On Slippage Tests and Multiple Decision (Selection and Ranking) Procedures.

摘要

Suppose k independent populations are given and we are interested in t (1 < or = t < k) of these populations which are better than the rest. Two different formulations, namely testing t slippages, and selection of a subset (of a random size or a fixed size t) containing the t better populations, are considered herein. Both of these formulations provide answers to certain relevant questions which are not answered by the classical tests of homogeneity.