[R-sig-ME] Population fit with glm works fine: totally off with glmer

[Dieter Menne]

Ben Bolker <bbolker at ...> writes:

>   This (using the aforementioned Dorie package)
> 
> library(blme)
> (fitbglmer <-   summary(g3 <- bglmer(Satiated~MealVol*Group+(1|Subject), 
>                          family=binomial, data=sdata)))
> 
> ## requires LATEST version of coefplot2 from r-forge:
> ##  packages won't be rebuilt until tomorrow, probably
> library(coefplot2)
> coefplot2(list(gf1,gf2,gf3),col=c(1,2,4))
> coefplot2(list(gf1,gf2,gf3),xlim=c(-0.05,0.15))
> 

Thanks, Ben, I had already started with MCMCglmm, but did not know about blme.

Dieter