The determination of sample sizes in the comparison of two multinomial proportions from ordered categories

Myoung Keun Lee, Hae Hiang Song, Seung Ho Kang, Chul W. Ahn

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

We consider sample size determination for ordered categorical data when the alternative assumption is the proportional odds model. In this paper the sample size formula proposed by WHITEHEAD (Statistics in Medicine, 12, 2257-2271, 1993) is compared with the methods based on exact and asymptotic linear rank tests with Wilcoxon and trend scores. We show that Whitehead's formula, which is based on a normal approximation, works well when the sample size is moderate to large but recommend the exact method with Wilcoxon scores for small sample sizes. The consequences of misspecification in models are also investigated.

Original languageEnglish
Pages (from-to)395-409
Number of pages15
JournalBiometrical Journal
Volume44
Issue number4
DOIs
Publication statusPublished - 2002 Aug 19

Fingerprint

Ordered Categories
Sample Size
Proportion
Linear Rank Test
Ordered Categorical Data
Proportional Odds Model
Sample Size Determination
Normal Approximation
Misspecification
Exact Method
Small Sample Size
Medicine
Statistics
Alternatives
Sample size
Model
Trends

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Cite this

Lee, Myoung Keun ; Song, Hae Hiang ; Kang, Seung Ho ; Ahn, Chul W. / The determination of sample sizes in the comparison of two multinomial proportions from ordered categories. In: Biometrical Journal. 2002 ; Vol. 44, No. 4. pp. 395-409.
@article{a9b43c0a105e446bb3b87c382845e5c7,
title = "The determination of sample sizes in the comparison of two multinomial proportions from ordered categories",
abstract = "We consider sample size determination for ordered categorical data when the alternative assumption is the proportional odds model. In this paper the sample size formula proposed by WHITEHEAD (Statistics in Medicine, 12, 2257-2271, 1993) is compared with the methods based on exact and asymptotic linear rank tests with Wilcoxon and trend scores. We show that Whitehead's formula, which is based on a normal approximation, works well when the sample size is moderate to large but recommend the exact method with Wilcoxon scores for small sample sizes. The consequences of misspecification in models are also investigated.",
author = "Lee, {Myoung Keun} and Song, {Hae Hiang} and Kang, {Seung Ho} and Ahn, {Chul W.}",
year = "2002",
month = "8",
day = "19",
doi = "10.1002/1521-4036(200206)44:4<395::AID-BIMJ395>3.0.CO;2-D",
language = "English",
volume = "44",
pages = "395--409",
journal = "Biometrical Journal",
issn = "0323-3847",
publisher = "Wiley-VCH Verlag",
number = "4",

}

The determination of sample sizes in the comparison of two multinomial proportions from ordered categories. / Lee, Myoung Keun; Song, Hae Hiang; Kang, Seung Ho; Ahn, Chul W.

In: Biometrical Journal, Vol. 44, No. 4, 19.08.2002, p. 395-409.

Research output: Contribution to journalArticle

TY - JOUR

T1 - The determination of sample sizes in the comparison of two multinomial proportions from ordered categories

AU - Lee, Myoung Keun

AU - Song, Hae Hiang

AU - Kang, Seung Ho

AU - Ahn, Chul W.

PY - 2002/8/19

Y1 - 2002/8/19

N2 - We consider sample size determination for ordered categorical data when the alternative assumption is the proportional odds model. In this paper the sample size formula proposed by WHITEHEAD (Statistics in Medicine, 12, 2257-2271, 1993) is compared with the methods based on exact and asymptotic linear rank tests with Wilcoxon and trend scores. We show that Whitehead's formula, which is based on a normal approximation, works well when the sample size is moderate to large but recommend the exact method with Wilcoxon scores for small sample sizes. The consequences of misspecification in models are also investigated.

AB - We consider sample size determination for ordered categorical data when the alternative assumption is the proportional odds model. In this paper the sample size formula proposed by WHITEHEAD (Statistics in Medicine, 12, 2257-2271, 1993) is compared with the methods based on exact and asymptotic linear rank tests with Wilcoxon and trend scores. We show that Whitehead's formula, which is based on a normal approximation, works well when the sample size is moderate to large but recommend the exact method with Wilcoxon scores for small sample sizes. The consequences of misspecification in models are also investigated.

UR - http://www.scopus.com/inward/record.url?scp=0035998046&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035998046&partnerID=8YFLogxK

U2 - 10.1002/1521-4036(200206)44:4<395::AID-BIMJ395>3.0.CO;2-D

DO - 10.1002/1521-4036(200206)44:4<395::AID-BIMJ395>3.0.CO;2-D

M3 - Article

VL - 44

SP - 395

EP - 409

JO - Biometrical Journal

JF - Biometrical Journal

SN - 0323-3847

IS - 4

ER -