Displaying 1 result from an estimated 1 matches for "kirkspf2".
2006 Aug 14
2
lme() F-values disagree with aov()
...n data from a between-within subjects experiment. The correct
ANOVA table is known because this is a textbook example (Experimental Design by
Roger Kirk Chapter 12: Split-Plot Factorial Design). The lme() F-values differ from
the known results. Please help me understand why.
d<-read.table("kirkspf2.dat",header=TRUE)
for(j in 1:4) d[,j] <- factor(d[,j]) ### Make vars into type "factor"
##lme() results
library(nlme)
fit<-lme(y~a*b*c,random=~1|s, data=d)
anova(fit)
##correct anova table
##subjects are nested within a; a between, b & c within
fit2<-aov(y ~ a*b*c + E...