Question on overdispersion

Dear Nameless,

The quasi distribution can no longer be used in lme4 because a) the
results were not very reliable b) there is an alternative to model
overdispersion.

The alternative is to expand your dataset to bernoulli trials. Then add
a random effect with one level per observation. This random effect will
model additive overdisperion. The quasi distributions model
overdisperion multiplicative.

In the example below, the random effect of RowID has 0 variance. Hence
no overdispersion.

dataset <- data.frame(
	male_chick_no = c(2,4,1,0,3,5,2), 
	female_chick_no=c(1,0,3,3,1,0,2), 
	FemaleID=c("A","A","B","B","C","D","E"))

longFormat <- do.call(rbind, lapply(seq_len(nrow(dataset)), function(i){
	with(dataset, data.frame(Sex = c(rep("M", male_chick_no[i]),
rep("F", female_chick_no[i])), FemaleID = FemaleID[i]))
}))
longFormat$FemaleID <- factor(longFormat$FemaleID)
longFormat$RowID <- factor(seq_len(nrow(longFormat)))
longFormat$Male <- longFormat$Sex == "M"

library(lme4)
fit1 <- glmer(Male ~ (1|FemaleID), data = longFormat, family = binomial)
fit2 <- glmer(Male ~ (1|FemaleID) + (1|RowID), data = longFormat, family
= binomial)
anova(fit1, fit2)

Best regards,

Thierry

PS sig-mixed-models is a better mailinglist for this kind of questions.

------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek
team Biometrie & Kwaliteitszorg
Gaverstraat 4
9500 Geraardsbergen
Belgium

Research Institute for Nature and Forest
team Biometrics & Quality Assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium

tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be
www.inbo.be

To call in the statistician after the experiment is done may be no more
than asking him to perform a post-mortem examination: he may be able to
say what the experiment died of.
~ Sir Ronald Aylmer Fisher

The plural of anecdote is not data.
~ Roger Brinner

The combination of some data and an aching desire for an answer does not
ensure that a reasonable answer can be extracted from a given body of
data.
~ John Tukey

Question on overdispersion

Thread (4 messages)