SOME POPULAR NORMALITY TESTS FOR UNIVARIATE DISTRIBUTIONS

Abstract

The normal distribution has long been focal point of much of statistical study. Normality of an underlying data distribution can have an effect on the properties of estimation or inferential procedures. Although so important it was not before the introduction of the Shapiro-Wilk test that normality tests gained attraction. Today we encounter vast literature on these tests. Consequently, the normality tests such as Pearson's in sqr b(1), b(2) and Fisher's g(1), g(2) along with Shapiro-Wilk's W statistic have been indispensable outputs in statistical packages. In spite of this importance Turkish literature has been lacking from these tests. Therefore the introduction of this topic is aimed at by taking up normality tests such as Pearson and Fisher ones.