|
Activity Number:
|
180
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 7, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #305309 |
|
Title:
|
Mixture Models and Informative Priors for Analyzing Incomplete Longitudinal Data
|
|
Author(s):
|
Joseph W. Hogan*+
|
|
Companies:
|
Brown University
|
|
Address:
|
Center for Statistical Sciences, Providence, RI, 02912,
|
|
Keywords:
|
repeated binary data ; semiparametric models ; varying coefficient models ; nonidentifiability ; nonignorable nonresponse ; HIV/AIDS
|
|
Abstract:
|
Mixture models provide an attractive framework for analyzing incomplete data because the parameter space frequently can be divided into one set of parameters indexing Y(obs) and another indexing Y(mis), given Y(obs). A subset of the parameters indexing [Y(mis)|Y(obs)] are completely nonidentifiable. We give examples of principled use of informative priors for capturing assumptions about the missing data mechanisms, including uncertainty about the assumptions. Examples use data from a longitudinal HIV cohort study and a longitudinal clinical trial of smoking cessation.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
