[R-sig-ME] replicated strip-plot design

Antonella Plaia plaia at unipa.it
Mon Feb 28 15:19:51 CET 2011

Hi everyone,
I need some help to write the right model to analyze data from a  
strip-plot design replicated over time.
These are the variables:
A: 3 level factor as row treatment,
B: 3 level factor as column treatment,
C: 2 blocks
Y: continuous response variable.
Data have been observed once a year for 17 years, together with some  
meteorological variables.
Time can be considered from two different point of views: as a count  
variable (1:17) to model the eventual additive effect over time,
but also as a random factor to take into account the effect of  
different meteorological conditions.
Therefore I have these variables too:
Year: count variable in the range 1:17,
W: a countinuous variable which summarizes weather condition.

And this is the model I have considered:

lmer(y~Year*A*B-1+(1|as.factor(Year))+(0+W|C:A)+(0+W|C:B), data=stripplot).
Is it the right one?

Thanks in advance for the help,

Prof. Antonella Plaia
DSSM - University of Palermo
viale delle scienze ed. 13
90128 Palermo - Italy
Tel. +39 091 6626244 - Fax +39 091 485726

This message was sent using the University of Palermo web mail interface.

More information about the R-sig-mixed-models mailing list