作者: Olive Jean Dunn , Frank J. Massey
DOI: 10.1080/01621459.1965.10480814
关键词: Exact statistics 、 Mathematics 、 Confidence interval 、 Credible interval 、 Multivariate t-distribution 、 Confidence region 、 Applied mathematics 、 CDF-based nonparametric confidence interval 、 Robust confidence intervals 、 Statistics 、 Confidence distribution
摘要: Abstract Various methods based on Student t variates have been suggested and used for obtaining simultaneous confidence intervals several means, or contrasts among means. Determination of an overall level such involves evaluating the probability mass a multivariate distribution over hypercube centered at origin, with sides paralleling coordinate planes, bounds this mass. Since distributions involve many nuisance parameters, impossible number tables would be necessary in order to make exact intervals. In virtual absence tables, approximations become important. paper, attempt has made investigate adequacy certain [2], [5], [8] by computing some particular cases. These compared approximations. This paper is concerned two-sided intervals, rather than one-side...