) and
(that variance + r*sum_over_treatments((treatment effect)^2)/(t-1))

I'd like to put confidence limits on the
sum_over_treatments((treatment effect)^2)/(t-1)
part.

Is there a "standard" way to do this?---I don't recall ever seeing it
done. In any event, does anyone see some way to do it, either
parametrically (assuming all the usual nice stuff), or by bootstrapping?

Thanks for any help.

Dave Parkhurst

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news