Dear list members,
When estimating a population averaged (marginal) model with -xtlogit,
pa- I receive the error message "estimates diverging (missing
predictions)". This happens only if I use an unstructured correlation
structure.
I am analysing unbalanced panel data with missing observations for
many individuals. I suppose this could be an explanation. What,
however, would be a suitable solution, switching to an exchangeable
structure? Or shouldn't I use marginal models in unbalanced panels?
Thanks
Thomas
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
