Dear all, One aim of my study is to assess the effect of temporal variation in habitat availability on the number of male singing per point count (p). To this end, 60 point counts were randomly distributed in the study area, but these points were fixed over the months (I recorded the number of singing males at the same points but at successive months). At each point, I recorded the cover of habitat available (e.g., shrubs, cereals, alfalfa...etc). I'm wondering if I should consider the ID as a random factor in this case? If yes, is this command correct? model=glmer(p~cover of habitat + (1|month/ID), famlily=poisson, data=data) Thank you very much for your help. Regards --- Sa?d Hanane, PhD Service d'?cologie, de Biodiversit? et de Conservation des Sols Centre de Recherche Foresti?re Chariae Omar Ibn Al Khattab, BP 763, Rabat-Agdal/Maroc.
GLMM replicated data
2 messages · Saâd HANANE, Mollie Brooks
You probably want crossed random effects of month and point ID: model=glmer(p~cover of habitat + (1|month)+(1|ID), famlily=poisson, data=data) Does your variable "cover of habitat" change over the months to possibly capture the effect of temporal variation in habitat availability that you want? If "cover of habitat" is constant through time, then all of the temporal variation will be represented in your random effect of month. If "cover of habitat" changes through time, then it will include part of the temporal variation, but some will still go into the random effect of month. Using the Poisson distribution is an assumption that is usually violated in ecology. After fitting the model, you could check for over- or underdispersion and then consider a negative binomial distribution if it?s overdispersed or a Conway Maxwell Poisson if it?s underdispersed. I wrote about dispersion issues in this paper: Brooks, M. E., K. Kristensen, M. R. Darrigo, P. Rubim, M. Uriarte, E. Bruna, and B. M. Bolker. 2019. Statistical modeling of patterns in annual reproductive rates. Ecology 00(00):e02706. 10.1002/ecy.2706. Let me know if you want a copy. Cheers, Mollie
On 13 Dec 2021, at 16.49, Sa?d HANANE <sdhan333 at gmail.com> wrote: Dear all, One aim of my study is to assess the effect of temporal variation in habitat availability on the number of male singing per point count (p). To this end, 60 point counts were randomly distributed in the study area, but these points were fixed over the months (I recorded the number of singing males at the same points but at successive months). At each point, I recorded the cover of habitat available (e.g., shrubs, cereals, alfalfa...etc). I'm wondering if I should consider the ID as a random factor in this case? If yes, is this command correct? model=glmer(p~cover of habitat + (1|month/ID), famlily=poisson, data=data) Thank you very much for your help. Regards --- Sa?d Hanane, PhD Service d'?cologie, de Biodiversit? et de Conservation des Sols Centre de Recherche Foresti?re Chariae Omar Ibn Al Khattab, BP 763, Rabat-Agdal/Maroc. [[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models