The Hoeffding assertion about weak convergence of the distributions of the expected values of the order statistics based on i.i.d. observations is extended to the case of dependent sample data.

A computational technique that transform integrals over RK, or some of its subsets, into the hypercube [0, 1]K can be exploited in order to solve integrals via Monte Carlo integration without the need to simulate from the original distribution; all that is needed is to simulate iid uniform [0,...