Message-ID: <loom.20110418T151845-29@post.gmane.org>
Date: 2011-04-18T13:22:16Z
From: Ben Bolker
Subject: reporting lme results
Pennell, Tanya <tp241 <at> exeter.ac.uk> writes:
> I have used lme4 and I have found a significant result
> when using anova to compare model 1 and model 2 (where I
> took out an interaction).
>
> The result looks like this:
> model.3: DIFFERENCE ~ (1 | MALE.ID)
> model.2: DIFFERENCE ~ MALE.SPECIES + (1 | MALE.ID)
> Df AIC BIC logLik Chisq Chi Df Pr(>Chisq)
> model.3 3 1379.7 1387.1 -686.86
> model.2 4 1374.1 1384.0 -683.05 7.6235 1 0.005761 **
> ---
> Signif. codes: 0 ?***? 0.001 ?**? 0.01 ?*? 0.05 ?.? 0.1 ? ? 1
>
> Just wondering how I would report this in a scientific paper?
>
I would say (for example) that a likelihood ratio test of
the effect of species on the difference was significant
(log-likelihood difference=3.81, p=0.0058).
Or something like that.
Questions like this might be more appropriate on the
r-sig-mixed-models list.