摘要:

AbstractRecently, CRESSIE and WHITFORD (1986) showed that Welch's test ofH0:μ1= μ2can be biased, under nonnormality, where μ1and μ2are the means of two independent treatment groups. They suggested, therefore, that a two‐sample analog of Johnson's test be used instead. One goal in this paper is to examine the extent to which a two‐sample analog of Johnson's test improves upon Welch's technique in terms of Type I errors and power when sample sizes are small or moderately large. Several alternative procedures are also considered including an additional modification of Johnson's procedure, a procedure suggested by DUNNETT (1982) that uses Tiku's modified maximum likelihood estimate with 10% trimming, two versions of Efron's bootstrap, and a test recently proposed by WILCOX (1989). The paper also describes situations where Welch's procedure is not robust in terms of Type I errors. This is important because based on results currently available, Welch's procedure is thought to be nonrobust in terms of power, but robust in terms of Type I

ISSN:0323-3847    

DOI:10.1002/bimj.4710320702

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractA method is described for estimating the probability that transfusion of a given quantity of blood from an AIDS‐infected donor will result in infection in the recipient. The method is based on maximum likelihood, and requires the assumption that individual viruses act independently in causing infection. Results of simulations indicate that the resulting estimates have low bias. The method is applicable to a wide variety of other problem

ISSN:0323-3847    

DOI:10.1002/bimj.4710320703

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

ISSN:0323-3847    

DOI:10.1002/bimj.4710320704

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractThe problem of survival analysis with covariates when observations are subject to irregular interval censoring is considered. A proportional hazards model is assumed. A pseudo likelihood based on the comparison of all pairs of individuals whose survival times can be ordered, is proposed. Statistical properties of the maximum pseudo likelihood estimator and its score test are explored. Several examples are presented, one of which is an examination of the effect of covariates on HIV seroconversion.

ISSN:0323-3847    

DOI:10.1002/bimj.4710320705

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractRelative risk estimation in case‐control studies is based on the premise that the control group represents the underlying population. Often more than one control group is collected in order to minimize the possibility of accepting a false result.In this paper it is assumed that a case is matched individually to two different controls and that one control group may lack representativeness with respect to some risk factors. It is discussed whether this group may be used for relative risk estimation and a polychotomous logistic regression model is suggested in which these differences between the control groups are taken into account. A practical method for model search is given.Data of two case‐control studies are used to demonstrate the method. In a simulation study its efficiency is investigated. Some computations illustrate the effect of ignoring a bias in one control gr

ISSN:0323-3847    

DOI:10.1002/bimj.4710320706

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractPaper presents the new method for selection of variables with the greatest discriminatory ability in the location model for mixed variable discriminant analysis. The procedure is derived applying the theory of graphical models selection. An example of application is presented to compare results with the model choice based on Akaike criterion.

ISSN:0323-3847    

DOI:10.1002/bimj.4710320707

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractThe use of the common test statistics in obtaining appropriate null hypothesis or model for one‐dimensional qualitative data is re‐examined. We propose alternatives which do not depend on the distributional approximation of the test statistics. We examine these alternatives using data from a medical set‐up and conclude that these alternatives are adequate for descriptive purposes which is what is usually desirable for such data. An extra caution is required however, if estimates are of importance especially for such data which depends on a predictor var

ISSN:0323-3847    

DOI:10.1002/bimj.4710320708

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractFor the analysis of square contingency tables with ordered categories, this note proposes some models having the structure of symmetry or asymmetry plus the main‐diagonal exponential frequencies. These models are applied to the data from the Los Angeles endometrial cancer study in BRESLOW and DAY (1980, p. 185

ISSN:0323-3847    

DOI:10.1002/bimj.4710320709

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

ISSN:0323-3847    

DOI:10.1002/bimj.4710320711

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY

摘要:

AbstractPlanar random figures can be described by series of landmarks. Bookstein's model assumes that the landmarks result from independent individual fluctuations around fixed fictive landmark centres. The aim of this paper is to estimate the distances between the centers and the variances of fluctuations around them. Furthermore, two new shape variable estimators are suggested and compared with an estimator of Bookstein.

ISSN:0323-3847    

DOI:10.1002/bimj.4710320712

出版商:WILEY‐VCH Verlag    

年代:1990

数据来源: WILEY