AN EMPIRICAL-ASSESSMENT OF SADDLEPOINT APPROXIMATIONS FOR TESTING A LOGISTIC-REGRESSION PARAMETER

被引：17
作者：
BEDRICK, EJ [1 ]
HILL, JR [1 ]
机构：
[1] EDS RES,ALBUQUERQUE,NM 87109
来源：
BIOMETRICS | 1992年 / 48卷 / 02期
关键词：
BARTLETT TEST; CONDITIONAL TESTS; EDGEWORTH EXPANSIONS; EXACT INFERENCE; TREND TEST;
D O I：
10.2307/2532307
中图分类号：
Q [生物科学];
学科分类号：
07 ; 0710 ; 09 ;
摘要：
Saddlepoint methods provide quick and easy approximations to significance levels for conditional tests of logistic regression parameters. We evaluate the accuracies of saddlepoint approximations for three well-known conditional tests: Bartlett's test for no three-factor interaction in a 2 x 2 x 2 table, the test for trend in a series of probabilities, and the exact test of no association in stratified 2 x 2 tables with a common odds ratio. General recommendations are suggested regarding the use of saddlepoint approximations for exact conditional significance levels.
引用
收藏
页码：529 / 544
页数：16
相关论文
共 30 条
[1]  
Agresti A., 1984, ANAL ORDINAL CATEGOR
[2]  
ANDERSEN EB, 1980, DISCRETE DATA MODELS
[3]  
BAKER RJ, 1978, GLIM SYSTEM
[4]  
BARNDORFFNIELSE.OE, 1989, ASYMPTOTIC TECHNIQUE
[5]  
BARNDORFFNIELSEN O, 1979, J ROY STAT SOC B MET, V41, P279
[6]  
Bartlett M.S., 1935, J R STAT SOC, V2, P248, DOI [10.2307/2983639, DOI 10.2307/2983639]
[7]  
Breslow NE, 1980, IARC SCI PUBLICATION, V1
[8]  
CHRISTENSEN R, 1990, LOG LINEAR MODELS
[9]  
Cox DR, 1989, ANAL BINARY DATA, V2nd
[10]  
CRAMER H, 1946, MATH METHODS STATIST
123