|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||||||||
|
|
|
|
||||||||||||||||
|
|
||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||||
|
|
|
|||||||||||||||||
|
|
ABSTRACT Dr. Agnes Hsiung, National Health Research Institute, Taiwan Multiple event-time data appear often in clinical studies to evaluate
treatments for chronic diseases and in longitudinal follow-up studies.
When the sampling unit is a cluster of subjects like in family studies
or for paired design in clinical studies, some frailty variables (
unobservable ) need to be incorporated into the model. In this talk, we
will explore several approaches and apply the proposed methods to real
example data for illustration.
Return to Longitudinal/Survival Working Group List | Return to Home Page |
