hello,
I have a question about how to specify a Bugs model for several
responses, some of them binary and some continuous, which are correlated
through presence of underlying normal random effects. All what I can
figure out is a (not very elegant) artificial fix which adds a small
source of normal variability for Y4 in the step
Y4[i]~dnorm(mu[i],1.0E3).
That is:
model
{
for(i in 1:N)
{
Y1[i]~dbern(p1[i])
Y2[i]~dbern(p2[i])
Y3[i]~dbern(p3[i])
Y4[i]~dnorm(mu[i],1.0E3)
e[i,1:4]~dmnorm(null[1:4],e.tau[1:4,1:4])
logit(p1[i])~B1.0+B1.1*x[i]
logit(p2[i])~B2.0+B2.1*x[i]
logit(p3[i])~B3.0+B3.1*x[i]
mu[i]~B4.0+B4.1*x[i]
}
...
priors on B's, e.tau
and some other stuff
}
Ideally, I would like to have something like Y4[i]<-mu[i]+e[i,4], which
I think, is not permissible.
Thanks in advance for suggestions how to solve this problem (probably
trivial for anybody more experienced than I am).
Marek Brabec
-------------------------------------------------------------------
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
