Comparisons of the Goodness of Fit Tests for the Geometric Distribution


Özonur D., Gökpınar E., Gökpınar F., Bayrak H., Gül H. H.

GAZI UNIVERSITY JOURNAL OF SCIENCE, vol.26, no.3, pp.369-375, 2013 (Scopus)

  • Publication Type: Article / Article
  • Volume: 26 Issue: 3
  • Publication Date: 2013
  • Journal Name: GAZI UNIVERSITY JOURNAL OF SCIENCE
  • Journal Indexes: Scopus
  • Page Numbers: pp.369-375
  • Gazi University Affiliated: Yes

Abstract

This article gives some goodness of fit tests for the Geometric distribution. The tests used in this study are Anderson Darling test, Cramer-von Mises test, Kolmogorov Smirnov test, test based on partition of Chi-square and some new alternatives based on smooth tests. The performance of these tests is compared by simulation study according to type-I errors and powers of tests. The power comparisons indicate that the modified version of smooth test statistic has the highest power value of test for Negative Binomial, Binomial and Poisson distributions. Also, Kolmogorov Smirnov, Anderson Darling, Cramer-von Mises and  test statistics can be preferred for large sample sizes.