-
作者：EFRON, B; THISTED, R
作者单位：Stanford University
摘要：Shakespeare wrote 31,534 different words, of which 14,376 appear only once, 4343 twice, etc. the question considered is how many words he knew but did not use. A parametric empirical Bayes model due to Fisher and a nonparametric model due to Good and Toulmin are examined. The latter theory is augmented using linear programming methods. The models are equivalent to supposing that Shakespeare knew at least 35,000 more words.
-
作者：COX, NR
作者单位：Imperial College London
摘要：When observations on 2 variables are sampled from 2 or more distinct populations, the linear structural relation between the 2 variables can be estimated consistently. The maximum likelihood estimators of the parameters in the model are found; they have different forms in different parts of the sample space. A procedure for testing hypotheses about the parameter usually of principal interest, the slope of the relation, is derived; it is complicated by the unusual nature of the parameter estima...
-
作者：DOKSUM, KA; SIEVERS, GL
作者单位：University of California System; University of California Berkeley; Western Michigan University
摘要：Statistical methods that give detailed descriptions of how populations differ are considered. These descriptions are in terms of a response function .DELTA.(.chi.) with the property that .times. + .DELTA.(.times.) has the same distribution as Y. The methods are based on simultaneous confidence bands for the response function computed from independent samples from the 2 populations. General and parametric models are considered and comparisons between the various methods are made.
-
作者：LAM, K; CHIU, WK
作者单位：University of Hong Kong
摘要：In the (.delta.*, P*) approach to the selection of the best of several normal populations there is a fundamentally important open question of whether the probability of correct selection increases with each of the sample sizes. Such monotonicity generally does not hold.
-
作者：CHMIEL, JJ
作者单位：Stanford University
摘要：Theoretical results for the Spearman-type variance estimator in the quantal response bioassay problem are determined when the mean of the underlying tolerance distribution is known, and also when it is unknown. For an experiment with infinitely many dose levels, the estimator is proved to be finite with probability 1, and its mean and variance are given. Allowing the dose interval to approach 0, the estimator is shown to be asymptotically normal and asymptotically highly efficient. Similar asy...
-
作者：COX, TF; LEWIS, T
作者单位：University of Hull
摘要：A new distance-based method is proposed for investigating the pattern in the plane formed by points, which may be assumed to be the positions of centers of trees in a forest stand. For each of several randomly selected origins the distance from the origin to the nearest tree, and the distance of this tree to its nearest neighbor are measured, and their ratio calculated. The values of this ratio which exceed 1/2 are shown to have the property of distinguishing between regular, random and cluste...
-
作者：FIENBERG, SE; LARNTZ, K
作者单位：University of Minnesota System; University of Minnesota Twin Cities
摘要：Log linear model representations for the Bradley-Terry paired comparisons and Luce multiple comparisons models are developed. Various multivariate extensions of the Bradley-Terry model are considered. A model is presented, similar to one of Davidson and Bradley (1969), but which is based on the log odds ratio as a measure of association rather than the .phi. coefficient of correlation. By the adaptive use of log linear model theory, complex data sets involving paired and multiple comparisons a...
-
作者：DAVIS, AW
作者单位：Commonwealth Scientific & Industrial Research Organisation (CSIRO)
-
作者：BARNDORFFNIELSEN, O
作者单位：Aarhus University
摘要：Various definitions of nonformation, i.e., the concept of a submodel and the corresponding part of the data containing no available information with respect to the parameter of interest, are discussed and exemplified. Certain of these have pointwise versions, reflecting the possibility that some experimental outcomes may be informative while others are not.
-
作者：BANCROFT, GA; DUNSMORE, IR
作者单位：University of Sheffield
摘要：In life testing suppose that the units under test can fail due to 1 of several competing risks. The estimation of parameters in such compound models is briefly discussed. Bayesian predictive distributions and probabilities are derived for such situations, and the predictions obtained are compared with some of the classical analogs.
