in probability theory, a proof technique that allows one to compare two unrelated random variables by creating a random vector whose marginal distributions correspond to the two random variables