TESTING WHETHER THE SURVIVAL DISTRIBUTION IS NEW BETTER THAN USED OF SPECIFIED AGE
成果类型：
Article
署名作者：
EBRAHIMI, N; HABIBULLAH, M
署名单位：
University of Wisconsin System
刊物名称：
BIOMETRIKA
ISSN/ISSBN：
0006-3444
DOI：
10.1093/biomet/77.1.212
发表日期：
1990
页码：
212215
关键词：
摘要：
The cumulative distribution function F, of a nonnegative random variable, is said to be new better than used of specified age t0 if .hivin.F (x + t0) .ltoreq. .hivin.F (x) .hivin.F (t0) for all .gtoreq. 0, where .hivin.F = 1 - F. Let A = {F:.hivin.F (x = t0) = .hivin.F (x) .hivin.F (t0), for all x .gtoreq. 0}. We propose a class of test statistics Tk, indexed by a positive integer k, for testing the null hypothesis that F belongs A against the alternative hypothesis that F is new better than used of specificed age t0 but not in A. We compare the performance of our proposed test statistics with the test given by Hollander, Park and Proschan (1986). The proposed tests perform better than their test for some alternatives.