Goodness-of-Fit Tests for the Gamma Distribution based on Residual Extropy

Abstract

In this article, we propose test statistics based on so-called residual extropy to test the conformity of a random sample with a two-parameter gamma distribution. First, we characterize the gamma distribution based on residual extropy. Then we form tests as the integrated deviation (or square deviation) between the sample and population residual extropies. Finally, Monte Carlo simulations will be conducted to calculate critical values and powers of the proposed tests and as well as the classical EDF tests; namely, Kolmogorov-Smirnov (KS), Cramer-von Mises (CvM), and Anderson-Darling (AD) tests. Power comparisons show that the proposed tests outperform the classical tests for a broad spectrum of alternative distributions.