Exact and Approximate Critical Values of Kolmogorov-Smirnov Test for Discrete Random Variables

There are several authors who have studied the problem of the application of Kolmogorov- Smirnov test in the case of discrete random variables. Indeed, as defined by its author, the Kolmogorov-Smirnov test is valid when the distribution function of the universe F(x) is continuous, and when is not satisfied this assumption the test is no longer applicable accurately. By the way Kolmogorov and Noether showed that if the distribution of the universe is discontinuous, the critical values are minor or equal as the the corresponding values of the continuous case. The objective of this paper is to present a procedure for determining the exact critical values of Kolmogorov-Smirnov test for discrete random variables.