Message-ID: <loom.20081220T192651-900@post.gmane.org>
Date: 2008-12-20T19:29:56Z
From: Dieter Menne
Subject: How test contrasts/coefficients of Repeated-Measures ANOVA?
Ben Meijering <b.meijering <at> student.rug.nl> writes:
> I'm doing a Repeated-Measures ANOVA, but I don't know how to test its
> contrasts or where to find the p-values of its coefficients. I know
> how to find the coefficient estimates of a contrast, but not how to
> test these estimates.
>
> First I do something like:
>
> y.aov <- aov(y ~ fac1 * fac2 + Error(subj/(fac1 * fac2)), data=data)
>
> Then, with
>
> coef(y.aov)
>
> I get the coefficients (which are, by default, treatment contrasts?),
> but p-values are not provided.
See model.table and se.contrast; and consider using lme in package nlme as an
alternative which make live much easier, if it can handle your (over?)complex
nesting.
Dieter