THE COMPARISON OF HYPOTHESIS TESTS DETERMINING NORMALITY AND SIMILARITY OF SAMPLES

Year 2017, Volume: 13 Issue: 2, 21 - 36, 30.11.2017

Levent Şahintürk Burcu Özcan

Abstract

A number of hypothesis tests are used to obtain information
about the characteristics of one or more populations. While parametric tests
are based on the assumption of normal distribution, non-parametric tests are
performed with highly ordered series from the original series. The main purpose
of this study is to check whether two independent samples taken from two
different populations of normally distributed samples fit the normal
distribution with Kolmogorov-Smirnov and Shapiro-Wilk tests. And to determine
the similarities of Kolmogorov-Smirnov and Sign test to samples with normal
distribution by Wilcoxon test and those with non-normal distribution. As a
result, it is aimed to compare the strength and effectiveness of the applied
tests. We used the MATLAB function in our work which is considered to be useful
for researchers.

Keywords

Parametric Tests, Non-parametric Tests, Kolmogorov-Smirnov, Shapiro-Wilk, Wilcoxon, Sign Tests

References

• [1] Weakliem, D.L. (2016). Hypothesis Testing and Model Selection in the Social Sciences. A Division of Guilford Publications, Inc. New York, pp.1/202.
• [2] Gregory W.Corder and Dale I. Foreman Nonparametric Statistics for Non-Statisticians, Copyright 2009 John Wiley & Sons, Inc.
• [3] Karagöz Y. Nonparametrik Tekniklerin Güç ve Etkinlikler, Elektronik Sosyal Bilimler Dergisi 33.33 (2010).
• [4] Yılmaz Y. and Yılmaz Y. Parametrik olmayan testlerin pazarlama alanındaki araştırmalarda kullanım,: 1995-2002 arası yazın taraması (2005).
• [5] Gamgam H. and Altunkaynak B. Parametrik Olmayan Yöntemler 2013. [p35]
• [6] Üstündağ G. Bazı Parametrik Olmayan İstatistiksel Yöntemlerin İncelenmesi, Çukurova Üniversitesi Yüksek Lisans Tezi. 2004.
• [7] Razali N. M. and Wah Y. B. Power comparisons of shapiro-wilk, kolmogorov- smirnov, lilliefors and anderson-darling tests, Journal of statistical modeling and analytics 2.1 (2011): 21-33.
• [8] Demirci N. Parametrik Olmayan Testler. Balıkesir University Lesson Notes.
• [9] Çolak E. Mann-Whitney-U ve Wilcoxon Testleri, Eskişehir Osmangazi University Lesson Notes.
• [10] Brewer J. K. On the power of statistical tests in the American Educational Research Journal, American Educational Research Journal 9: 391–401.
• [11] Egboro F. O. The Implications of Parametric and Non-Parametric Statistics in Data Analysis in Marketing Research, International Journal of Humanities and Social Science Vol. 5, No. 6; June 2015.
• [12] Horton N. J. Switzer SS: Statistical methods in the journal, New Engl J Med. 2005, 353 (18): 1977-1979. 10.1056/NEJM200511033531823.
• [13] Feys J. Nonparametric Tests for the Interaction in Two-way Factorial Designs Using R, The R Journal, Volume 8/1, August 2016.
• [14] Murakami H. The power of the modified Wilcoxon rank-sum test for the one- sided alternative. A Journal of Theoretical and Applied Statistics 49.3 (2015): 781-794.
• [15] Oyekaand I.C.A. and Ebuh G.U. Modified Wilcoxon Signed-Rank Test, Open Journal of Statistics, no.2, pp. 172–176, 2012.
• [16] M. Amezziane and C. Schmegner "A class of one sample tests based on the Mann–Whitney–Wilcoxon functional." Journal of Statistical Computation and Simulation 85.3 (2015): 587-595.