A NONPARAMETRIC TEST IN THE TWO-SAMPLE PROBLEM WHEN THE FIRST POPULATION IS SYMMETRIC

In the two-sample problem, where the distribution of the first population is symmetric, we propose a test procedure to examine whether the distribution of the first population is the same as the second. And it is shown by a simulation study that the power of the proposed test is almostly better than those of the median test and Wilcoxon's test in some circumstances. Further we derive a test procedure in the large sample case.