[R-sig-ME] effect sizes in lmer

[David Winsemius]

On Oct 12, 2013, at 1:01 PM, Joshua Hartshorne wrote:

> Are there any good methods for determining effect size for lmer? I am using
> anova(mod1,mod2) for my test statistics, so it would be great to get effect
> sizes out of that. In a 2011 discussion here, it was suggested that the
> log-likelihood ratio test is superior to effect size measures, so maybe
> even just a citation that makes that point would work?

You should read that discussion again to see if you can discern what was the point under discussion especially seeing it you can determine: superior with respect to what purpose?

> Your suggestions would be greatly appreciated. I'm under re-review at a
> journal that requires effect size measures.

The journal's purpose of requiring effect sizes is to prevent authors from focussing solely getting p-values, and then failing to explain the implications of model estimates. Don't you have access to the fixef and ranef functions?

-- 

David Winsemius
Alameda, CA, USA