[R-sig-ME] design of nested mixed effects model with repeated measures

José Manuel Blanco Moreno jmblanco at ub.edu
Fri Dec 24 20:53:19 CET 2010


El 24/12/2010 17:45, Andrew Dolman escribió:
> Patch is not random either and even if it were it only has 2 levels
> and so cannot properly be fit as random. It's also crossed with block
> and with fire. (It's also a confusing name. Consider calling it
> habitat, or canopy(T,F))
>
I may be misunderstanding the design, but I would be somewhat reluctant to throw 
away the random effect for patch. Indeed I think that Andrew is right when 
considering renaming this factor as canopy or habitat (a factor wih two levels); 
however, if the "replicates" are taken from beneath a single canopy (instead of 
selecting different canopies from a plot), should not this nesting be included 
in the random effects, even if the structure seems complicated?
Moreover, I do not think that canopy and plot are crossed, since you cannot move 
one canopy from one plot to another, and each canopy on each plot is actually a 
different individual.
Forgive me if I got it completely wrong. And merry Christmas by the way.


-- 
---------------------------------------
José M. Blanco Moreno

Dept. de Biologia Vegetal (Botànica)
Universitat de Barcelona
Av. Diagonal 645
08028 Barcelona SPAIN
---------------------------------------

phone: (+34) 934 039 863
fax: (+34) 934 112 842
e-mail: jmblanco at ub.edu




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