Limiting conditional distribution for tests of independence in the two way table

Seung Ho Kang, Jerome Klotz

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

For testing the hypothesis of independence in the r x c two way contingency table, we consider the likelihood ratio statistic -2 ln(λ) and the χ2 statistic conditionally given the sufficient statistic of row and column margin totals. We use Stirling's formula in the conditional characteristic function of normalized cell counts to give a direct proof that the limiting null conditional distribution is multivariate normal Z : script Nrc(0, Λ). We then establish the two statistics have a limiting conditional distribution equivalent to ZT Λ-Z, which has a central χv2, distribution with degrees of freedom v = (r - 1)(c - 1) = rank(Λ).

Original languageEnglish
Pages (from-to)2075-2082
Number of pages8
JournalCommunications in Statistics - Theory and Methods
Volume27
Issue number8
DOIs
Publication statusPublished - 1998 Jan 1

Fingerprint

Test of Independence
Conditional Distribution
Limiting Distribution
Table
Stirling's formula
Sufficient Statistics
Likelihood Ratio Statistic
Multivariate Normal
Null Distribution
Contingency Table
Characteristic Function
Margin
Statistic
Count
Limiting
Degree of freedom
Statistics
Testing
Cell
Independence

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Cite this

@article{6912076c2a6b4fa999e51f13c1d28fe1,
title = "Limiting conditional distribution for tests of independence in the two way table",
abstract = "For testing the hypothesis of independence in the r x c two way contingency table, we consider the likelihood ratio statistic -2 ln(λ) and the χ2 statistic conditionally given the sufficient statistic of row and column margin totals. We use Stirling's formula in the conditional characteristic function of normalized cell counts to give a direct proof that the limiting null conditional distribution is multivariate normal Z : script Nrc(0, Λ). We then establish the two statistics have a limiting conditional distribution equivalent to ZT Λ-Z, which has a central χv2, distribution with degrees of freedom v = (r - 1)(c - 1) = rank(Λ).",
author = "Kang, {Seung Ho} and Jerome Klotz",
year = "1998",
month = "1",
day = "1",
doi = "10.1080/03610929808832210",
language = "English",
volume = "27",
pages = "2075--2082",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "8",

}

Limiting conditional distribution for tests of independence in the two way table. / Kang, Seung Ho; Klotz, Jerome.

In: Communications in Statistics - Theory and Methods, Vol. 27, No. 8, 01.01.1998, p. 2075-2082.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Limiting conditional distribution for tests of independence in the two way table

AU - Kang, Seung Ho

AU - Klotz, Jerome

PY - 1998/1/1

Y1 - 1998/1/1

N2 - For testing the hypothesis of independence in the r x c two way contingency table, we consider the likelihood ratio statistic -2 ln(λ) and the χ2 statistic conditionally given the sufficient statistic of row and column margin totals. We use Stirling's formula in the conditional characteristic function of normalized cell counts to give a direct proof that the limiting null conditional distribution is multivariate normal Z : script Nrc(0, Λ). We then establish the two statistics have a limiting conditional distribution equivalent to ZT Λ-Z, which has a central χv2, distribution with degrees of freedom v = (r - 1)(c - 1) = rank(Λ).

AB - For testing the hypothesis of independence in the r x c two way contingency table, we consider the likelihood ratio statistic -2 ln(λ) and the χ2 statistic conditionally given the sufficient statistic of row and column margin totals. We use Stirling's formula in the conditional characteristic function of normalized cell counts to give a direct proof that the limiting null conditional distribution is multivariate normal Z : script Nrc(0, Λ). We then establish the two statistics have a limiting conditional distribution equivalent to ZT Λ-Z, which has a central χv2, distribution with degrees of freedom v = (r - 1)(c - 1) = rank(Λ).

UR - http://www.scopus.com/inward/record.url?scp=0041972057&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0041972057&partnerID=8YFLogxK

U2 - 10.1080/03610929808832210

DO - 10.1080/03610929808832210

M3 - Article

VL - 27

SP - 2075

EP - 2082

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 8

ER -