A rank test of the crossing point of two continuous cdf's (Q1906601): Difference between revisions

Based on a new criterion function, this paper proposes a new test statistic for a hypothesis testing problem: whether or not there exists a unique crossing point between two continuous distribution functions. Tables of critical values for a small sample and a large sample test are given. A Monte Carlo study shows that the power of this kind of tests is more significant not only than the Kolmogorov-Smirnov test but also than the Hawkins-Kochar test [\textit{D. L. Hawkins} and \textit{S. C. Kochar}, Ann. Stat. 19, No. 3, 1626-1638 (1991; Zbl 0729.62046)].