Dear Listers,
Weighted GEE for a longitudinal study with five observations and a binary
outcome
The GEE method for analysing longitudinal data assumes that dropouts are
MCAR. However, I have read somewhere that there is a modified weighted GEE
that can handle MAR-data. Does anyone know how to do this in Stata.
(Another option would be to analyze data with a mixed (random) effects
model. However a mixed effect model would produce subject specific
parameters why these results would not be comparable with results from a
GEE-model that have been reported by others. It would also be nice to
compare the results from the ordinary GEE and the weighted GEE in our field
of work.)
/Tomas
*
* 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/
