Dear Jarrod,
recently you suggested that to overcome overdispersion in a binomial
distribution by fitting a observation-level random effect.
I was wondering if this method has been described anywhere so I can put a
reference in my report.
Thanks a lot,
Greetings David Jansen
Hi Thierry + na
Hi Thierry + nameless,
It is not necessary to expand the binomial into Bernoulli trials (nor
advisable if n and/or the binomial size are large). You can just fit
observation-level random effects:
dataset$resid<-as.factor(1:dim(dataset)[1])
fit3 <- glmer(cbind(male_chick_no, female_chick_no
2 matches
Mail list logo
