[R-sig-ME] ICC for glmer Poisson

[Paul Johnson]

Hi,

Have a look at 

Nakagawa, Shinichi, and Holger Schielzeth. 2010. “Repeatability for Gaussian and Non-Gaussian Data: A Practical Guide for Biologists.” Biological Reviews of the Cambridge Philosophical Society 85 (4): 935–56. doi:10.1111/j.1469-185X.2010.00141.x.
dx.doi.org/10.1111/j.1469-185X.2010.00141.x

and the related rptR package:

Stoffel, M., Nakagawa, S. & Schielzeth, H. (2017) rptR: Repeatability estimation and variance decomposition by generalized linear mixed-effects models.. Methods Ecol Evol. Accepted Author Manuscript. doi:10.1111/2041-210X.12797

There are various ways to get CIs for variances, the easiest is applying confint() to a glmer fit (in fact I think this gives CIs for the SDs). For getting confidence (credible) intervals from variance estimates though I think MCMC (e.g. MCMCglmm) works best.
For testing random effects, see:
http://bbolker.github.io/mixedmodels-misc/glmmFAQ.html#testing-significance-of-random-effects

Good luck,
Paul


> On 2 Feb 2018, at 17:31, dani <orchidn at live.com> wrote:
> 
> Hello everyone,
> 
> 
> I am looking for information regarding the calculation of the ICC for a Poisson Glmer model.
> 
> I remember seeing somewhere that the within group variance is considered to be 1 in the calculation of the ICC for Poisson, but I do not have a reference for that. I did some research online but I could not find much.
> 
> Any help would be very appreciated.
> 
> 
> Also, is there a way to determine significance and CIs for variances? I was thinking that variance can be tested by comparing a Glm model (without random effects) and the Glmer model (with random effects) using anova(), but I have no idea how to calculate the CIs.
> 
> 
> Thank you all!
> 
> MD
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models