Two Powerful Tests for Normality

摘要

In this paper, two powerful tests for normality are proposed based on the Noughabi's entropy estimator (J Stat Comput Simul 80:1151-1162,2010). The power values of the proposed tests are computed and compared with the most popular normality test i.e., Shapiro-Wilk's test. According to (Judge et al. in The theory and practice of econometrics, Wiley, New York, 1980) the Shapiro-Wilk statistic has become the most popular normality test because of its performance in Monte Carlo simulations. Moreover, the Shapiro-Wilks test is one of the tests used in S AS software for testing normality. Finally, some illustrative examples are presented and analyzed.