Thank you for your response. Sorry for insisting but I haven’t
understood if the data do not permit to include an individual effect or
if it is just the models I run.
We have several flowers and several fruits per individual plant. So
there is a correlation between the flowers/fruits of one plant. But when
we run the models here after which are apparently adapted for this type
of data (and which should treat the flowers/fruits and not the
individual plants) the number of observations given is equal to the
number of individuals and not to the number of fruits + the number of
flowers.
Do you think that our model is good and that we cannot ask for an
individual effect in our case or do you think that there are some other
functions which permit to take account of the dependence between
flowers/fruits of a same individual?
Thank you by advance
Emmanuelle
y < cbind(indnbfruits,indnbflowers);
model1
<glm(y~red*yellow+I(red^2)+I(yellow^2)+densite8+I(densite8^2)+freq8_4+I
(freq8_4^2), quasibinomial);
model2 <
glmmPQL(y~red*yellow+I(red^2)+I(yellow^2)+densite8+I(densite8^2)+freq8_4
+I(freq8_4^2), random=~1num, quasibinomial);
Emmanuelle TASTARD
UMR 5174 'Evolution et Diversité Biologique'
Université Paul Sabatier Bat 4R3
31062 TOULOUSE CEDEX 9 France
tel : 05 61 55 67 59
Message d'origine
De : rhelpbounces@stat.math.ethz.ch
[mailto:rhelpbounces@stat.math.ethz.ch] De la part de
Lorenz.Gygax@fat.admin.ch
Envoyé : lundi 10 octobre 2005 10:22
À : rhelp@stat.math.ethz.ch
Objet : Re: [R] interpretation output glmmPQL
> We study the effect of several variables on fruit set for 44
> individuals (plants). For each individual, we have the number
> of fruits, the number of flowers and a value for each variable.
> ...
>  Glm does not take account of the correlation between the
> flowers of a unique individual. So we would like to add a
> random effect 'individual' but the model2 (here after) gives an
> output similar to the one of model1 for estimated coefficients
> and pvalues.
> ...
> Does it mean that there is no individual effect or is my model
> not good (number of groups (individuals)=number of observations,
> is it possible?).
If you have only one observation per indiviudal plant, how could there
be
dependence within the plant? This would only make sense if your
observations
were the individual flowers. Data on those could be correlated within
plant
and then a random term for the plant is meaningful.
Cheers, Lorenz

Lorenz Gygax
Centre for proper housing of ruminants and pigs
Swiss Federal Veterinary Office
agroscope FAT Tänikon, CH8356 Ettenhausen / Switzerland
[[alternative HTML version deleted]]
This archive was generated by hypermail 2.1.8 : Sun 23 Oct 2005  18:38:00 EST