|
Activity Number:
|
338
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #301193 |
|
Title:
|
Simulation Studies of the Effects of Sample Size, and Stratification Factors on Performance of the Fisher's Exact and Cochran-Mantel-Haenszel Tests
|
|
Author(s):
|
David Carlin*+ and Harry Yang
|
|
Companies:
|
MedImmune, Inc. and MedImmune, Inc.
|
|
Address:
|
35 W. Watkins Mill Rd., Gaithersburg, MD, 20878,
|
|
Keywords:
|
simulation ; Fisher's exact ; Cochran-Mantel-Haenszel ; sample size ; power
|
|
Abstract:
|
Sample size determination, selections of stratification factors and statistical test methods are important steps in planning a clinical trial. Since the Cochran-Mantel-Haenszel test is based on an asymptotic distribution requiring large sample size, the Fisher's exact test normally serves as an alternative method when the sample size is small. However, it does not allow for adjustment of mean and variance estimation based on stratification factors. Therefore, it may decrease the chance for detecting treatment effect. Presented in this paper are results of a simulation study conducted to evaluate the impact of sample size and stratification factors on the performance of the Fisher's exact and Cochran-Mantel-Haenszel (CMH) tests. The study demonstrates that even for a moderate sample size (n>=60), CMH test is more powerful in detecting differences. Also included are several applications of simulation methods to determine sample size not readily calculated by commercially available software packages, and develop stochastic process control criteria.
|
- 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 2004 program |
