|
Activity Number:
|
485
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 7, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #300869 |
|
Title:
|
A Parametric Method for Testing Proportional Odds Assumption
|
|
Author(s):
|
Qing Xu*+ and Jong-Hyeon Jeong+
|
|
Companies:
|
U.S. Food and Drug Administration and University of Pittsburgh
|
|
Address:
|
10903 New Hampshire Avenue,Bldg. #22, Room1213, Silver Spirng, , MD, 20993-0002, Department of Biostatistics,School of Public Health, Pittsburgh, , PA, 15261,
|
|
Keywords:
|
proportional odds ; survival data ; parametric method
|
|
Abstract:
|
We develop a parametric method for detecting the validity of the proportional odds assumption between two groups for survival data. This study was based on the premise that the test procedures taking advantage of knowledge of the distributional information about the data will improve the sensitivity of an existing nonparametric test method. In the proposed method, a parametric form of the weight function is supplemented into an existing nonparametric test statistic for testing the proportional odds assumption. We evaluate type I error probabilities and power of the new method by using the simulated survival data following the log-logistic distribution. The results indicate that the parametric test performs with a higher sensitivity than the existing one, as expected, providing an alternative method of testing the proportional odds assumption often encountered in clinical trial data.
|
- 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 |
