In the past I have used two-part models to deal with distributions having a
high proportion of "true" zeroes. I understand that a two-part model has now
been extended to random effects longitudinal models in a way which also
allow the possibility of correlation in the logistic and linear random
coefficients. Tooze et al have developed a SAS macro called MIXCORR to do
this and Muthen and Muthen have implemented a similar model in Mplus. My
understanding is that gllamm could handle such models, but in reviewing the
manual it was not apparent to me how. I have been unable to find this
specific topic in the statalist archives. Has anyone had experience using
gllamm for such models (either with cross-sectional or longitudinal data) or
has anyone adapted the MIXCORR macro to Stata?
Thanks very much.
[The Tooze approach is described in: Tooze, AJ, Grunwald, GK, Jones, RH
(2002). Analysis of repeated measures data with clumping at zero.
Statistical Methods in Medical Research, 11, 341-355.]
Dan Chandler
*******
436 Old Wagon Road
Trinidad, CA 95570
Fax or phone: (707) 677-0895
[email protected]
*
* 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/
