An alternative to the kolmogorov-smirnov test for goodness of fit
Author:
- 1 January 1987
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Theory and Methods
- Vol. 16 (12) , 3519-3534
Abstract
A test is proposed which requires a better fit in the extremes of a distribution than the Kolmogorov-Smirnov test for H0. not to be rejected. Critical values are calculated for sample sizes up to 100, and approximate critical values are found for larger samples. The power of the test is obtained for a number of distributions, and it is shown that the test is more powerful than some existing tests for a wide range of casesKeywords
This publication has 8 references indexed in Scilit:
- A Modified Kolmogorov-Smirnov Test Sensitive to Tail AlternativesThe Annals of Statistics, 1983
- Sheffer Polynomials for Computing Exact Kolmogorov-Smirnov and Renyi Type DistributionsThe Annals of Statistics, 1981
- Rectangle Probabilities for Uniform Order statistics and the Probability That the Empirical Distribution Function Lies Between Two Distribution FunctionsThe Annals of Mathematical Statistics, 1971
- On Certain Confidence Contours for Distribution FunctionsThe Annals of Mathematical Statistics, 1954
- Numerical Tabulation of the Distribution of Kolmogorov's Statistic for Finite Sample SizeJournal of the American Statistical Association, 1952
- Asymptotic Theory of Certain "Goodness of Fit" Criteria Based on Stochastic ProcessesThe Annals of Mathematical Statistics, 1952
- Table for Estimating the Goodness of Fit of Empirical DistributionsThe Annals of Mathematical Statistics, 1948
- On the Kolmogorov-Smirnov Limit Theorems for Empirical DistributionsThe Annals of Mathematical Statistics, 1948