-
作者：MARTIN, RJ
-
作者：PEDERSEN, BV
摘要：Conditional density and distribution functions were approximated by the mixed Edgeworth-saddlepoint expansion. This technique is described, and its use and character are illustrated by 3 examples [time-dependent Poisson process, comparing rates of Poisson processes and truncated normal density for survival data] concerning miscellaneous exponential families of order 2.
-
作者：ROSNER, B
摘要：Maximum likelihood was used to estimate a mother-child correlation, .rho.ms, when there are a variable number of offspring per family. The technique is iterative and uses standard Newton-Raphson procedures to facilitate convergence. A comparison is made using Monte Carlo methods between the maximum likelihood estimator and other more traditional estimators of interclass correlations. The maximum likelihood estimator is roughly equivalent in mean squared error to the pairwise estimator.
-
作者：HOEL, PG; JENNRICH, RI
摘要：Design problems associated with estimating response probabilities in low dose carcinogen experiments are considered. Optimal designs are developed for a specific class of models and examples are used to investigate the potential gains that may result from their use. Their robustness with respect to the prior information needed for their construction is considered. An algorithm for producing the required optimal design is presented.
-
作者：PRENTICE, RL; PYKE, R
作者单位：University of Washington; University of Washington Seattle
摘要：The probability of disease development in a defined time period is described by a log regression model. A model for the regression variable, given disease status, is induced and is applied to case control data. The odds ratio estimators and the asymptotic variance matrices may be obtained by applying the original log regression model to the case control study as if the data are obtained in a prospective study. This result gives a flexible and convenient method of analysis for a range of case c...
-
作者：AITCHISON, J
摘要：To obtain a sufficiently large diagnostic training set for differential diagnosis it is often necessary to use cases from 2 or more clinics. Important questions that arise are whether there are any differences between clinics in their methods of measuring the diagnostic features of cases, to what extent ignoring such differences invalidates any diagnostic system devised for these clinics and in what manner interclinic calibration information may be employed to make efficient use of the complet...
-
作者：BEGG, CB; MEHTA, CR
作者单位：Harvard University; Harvard T.H. Chan School of Public Health; Harvard University Medical Affiliates; Dana-Farber Cancer Institute
摘要：An adaptive procedure is proposed for a comparative clinical trial of 2 treatments in which the response variable has a normal distribution. Patients are assumed to arrive in pairs, and at each stage, after each pair has been treated, a decision is made either to continue to the next stage, or to suspend randomization and to assign patients to the best treatment. This leads to stopping rules which are tabulated for a variety of patient horizons. In general, trials will be relatively short, and...
-
作者：BRETH, M
摘要：Bayesian confidence bands for a distribution function are converted into confidence intervals for specified population quantiles and also for the mean. Bayesian tolerance limits are similarly constructed.
-
作者：TARONE, RE
摘要：Using the C(.alpha.) procedure of Neyman (1959), tests were derived for the goodness of fit of the binomial distribution which are asymptotically optimal against generalized binomial alternatives proposed by Altham (1978) and Kupper and Haseman (1978). The C(.alpha.) tests optimal against correlated binomial alternatives, beta-binomial alternatives and a general class of mixture alternatives suggested by Wisniewski (1968) are shown to be closely related to the binomial variance test. A numeric...
-
作者：FAREWELL, VT
摘要：The logistic model (Cox, 1970 2) was widely used in the analysis of retrospective studies [of disease incidence]. Two methods of analysis, one based on work by Mantel (1973) and Anderson (1972) and the other based on a conditional likelihood are discussed. The motivation for the 1st method is clarified and some efficiency calculations are presented.
