|
Activity Number:
|
251
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #307458 |
|
Title:
|
Data Analysis under Case-Cohort Designs with Clustered Binary Outcome Data
|
|
Author(s):
|
Shou-En Lu*+ and Yong Lin and Joanna H. Shih
|
|
Companies:
|
University of Medicine & Dentistry of New Jersey and University of Medicine & Dentistry of New Jersey and National Cancer Institute
|
|
Address:
|
683 Hoes Lane, W., Piscataway, NJ, 08854,
|
|
Keywords:
|
case-cohort design ; estimating function ; intra-cluster association ; binary outcomes
|
|
Abstract:
|
Cluster case-cohort design (Lu and Shih 2006) is a cost effective design that incorporates cluster structure in a cohort into the sampling frame. In contrast to assuming all cohort members to be independent, it accounts for the dependency between cluster members and extends the well-known case-cohort design of Prentice (1986) which only applies for cohorts of independent individuals. This paper aims to develop a statistical methodology for analyzing binary outcomes under the cluster case-cohort design. Statistical inference is developed to estimate the regression parameters in the marginal logistic regression model and the intra-cluster association. Statistical properties of the proposed estimators are developed. The performance and statistical efficiencies of the proposed estimators are investigated with simulation studies. A data example is used to illustrate the proposed methodology.
|
