Testing hypotheses concerning unions of linear subspaces

Roger L. Berger, Dennis F. Sinclair

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

The likelihood ratio test (LRT) for hypotheses concerning unions of linear subspaces is derived for the normal theory linear model. A more powerful test, an intersection-union test, is proposed for the case in which the subspaces are not all of the same dimension. A theorem is proved that may be used to identify hypotheses that concern unions of linear subspaces. Some hypotheses about the spacings between normal means are shown to concern unions of linear subspaces and therefore can be tested using the LRT. Finally, the computation of the LRT statistic is discussed.

Original languageEnglish (US)
Pages (from-to)158-163
Number of pages6
JournalJournal of the American Statistical Association
Volume79
Issue number385
DOIs
StatePublished - 1984
Externally publishedYes

Fingerprint

Testing Hypotheses
Union
Subspace
Likelihood Ratio Test
Likelihood Ratio Test Statistic
Spacing
Linear Model
Intersection
Hypothesis testing
Likelihood ratio test
Theorem

Keywords

  • Intersection-union test
  • Likelihood ratio test
  • Linear model
  • Ordered means
  • Spacings between means

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Testing hypotheses concerning unions of linear subspaces. / Berger, Roger L.; Sinclair, Dennis F.

In: Journal of the American Statistical Association, Vol. 79, No. 385, 1984, p. 158-163.

Research output: Contribution to journalArticle

Berger, Roger L. ; Sinclair, Dennis F. / Testing hypotheses concerning unions of linear subspaces. In: Journal of the American Statistical Association. 1984 ; Vol. 79, No. 385. pp. 158-163.
@article{c8df49e540214ef981ad0205572d91b4,
title = "Testing hypotheses concerning unions of linear subspaces",
abstract = "The likelihood ratio test (LRT) for hypotheses concerning unions of linear subspaces is derived for the normal theory linear model. A more powerful test, an intersection-union test, is proposed for the case in which the subspaces are not all of the same dimension. A theorem is proved that may be used to identify hypotheses that concern unions of linear subspaces. Some hypotheses about the spacings between normal means are shown to concern unions of linear subspaces and therefore can be tested using the LRT. Finally, the computation of the LRT statistic is discussed.",
keywords = "Intersection-union test, Likelihood ratio test, Linear model, Ordered means, Spacings between means",
author = "Berger, {Roger L.} and Sinclair, {Dennis F.}",
year = "1984",
doi = "10.1080/01621459.1984.10477079",
language = "English (US)",
volume = "79",
pages = "158--163",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "385",

}

TY - JOUR

T1 - Testing hypotheses concerning unions of linear subspaces

AU - Berger, Roger L.

AU - Sinclair, Dennis F.

PY - 1984

Y1 - 1984

N2 - The likelihood ratio test (LRT) for hypotheses concerning unions of linear subspaces is derived for the normal theory linear model. A more powerful test, an intersection-union test, is proposed for the case in which the subspaces are not all of the same dimension. A theorem is proved that may be used to identify hypotheses that concern unions of linear subspaces. Some hypotheses about the spacings between normal means are shown to concern unions of linear subspaces and therefore can be tested using the LRT. Finally, the computation of the LRT statistic is discussed.

AB - The likelihood ratio test (LRT) for hypotheses concerning unions of linear subspaces is derived for the normal theory linear model. A more powerful test, an intersection-union test, is proposed for the case in which the subspaces are not all of the same dimension. A theorem is proved that may be used to identify hypotheses that concern unions of linear subspaces. Some hypotheses about the spacings between normal means are shown to concern unions of linear subspaces and therefore can be tested using the LRT. Finally, the computation of the LRT statistic is discussed.

KW - Intersection-union test

KW - Likelihood ratio test

KW - Linear model

KW - Ordered means

KW - Spacings between means

UR - http://www.scopus.com/inward/record.url?scp=0010597073&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0010597073&partnerID=8YFLogxK

U2 - 10.1080/01621459.1984.10477079

DO - 10.1080/01621459.1984.10477079

M3 - Article

AN - SCOPUS:0010597073

VL - 79

SP - 158

EP - 163

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 385

ER -