A Bayesian method for combining statistical tests

被引：4
|
作者：
Koziol, JA
Tuckwell, HC
机构：
[1] Scripps Res Inst, Dept Mol & Expt Med, La Jolla, CA 92037 USA
[2] Univ Paris 06, INSERM, U444, F-75012 Paris, France
来源：
JOURNAL OF STATISTICAL PLANNING AND INFERENCE | 1999年 / 78卷 / 1-2期
关键词：
combining independent p-values; Fisher's combination procedure; meta-analysis;
D O I：
10.1016/S0378-3758(98)00222-5
中图分类号：
O21 [概率论与数理统计]; C8 [统计学];
学科分类号：
020208 ; 070103 ; 0714 ;
摘要：
Within the context of Liptak's (1958, Magyar. Tud. Akad. Mat. Kutato Int. Kozl. 3, 171-197) formulation of the problem of combining independent p-values, a class of Bayes tests is constructed. Fisher's well-known combination procedure, based on the product of the p-values, is found to be a Bayes test in this setting with a noninformative prior. Good's weighted version of Fisher's procedure is shown to be an excellent approximation to other Bayes tests. (C) 1999 Elsevier Science B.V. All rights reserved.
引用
收藏
页码：317 / 323
页数：7
相关论文
共 50 条
12345