Variance explained by random factor

Many apologise but the glm model I compared was ma not ma1 and thus did have the interaction term:

ma<-glm(RoundedOverlap~sess+breedfem+sess:breedfem ,family=poisson,data=Male)
mixed<-lmer(RoundedOverlap~sess+breedfem+sess:breedfem+(1|Site),family=poisson,data=Male)



-----Original Message-----
From: dmbates at gmail.com [mailto:dmbates at gmail.com] On Behalf Of Douglas Bates
Sent: 14 August 2008 10:22
To: Ken Beath
Cc: Renwick, A. R.; r-sig-mixed-models at r-project.org
Subject: Re: [R-sig-ME] Variance explained by random factor

On 14/08/2008, at 1:17 AM, Renwick, A. R. wrote:

I am currently trying to run a lmer model with poisson distrubution.
I tested the model with a model without the random effect and it
inferred that I should include the random effect:

ma1<-glm(RoundedOverlap~sess+breedfem,family=poisson,data=Male)

mixed<-lmer(RoundedOverlap~sess+breedfem+sess:breedfem+(1|Site),famil
y=poisson,data=Male)

#test to see if sig difference between glm and glmm
as.numeric(2*(logLik(mixed)-logLik(ma)))
#99.16136
pchisq(99.16136,1,lower=FALSE)
#2.327441e-23  so should use a GLMM

The problem may be due to the random effects model containing an
interaction term sess:breedfem that the glm doesn't.

I agree.  The result from the likelihood ratio test is actually evaluating the significance of the interaction term, not the random effects term.

_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

The University of Aberdeen is a charity registered in Scotland, No SC013683.