-
作者：MCCULLAGH, P
作者单位：University of London; Imperial College London
摘要：A parametric model is developed for the analysis of square contingency tables with ordered categories. Order among the categories is a built-in feature of the new model and it is unnecessary to assign arbitrary scores to the row and column variables. Special cases of the proposed model include conditional symmetry and symmetry. The relationship with marginal homogeneity is described. The model for quasisymmetry is considered and is invariant under general permutation transformations of the ind...
-
作者：PAUL, SR; PLACKETT, RL
摘要：Data presented as contingency tables and classified by qualitative or quantitative methods are usually analyzed on the basis of a Poisson log linear model. The effect of Poisson mixtures, especially as represented by the negative binomial distribution, on statistical inference was examined. The main finding is that probabilities of rejection are increased, sometimes considerably. In the case of heterogeneous binary data, attention is given to the problems of analysis implied by Poisson mixtures.
-
作者：SEHEULT, A
摘要：A Neyman-Pearson test of identification in forensic science reflects, all the properties of a Bayes factor approach presented by Lindley in a previous volume of this journal.
-
作者：BRYANT, P; WILLIAMSON, JA
作者单位：International Business Machines (IBM); IBM USA; University of Colorado System; University of Colorado Boulder
摘要：Maximum likelihood techniques as applied to classification and clustering problems are examined, and the classification maximum likelihood technique, in which individual observations are assigned on an all-or-nothing basis to 1 of several classes as part of the maximization process, is shown to give results asymptotically biased. This extends Marriott''s work for normal component distributions. Numerical examples are presented for normal component distributions and for a problem in genetics. B...
-
作者：SCHUCANY, WR; PARR, WC; BOYER, JE
摘要：Several measures of dependence are examined for the Farlie-Gumbel-Morgenstern system, F, of joint distributions. Some bounds and relationships are shown for product-moment correlation coefficients and rank correlation coefficients. While the F system is an interesting construction for specificied marginals, it is limited to describing weak dependence. A quality control application using a related screening variable is given.
-
作者：EAVES, D; BURR, AH
作者单位：Simon Fraser University
摘要：The threshold hypothesis holds that the vector of L population proportions of flashes seen corresponding to L different flash intensities lies on a certain discrete family of curves in the unit L-cube. Tests are described of this sharp hypothesis against vague alternative hypotheses that these proportions are more or less unrelated. The calculations are described for the posterior distributions of the threshold number (N) of photons needed for flash perceptions and of a photon receptivity coef...
-
作者：GLAZEBROOK, KD
摘要：A model for the allocation problem in a controlled clinical trial is proposed and is more general than the two-armed bandit. The trial involves more than 2 treatments and experiments to have a large set of possible outcomes. The optimal strategy for the allocation problem is straightforward to compute.
-
作者：MEEKER, WQ
摘要：A test for independence between 2 dichotomous characteristics when data are obtained sequentially is suggested. Truncated sequential test regions are found, and the exact properties of the tests are evaluated. A numerical example is given, and test properties are compared with those of comparable fixed sample size tests.
-
作者：LAGAKOS, SW; WILLIAMS, JS
作者单位：Harvard University; Colorado State University System; Colorado State University Fort Collins; State University of New York (SUNY) System; University at Buffalo, SUNY
摘要：A class of models was proposed for the analysis of censored survival data in situations where survival was affected by conditions which lead to censoring. The specific case where survival was exponential was examined in detail and the methods were applied to a lung cancer clinical trial [in humans].
-
作者：BINDER, DA
摘要：A parametric model for partitioning individuals into mutually exclusive groups is given. A Bayesian analysis is applied and a loss structure imposed. A model-dependent definition of a similarity matrix is proposed, and estimates based on this matrix are justified in a decision-theoretic framework. Some existing cluster analysis techniques are derived as special limiting cases. The results of the procedure applied to 2 data sets are compared with other analyses.
