Mean separation test in lmer/glmer models

Try Multcomp package, 
Ferrnando

-----Mensaje original-----
De: r-sig-mixed-models-bounces at r-project.org
[mailto:r-sig-mixed-models-bounces at r-project.org] En nombre de Bin Liu
Enviado el: viernes, 15 de junio de 2012 10:50 a.m.
Para: R-sig-mixed-models at r-project.org
Asunto: [R-sig-ME] Mean separation test in lmer/glmer models

Hi All,

I will be dealing with agricultural experimental design data which has
response (like Yield) and categorical covariates. The covariates can be
water, air, Light and Nitrate (all treatments have multiple levels). The
design is a randomized complete block design with all combinations of
treatments in different blocks. Different combination may have different
sample size. We take the block factor as random and fit a linear mixed
model (or GLMM) using lmer() in R. Our goal is to find the best combination
of treatments which has the most yield, and some combinations which are
significant better than overall mean. My thought about this is using mean
separation test, or multiple comparisons (using adjustments: Bonferroni,
Scheffe, Tukey). I tried searching for a while in R but didn't get enough
information. So can anyone give me some suggestions on useful R functions,
or packages, or research papers which may handle this problem in mixed
effect models? If not available, methods in linear/generalized linear
models without random models may help me too. Thanks a lot in advance!

Regards,

Mean separation test in lmer/glmer models

Thread (3 messages)