Comparing mixed models

ZoeC

New Member
#1
Hi, I am trying to compare mixed effects models for a conservation project (R script below). I have run a lmer (m4) with all my variables included (some are scale, some are categorical). My model 'm5' has one of the variables from 'm4' omitted and I want to compare the 2 models to see if the dropped variable was significant. My questions are:

1. Is the R message *refitting model(s) with ML (instead of REML)* of any significance?
2. Comparing the 2 models using anova - have I done this correctly? I'm confused to see negative AIC values. Do I compare the two AIC values (-26.007 and -47.716)?
3. I think I need to use an F test (not chi squared) - is this appropriate and if so how do I ask R to do that within the anova?

Many thanks. :)

> m4<-lmer(fr_per_fl~ele+prox+sl+veg+h+trunk+can+gap+th+(1|orc), DataT)

> AIC(m4)
[1] 43.0984

> m5<-lmer(fr_per_fl~ele+prox+sl+veg+h+trunk+can+gap+(1|orc),
DataT)

> AIC(m5)
[1] 40.54589

> anova(m5,m4)
*refitting model(s) with ML (instead of REML)*
Data: DataT

Models:

m5: fr_per_fl ~ ele + prox + sl + veg + h + trunk + can + gap + (1 |
m5: orc)
m4: fr_per_fl ~ ele + prox + sl + veg + h + trunk + can + gap + th +
m4: (1 | orc)

Df AIC BIC logLik deviance Chisq Chi Df Pr(>Chisq)
m5 20 -26.007 -5.1163 33.003 -66.007
m4 21 -47.716 -25.7809 44.858 -89.716 23.709 1 1.121e-06 ***
---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1