Use of Non-Parametric Approaches on Normality of Hydrologic Variables

Parametric approaches in statistical analysis assume that any given data are normally distributed. Therefore, the test of whether this conventional assumption is valid should be made in this context of the available data’s normality before being passed to the application of statistical tests. The pa...

Full description

Bibliographic Details
Main Authors: Kadri Yürekli, Müberra Erdoğan, Mehmet Murat Cömert
Format: Article
Language:English
Published: Turkish Science and Technology Publishing (TURSTEP) 2018-08-01
Series:Turkish Journal of Agriculture: Food Science and Technology
Subjects:
Online Access:http://www.agrifoodscience.com/index.php/TURJAF/article/view/1927
Description
Summary:Parametric approaches in statistical analysis assume that any given data are normally distributed. Therefore, the test of whether this conventional assumption is valid should be made in this context of the available data’s normality before being passed to the application of statistical tests. The paper is focused on the normality methodologies commonly used in literature, named Kolmogorov-Smirnov, Jarque-Bera, D’agostino, Anderson Darling, Shapiro-Wilk and Ryan Joiner. In the study, the seasonal maximum data from eight streamflow gauging stations in Yesilirmak Basin was used as material. The normality in the 59% of the whole data sets were obtained as the highest result by the Kolmogorov –Smirnov approach, when compared to the other normality tests considered in the study.
ISSN:2148-127X