Characterizations of the dispersive order of distributions by the Laplace transform (Q1807916): Difference between revisions

The author derives some characterizations of the dispersive stochastic order by means of the Laplace transform. Inequalities for the Laplace transforms of distribution functions and hazard rate functions are obtained as corollaries. Some orderings of order statistics are also obtained. For example, if \(\{X,X_1,X_2,\dots\}\) is a set of nonnegative, independent and identically distributed random variables, and if \(\{Y,Y_1,Y_2,\dots\}\) is another set of nonnegative, independent and identically distributed random variables, and if \(X_{(i:n)}\) is the \(i\)th order statistic in a sample of size \(n\) from \(\{X_1,X_2,\dots\}\), and \(Y_{(i:n)}\) is the \(i\)th order statistic in a sample of size \(n\) from \(\{Y_1,Y_2,\dots\}\), and if \(X\leq_{\text{disp}}Y\), then \(X_{(i:n)}+ Y_{(i-1:n)}\leq_{\text{Lt}} X_{(i- 1:n)}+ Y_{(i:n)}\) for \(2\leq i\leq n\), and \(X_{(n+ 1-i:n+1)}+ Y_{(n- i:n)}\leq_{\text{Lt}} X_{(n- i:n)}+ Y_{(n+ 1-i:n+1)}\) for \(0\leq i\leq n-1\), and \(X_{(i:n)}+ Y_{(i: n+1)}\leq_{\text{Lt}} X_{(i: n+1)}+ Y_{(i:n)}\) for \(1\leq i\leq n\), where \(\leq_{\text{disp}}\) and \(\leq_{\text{Lt}}\) denote the dispersive and the Laplace transform orders, respectively.