|
Activity Number:
|
469
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305838 |
|
Title:
|
Testing Equality of Medians in Two Independent Lognormal Distributions
|
|
Author(s):
|
Hongwei Wang*+ and Arvind K. Shah
|
|
Companies:
|
Merck & Co., Inc. and Merck & Co., Inc.
|
|
Address:
|
Merck Research Lab, Rahway, NJ, 07065,
|
|
Keywords:
|
lognormal distribution ; median ; confidence interval ; coverage ; testing equality
|
|
Abstract:
|
The distribution of some biomarkers follows lognormal distribution, where median is more appropriate measure for making inference. We are testing the equality of medians through the confidence intervals (CIs) on the difference in the medians of two independent lognormally distributed variables. The performance of six CIs is evaluated using simulated as well as actual data. The first is based on log-transformation and Delta method, the next two are based on back-transforming the limits of a CI on the difference of means on a log scale, the fourth is based on the asymptotic property of the median itself, and the fifth and sixth are based on the inter-quartile range and Hodges-Lehman estimator respectively. The average width and coverage of these confidence intervals are reported for varying mean, variance, and sample size. Some recommendations based on the results are also given.
|
- 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 2006 program |
