Nonparametric Tests for the One-Sided Alternatives

• 발행기관: 한국자료분석학회
• 저자명: Hyo-Il Park
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.13 No.1, 63~70쪽, 전체 8쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2011.02.28

영문 초록

In this study, we propose nonparametric test procedures for the one-sided alternatives using a group of quantile test statistics. This approach for the construction of nonparametric test may be useful when the underlying distributions are completely unknown. We consider maximal and summing types as the combining functions for the quantile statistics. Then we derive the limiting distributions for the proposed test statistics. Also we comment for using the supremum type of statistics. Then we illustrate our procedure with an example and compare the proposed tests with the Kolmogorov-Smirnov test by obtaining empirical powers through simulation study. For obtaining the numerical values for the example and simulation study, we apply the permutation principle.

목차

1. Introduction

2. One-sided tests

3. An example and simulation study

References

키워드

Nonparametric test Permutation principle Quantile test Two-sample problem