摘要:

Inferences are made concerning population proportions when data are not missing at random.Both one sample and two sample situations are considered with examples in clinical trials.The one samplesituation involves the existence of response related incomplete data in a study conducted to make inferences involving the proportion. The two sample problem involves comparing two treatments in clinical trials when there exists dropouts due to both the treatment and the response to the treatment.Bayes procedures are used in estimating parameters of interest and testing hypotheses of interest in these two situations. An ad-hoc approach to the classical inference is presented for each ofthe two situations and compared with the Bayesian approach discussed. To illustrate the theory developed, data from clinical trials of severe head trauma patients at the Medical College of Virginia Head Injury Center from 1984 to 1987 is considered

ISSN:0361-0918    

DOI:10.1080/03610919408813197

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

In statistical applications involving linear models, many inferences can only be made based on statistics that are functions of positive linear combinations of mean squares. The Satterthwaite procedure is commonly used to approximate the distribution of such a linear combination to a chi-squared with appropriate degrees of freedom. This article presents a new procedure, based on recent work on confidence intervals, for determining this approximate degrees of freedom. Numerical examples are also given to compare the performance of the proposed method with other approximations

ISSN:0361-0918    

DOI:10.1080/03610919408813198

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

We consider a general linear model (y,X,U) where U is a general positive definite matrix and X is possibly rank-deficient. We give updated formulae for various statistical quantities of interest (BLUEs, residual sum of squares, etc.) in the following situations: introduction of an additional observation, deletion of an observation, inclusion of a new regressor and deletion of a regressor. We give the formulae in statistical terminology so that their significance is better understood. We then give an application of these results to Regression Diagnostics in the linear model with correlated errors

ISSN:0361-0918    

DOI:10.1080/03610919408813199

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

In this article, we identify a situation in Meta analysis involving independent studies in which the conventional omnibus tests are orthogonal to a certain type of votecounting method. This, in turn, gives partial evidence pointing to the inefficiency of using a conventional omnibus test or a votecounting method alone. Also a method is suggested to combine an omnibus test and the votecounting method when an appropriate situation arises in practice. By simulation, it is demonstrated that the test due to this process of combination has somewhat more power than the omnibus tests for many classical problems

ISSN:0361-0918    

DOI:10.1080/03610919408813200

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

The procedure proposed by Buckley and James(1979)for estimating the regression coefficients in acensored data linear regression model has performed well in several studies and does not require distributional assumptions.Recently Hillis (1993) compares the finite sample properties of the methods proposed by Buckley and James(1979)Smith(1986)and Weissfeld and Schneider (1987) for estimatingthe covariance matrix of the Buckley James estimator and concludes that Smith's estimator is superior. However, Smith's method is defined only for the variance of the slope estimator in the simplelinear regression model. In this paper we consider a heuristic generalization of Smith's method tothe multiple linear regression model and compare the generalized Smith estimator with the Buckley James and WeissfeldSchneider estimators through simulations.The generalized version of Smith's method performs the best

ISSN:0361-0918    

DOI:10.1080/03610919408813201

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

A new clustering admissibility condition, metric admissibility,is introduced. This admissibility condition is important in clustering applications where it is desired that the cluster distances remain metric(satisfy the triangle inequality).The Lance and Williams infinite family of clusteringalgorithms is evaluated with respect to this admissibility condition. This family contains most ofthe commonly used agglomerative clustering algorithms. Necessary and sufficient conditions are given on the parameters of the Lance and Williams cluster distance function in order to assure metricadmissibility of the corresponding algorithms

ISSN:0361-0918    

DOI:10.1080/03610919408813202

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

Theoretical considerations and empirical results show that the one-dimensional quality of non-uniform random numbers is bad and the discrepancy is high when they are generated by the ratio of uniforms method combined with linear congruent ial generators. This observation motivates the suggestion to replace the ratio of uniforms method by transformed rejection (also called exact approximation or almost exact inversion), as the above problem does not occur for this method. Using the functionwith appropriate a and b as approximation of the inverse distribution function the transformed rejection method can be used for the same distributions as the ratio of uniforms method. The resulting algorithms for the normal, the exponential and the t-distribution are short and easy to implement. Looking at the number of uniform deviates required, at the code length and at the speed the suggested algorithms are superior to the ratio of uniforms method and compare well with other algorithms suggested in literature

ISSN:0361-0918    

DOI:10.1080/03610919408813203

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

摘要:

Use of Newton's method for computing the noncentrality parameter based on the specified power in sample size problems of chi-squared tests requires that we evaluate both the noncentral chi-squareddistribution function and its derivative with respect to the noncentrality parameter. A close relationship between computing formulas for them is revealed, by which their evaluations can be performed jointly. This property greatly reduces the amount of computation involved. The corresponding algorithm is provided in a step-by-step form.

ISSN:0361-0918    

DOI:10.1080/03610919408813204

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor

ISSN:0361-0918    

DOI:10.1080/03610919408813186

出版商:Marcel Dekker, Inc.    

年代:1994

数据来源: Taylor