Nonparametric tests for scale and location

Daniele Compagnone, Manfred Heinz Denker

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We prove asymptotic normality of simple linear rank statistics when the observations have U-statistic structure. This type of statistic can be used for testing the location parameter and/or the scale parameter in a scale-location model. It is shown that the tests for the location parameter are asymptotically efficient under normality using a normal score function and are also as efficient as the t-test when the dimension of the kernel function tends to infinity and when Wilcoxon scores are used. For scaling parameters we propose a generalization of tests by Lehmann and Moses for scale effects. We describe a simulation study which shows that these new tests perform better than the ones previously used.

Original languageEnglish (US)
Pages (from-to)123-154
Number of pages32
JournalJournal of Nonparametric Statistics
Volume7
Issue number2
DOIs
StatePublished - Jan 1 1996

Fingerprint

Non-parametric test
Location Parameter
Linear Rank Statistics
Location-scale Model
Scale Effect
Normal Function
Score Function
U-statistics
t-test
Scale Parameter
Kernel Function
Asymptotic Normality
Normality
Statistic
Infinity
Simulation Study
Scaling
Tend
Testing
Nonparametric test

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Compagnone, Daniele ; Denker, Manfred Heinz. / Nonparametric tests for scale and location. In: Journal of Nonparametric Statistics. 1996 ; Vol. 7, No. 2. pp. 123-154.
@article{9db9598d4e8f4b8281f568545c5834e8,
title = "Nonparametric tests for scale and location",
abstract = "We prove asymptotic normality of simple linear rank statistics when the observations have U-statistic structure. This type of statistic can be used for testing the location parameter and/or the scale parameter in a scale-location model. It is shown that the tests for the location parameter are asymptotically efficient under normality using a normal score function and are also as efficient as the t-test when the dimension of the kernel function tends to infinity and when Wilcoxon scores are used. For scaling parameters we propose a generalization of tests by Lehmann and Moses for scale effects. We describe a simulation study which shows that these new tests perform better than the ones previously used.",
author = "Daniele Compagnone and Denker, {Manfred Heinz}",
year = "1996",
month = "1",
day = "1",
doi = "10.1080/10485259608832695",
language = "English (US)",
volume = "7",
pages = "123--154",
journal = "Journal of Nonparametric Statistics",
issn = "1048-5252",
publisher = "Taylor and Francis Ltd.",
number = "2",

}

Nonparametric tests for scale and location. / Compagnone, Daniele; Denker, Manfred Heinz.

In: Journal of Nonparametric Statistics, Vol. 7, No. 2, 01.01.1996, p. 123-154.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Nonparametric tests for scale and location

AU - Compagnone, Daniele

AU - Denker, Manfred Heinz

PY - 1996/1/1

Y1 - 1996/1/1

N2 - We prove asymptotic normality of simple linear rank statistics when the observations have U-statistic structure. This type of statistic can be used for testing the location parameter and/or the scale parameter in a scale-location model. It is shown that the tests for the location parameter are asymptotically efficient under normality using a normal score function and are also as efficient as the t-test when the dimension of the kernel function tends to infinity and when Wilcoxon scores are used. For scaling parameters we propose a generalization of tests by Lehmann and Moses for scale effects. We describe a simulation study which shows that these new tests perform better than the ones previously used.

AB - We prove asymptotic normality of simple linear rank statistics when the observations have U-statistic structure. This type of statistic can be used for testing the location parameter and/or the scale parameter in a scale-location model. It is shown that the tests for the location parameter are asymptotically efficient under normality using a normal score function and are also as efficient as the t-test when the dimension of the kernel function tends to infinity and when Wilcoxon scores are used. For scaling parameters we propose a generalization of tests by Lehmann and Moses for scale effects. We describe a simulation study which shows that these new tests perform better than the ones previously used.

UR - http://www.scopus.com/inward/record.url?scp=0040172691&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0040172691&partnerID=8YFLogxK

U2 - 10.1080/10485259608832695

DO - 10.1080/10485259608832695

M3 - Article

VL - 7

SP - 123

EP - 154

JO - Journal of Nonparametric Statistics

JF - Journal of Nonparametric Statistics

SN - 1048-5252

IS - 2

ER -