|
Activity Number:
|
157
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 9, 2004 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
ENAR
|
| Abstract - #300356 |
|
Title:
|
A Marginalized Pattern-mixture Model for Longitudinal Binary Data with Possibly Nonignorable Nonresponse
|
|
Author(s):
|
Kenneth J. Wilkins*+ and Garrett M. Fitzmaurice
|
|
Companies:
|
Harvard School of Public Health and Harvard School of Public Health
|
|
Address:
|
655 Huntington Ave., Boston, MA, 02115,
|
|
Keywords:
|
binary data ; missing data ; dropout ; nonignorable nonresponse ; longitudinal data
|
|
Abstract:
|
This paper proposes a method for modeling longitudinal binary responses that are subject to nonignorable nonresponse. The approach presumes that the target of inference is the marginal distribution of the response at each occasion and its dependence on covariates, and can handle both monotone nd nonmonotone missingness. The approach involves a marginally specified pattern-mixture model that directly parameterizes both the marginal means at each occasion and the dependence of each response on indicators of nonresponse pattern. This formulation readily incorporates a variety of nonignorable nonresponse processes assumed within a sensitivity analysis. With identifying constraints in place, estimation of model parameters proceeds via solution to a set of modified generalized estimating equations. The proposed method provides an alternative to standard selection and pattern-mixture modeling frameworks, while featuring the advantages of each. The paper concludes with applications of the method to data from two longitudinal studies: a contraceptive clinical trial of dosage with substantial dropout, and a study of obesity in children for which there was intermittent nonresponse.
|
