A Simple Approach to Testing Homogeneity of Order-Constrained Means
针对Bartholomew提出的有序约束均值同质性似然比检验在零分布依赖序约束形式和样本量、计算复杂等问题，提出基于Fisher组合方法的替代程序，易于实施且功效相当。
Abstract The likelihood-ratio test for equality of partially ordered means, first proposed by Bartholomew, is known to possess power characteristics that are generally superior to those of competing procedures. A limitation of this test, however, is that the null distribution of the test statistic is heavily dependent on both the specific form of the order restriction and the group sample sizes. It is not readily available for most orderings, especially when the group sample sizes are unequal. Furthermore, computation of the test statistics often involves intricate algorithms. We propose a class of procedures based upon classical methods for combining independent tests. These procedures are generally applicable and easily implemented. A study of the power functions of the competing tests indicates that the power of the test based on Fisher's combination method compares reasonably with that of the likelihood-ratio test. It is concluded that the procedure based on Fisher's combination method is an effective alternative to the likelihood-ratio test, especially when the latter is difficult to implement.