Activity Number:
|
110
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 12, 2002 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Biometrics Section*
|
Abstract - #300569 |
Title:
|
Analysis of Cross Sectional Cluster Ranomized Designs With the Logistic-Normal Likelihood Ratio Test
|
Author(s):
|
Rafael Diaz*+
|
Affiliation(s):
|
Tulane University
|
Address:
|
2114 Burdette Street, New Orleans, Lousiana, 70118, USA
|
Keywords:
|
Groups of proportions ; logistic-normal ; test ; population average ; cluster-specific
|
Abstract:
|
The Logistic-normal likelihood ratio test has been proposed among the parametric techniques to compare two groups of proportions. However, it may be difficult to interpret intervention effects when using this test for cross-sectional cluster-randomized trials. This difficulty is due to the fact that the estimates in the Logistic-normal likelihood ratio test represent cluster-specific effects as opposed to population-average effects. The ready availability of software where this test has been implemented, and the optimal statistical properties of parametric techniques when the sample sizes are large enough, are two reasons to examine the usefulness of this cluster-specific technique for cross-sectional cluster randomized trials. This study examines two features of this test for this type of trial: its bias in estimating an intervention effect, and its power to detect one through Monte Carlo simulations for effect-sizes of interest.
|
- 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 2002 program |