A Likelihood Ratio Test for Stochastic Ordering
本文提出一种将似然比原理扩展到非参数环境的方法，用于检验随机排序，模拟显示该检验具有良好的检验功效，且是一致无偏的。
Abstract This article presents a method of extending the likelihood ratio principle to a nonparametric setting. The problem dealt with is a test for stochastic ordering. The resulting test is shown, via simulation, to have good power. It is shown to be consistent and unbiased.