-
作者:PATTERSON, HD; WILLIAMS, ER
作者单位:University of Edinburgh; University of Edinburgh
摘要:This paper describes an algorithm for constructing resolvable incomplete block designs for any number of varieties v and block size k such that v is a multiple of k. These designs are called .alpha.-designs. They include as special cases some lattice and resolvable cyclic designs. Additional designs with 2 block sizes differing by 1 plot are derived by omitting 1 or more varieties of the .alpha.-designs. The designs are available with high efficiency factors for a wide range of parameter values.
-
作者:YANG, MCK
作者单位:State University System of Florida; University of Florida
摘要:Suppose that a group of animals moves from an unknown location to a particular area. Their migration cannot be seen, but at each spot of this particular area it is possible to estimate the arrival time of the animals by continual observation. Designs for allocating the observation stations to obtain a desirable estimate of the direction angle along which the animals move into this area are discussed. This problem arose from lobster research. Since lobsters move undetected on the sea floor, we ...
-
作者:COX, TF
作者单位:University of Hull
摘要:Estimators for the density of points in a plane may be unbiased for the spatial pattern for which they were designed but are usually biased for most others. Two new estimators are proposed which are approximately unbiased for a wide range of patterns, and are used on data in the form of the positions of the centers of trees in a forest stand.
-
作者:MARCUS, R
作者单位:Tel Aviv University
摘要:Powers are compared of some tests of the equality of means in a 1-way analysis of variance model with ordered alternative. The test statistics are the likelihood ratio test, Williams''s test and a modified Williams''s test. [Drug dosage tests are mentioned].
-
作者:MCLACHLAN, GJ
作者单位:University of Queensland
-
作者:SHUSTER, JJ; DOWNING, DJ
作者单位:State University System of Florida; University of Florida; Marquette University
摘要:Methods for testing independence, quasiindependence and marginal symmetry in contingency tables are derived for a wide variety of sampling schemes including stratified multistage cluster sampling. The null hypothesis is a vector of linear and quadratic contrasts involving the probabilities. The asymptotic null distribution of the test static is chi-squared. The thoery can also be used to test equality of failure distributions in a stratified prospective multiclinic [human treatment] trial, joi...