I am using GLLAMM to run a random-intercept logistic regression. I can
find no information on what covariance structure is used (e.g.,
exchangable, identity, autoregressive), or how I can specify a
covariance structure.
Could somebody explain to me whether GLLAMM uses a covariance
structure, and if not, why not?
Below is some additional information on the model I am running:
The model is a repeated mesures study with four timepoints. The
outcome here (harddrug) is dichotomous.
The command I am using is:
xi: gllamm harddrug i.howstab time i.group3, i(id) link(logit)
family(binom) nip(30) adapt nolog eform
harddrug=drug use in last 30 days (0/1)
howstab=level of housing stability (0/1/2)
time=assessment wave (0/1/2/3)
group3=group assignment (0/1/2)
id=subject identifier (n=530)
Thanks for any help you can provide
Brian Weir
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
