关于Umesh-Loughin-Scherer统计量的一阶Rao-Scott校正

On the first-order Rao-Scott correction of the Umesh-Loughin-Scherer statistic.

作者信息

Bilder C R, Loughin T M

机构信息

Department of Statistics, Oklahoma State University, Stillwater 74078, USA.

出版信息

Biometrics. 2001 Dec;57(4):1253-5. doi: 10.1111/j.0006-341x.2001.01253.x.

DOI:10.1111/j.0006-341x.2001.01253.x

PMID:11764268

Abstract

Decady and Thomas (2000, Biometrics 56, 893-896) propose a first-order corrected Umesh-Loughin-Scherer statistic to test for association in an r x c contingency table with multiple column responses. Agresti and Liu (1999, Biometrics 55, 936-943) point out that such statistics are not invariant to the arbitrary designation of a zero or one to a positive response. This paper shows that, in addition, the proposed testing procedure does not hold the correct size when there are strong pairwise associations between responses.

摘要

迪卡迪和托马斯（2000年，《生物统计学》56卷，第893 - 896页）提出了一种一阶校正的乌梅什 - 洛金 - 谢勒统计量，用于检验具有多个列响应的r×c列联表中的关联性。阿格雷斯蒂和刘（1999年，《生物统计学》55卷，第936 - 943页）指出，这类统计量对于将正响应任意指定为零或一并不具有不变性。本文还表明，当响应之间存在强成对关联时，所提出的检验程序并不具有正确的显著性水平。