|
Activity Number:
|
457
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 7, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
ENAR
|
| Abstract - #300271 |
|
Title:
|
Estimation and Inference for Group-Randomized Trials with a Binary Outcome
|
|
Author(s):
|
Sherri L. Pals*+
|
|
Companies:
|
Centers for Disease Control and Prevention
|
|
Address:
|
MS E45, Atlanta, GA, 30333,
|
|
Keywords:
|
Group-randomized trials ; Intraclass correlation ; Binary variables
|
|
Abstract:
|
Group-randomized trials (GRTs) require design and analytic methods that take into account intraclass correlation due to assignment of intact social groups to study conditions. Binary outcome variables in GRTs are common, and investigators have increasingly responded to the challenge of developing methods for these trials. Early GRTs with binary outcomes were often analyzed with two-stage methods, involving analysis of group-level event rates. Generalized estimating equations (GEE) provided a method for taking intraclass correlation into account while avoiding the specification of the joint distribution; recently proposed corrections have addressed GEE's poor performance in small GRTs. The use of mixed-model approaches has increased dramatically due to availability in commercial software. Each of these methods will be reviewed, highlighting recent developments and areas for future work.
|
- 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 2008 program |
