[R-sig-ME] help with repeated measures on a split-plot experiment

[James Hudson]

I require assistance to properly code an lmer function. I have repeatedly
measured vegetation in a split-plot, agricultural-style experiment. A pdf of
the dat file is attached. If there is any additional information I should
include, please let me know.

The design follows:
Plots were assigned to 1 of 3 snow levels (1, 2 or 3). Then, each plot was
split into 1 of 2 temperature levels (A or B). These split-plots were
replicated 6 times each. All plots were then measured repeatedly over
several years.

I am interested in the fixed effects of the treatments (snow, temperature,
time, and their interactions).

My statistical background is not advanced enough for me to properly code the
random effects of both repeated measures and the split-plot design. I expect
an AR(1) structure for 'year.' Here's my attempt at the code (it doesn't
work):


cass<-as.data.frame(read.csv("cass.csv", header=TRUE))
lmer(response ~ snow*warm*year + (1|snow/warm/year) + (0+year|plot), cass))

Any help would be appreciated.

James Hudson
UBC Geography
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cass.pdf
Type: application/pdf
Size: 32828 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20080527/538f31c2/attachment.pdf>