Hello,
I have some concerns that I hope I could get comments on here (eventhough
they are not specific for BUGS):
My first question is if DIC is suitable for 'mixed' effects (or how to
interpret the comparisons). I have read in the book by Gelfand et al. 2003
that the focus of DIC changes between the marginal and the full
conditional model. In a mixed effects model the fixed effects are
intergrated out and could DIC of fixed effect model therefore be compared
with DIC from a mixed effects model (i.e. by adding a random component).
Has anyone any experience with this issue?
My second question concerns overparameterized models and DIC. I have seen
that neither AIC nor BIC is suitable for models where there are more
parameters than data observations, but what about DIC? My example here
concerns a model with a structured correlation matrix between the random
effects (one parameter per observation). What I want to compare is DIC
from a model without the correlation matrix and one with, is this sensible?
Best,
Patrik Waldmann
Department of Forest Genetics and Plant Physiology
Umeå Plant Science Centre
Swedish University of Agricultural Sciences
SE-901 83 Umeå
Sweden
-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.
For help with crashes and error messages, first mail [log in to unmask]
To mail the BUGS list, mail to [log in to unmask]
Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
Please do not mail attachments to the list.
To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
If this fails, mail [log in to unmask], NOT the whole list
|