Article
Summary:
Generalizing the result by Y. L. Tong, a chain of inequalities for probabilities in some types of multivariate distributions is proved. These inequalities embrace a large number of interesting special cases. Nine illustrations are given: cases of multivariate equicorrelated normal, $t,\chi^2$, Poisson, exponential distributions, normal and rank statistics for comparing many treatments with one control, order statistics used in estimating quantiles, and characteristic roots of covariance matrices in certain multiple sampling.
References:
[10] Z. Šidák:
Rectangular confidence regions for the means of multivariate normal distributions. J. Amer. Statist. Assoc. 62 (1967), 626-633.
MR 0216666
[11] Z. Šidák:
On probabilities of rectangles in multivariate Student distributions: their dependence on correlations. Ann. Math. Statist. 42 (1971), 169-175.
DOI 10.1214/aoms/1177693504 |
MR 0278354
[14] R. G. D. Steel:
A multiple comparison sign test: treatments versus control. J. Amer. Statist. Assoc. 54 (1959), 767-775.
MR 0109396 |
Zbl 0090.36003