[R] contrast matrix for aov
Darren Weber
DarrenLeeWeber at gmail.com
Thu Mar 10 01:03:42 CET 2005
How do we specify a contrast interaction matrix for an ANOVA model?
We have a two-factor, repeated measures design, with
Cue Direction (2) x Brain Hemisphere(2)
Each of these has 2 levels, 'left' and 'right', so it's a simple 2x2 design
matrix. We have 8 subjects in each cell (a balanced design) and we want to
specify the interaction contrast so that:
CueLeft>CueRght for the Right Hemisphere
CueRght>CueLeft for the Left Hemisphere.
Here is a copy of the relevant commands for R:
########################################
lh_cueL <- rowMeans( LHroi.cueL[,t1:t2] )
lh_cueR <- rowMeans( LHroi.cueR[,t1:t2] )
rh_cueL <- rowMeans( RHroi.cueL[,t1:t2] )
rh_cueR <- rowMeans( RHroi.cueR[,t1:t2] )
roiValues <- c( lh_cueL, lh_cueR, rh_cueL, rh_cueR )
cuelabels <- c("CueLeft", "CueRight")
hemlabels <- c("LH", "RH")
roiDataframe <- data.frame( roi=roiValues, Subject=gl(8,1,32,subjectlabels),
Hemisphere=gl(2,16,32,hemlabels), Cue=gl(2,8,32,cuelabels) )
roi.aov <- aov(roi ~ (Cue*Hemisphere) + Error(Subject/(Cue*Hemisphere)),
data=roiDataframe)
print(summary(roi.aov))
########################################
I've tried to create a contrast matrix like this:
cm <- contrasts(roiDataframe$Cue)
which gives the main effect contrasts for the Cue factor. I really want to
specify the interaction contrasts, so I tried this:
########################################
# c( lh_cueL, lh_cueR, rh_cueL, rh_cueR )
# CueRight>CueLeft for the Left Hemisphere.
# CueLeft>CueRight for the Right Hemisphere
cm <- c(-1, 1, 1, -1)
dim(cm) <- c(2,2)
roi.aov <- aov( roi ~ (Cue*Hemisphere) + Error(Subject/(Cue*Hemisphere)),
contrasts=cm, data=roiDataframe)
print(summary(roi.aov))
########################################
but the results of these two aov commands are identical. Is it the case that
the 2x2 design matrix is always going to give the same F values for the
interaction regardless of the contrast direction? OR, is there some way to get
a summary output for the contrasts that is not available from the print method?
More information about the R-help
mailing list