Abstract
Let Xi 1,Xi 2…Xi n,i = 1,2,…,K independent random samples of size ni from absolutely continuous distributions with distribution functions Fi. We assume that these distribution functions have zero as the common quantile of order a(0≤α≤1), i.e., Fi(0) = α for i = 1,2, …,k. It is also assumed that Fi's are identical in all respects except possibly their scale parameters i.e., We wish to test the null hypothesis H0:σ1 = σ2 = …=σk against the ordered alternative H1 : σ1≤σ2≤…≤σk with at least one strict inequality. Define for i< j, i, j = 1,2,…,k, Øij(Xic, Xjd) = 1 if 0 ≤ xic ≤ Xjd or Xjd ≤ xic ≤0, and -1 if 0 ≤ xjd ≤ Xjd or Xic≤ xjd ≤0, and 0 other-wise. Let Uij be the corresponding two-sample U-statistic. For testing H0 against H1, with Fi(0) = α i = 1,2,…k, we propose a class of distribution free tests based on the statistics. Large values of Tk are significant for testing H0 against H1. The optimum values of ais are obtained. The tests are quite efficient.