[R] ANCOVA with defined error terms

Richard M. Heiberger rmh at temple.edu
Thu Aug 20 05:21:57 CEST 2009

track.aov <- aov(mice ~ coon
                  + block*veget*fruit*time - block:veget:fruit:time
                  + Error(block/plot), data = track)

I think this is what you are looking for.  This model in words says,
What is the effect of the four-way crossing after adjusting for
the covariate coon?  We don't have enough degrees of freedom for
the full crossing, so remove the four-way interaction.

anova() gives the sequential sums of squares of (multi-degree of
freedom) effects.  Be sure that the five variables (block, veget, fruit,
time, plot) are factors.  I am assuming time is a factor here.

More information about the R-help mailing list