|
Activity Number:
|
548
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #307289 |
|
Title:
|
Hochberg's Step-Up Method: Cutting Corners off Holm's Step-Down Method
|
|
Author(s):
|
Yifan Huang*+ and Jason Hsu
|
|
Companies:
|
H. Lee Moffitt Cancer Center & Research Institute and The Ohio State University
|
|
Address:
|
12902 Magnolia Drive, Tampa, FL, 33612,
|
|
Keywords:
|
Hochberg's method ; Holm's method ; multiple testing ; step-up test ; step-down test ; partition testing
|
|
Abstract:
|
Holm's method and Hochberg's methods are popular multiple tests. They are viewed as step-down and step-up versions of the Bonferroni test. We will present our insight that both are special cases of partition testing, and the difference is while Holm's method tests each partition hypothesis using the largest order statistic setting critical value based on the Bonferroni inequality, Hochberg's method tests each partition hypothesis using all the order statistics setting a series of critical values based on Simes' inequality. Geometrically, Hochberg's step-up method "cuts corners" off the acceptance regions of Holm's step-down method by making assumption on the joint distribution of the test statistics. Partition testing making use of the joint distribution of the test statistics is more powerful than partition testing using probabilistic inequalities, thus is recommended.
|
