A Modified Skewness Measure for Testing Asymmetry
Author:
- 11 January 2010
- journal article
- Published by Taylor & Francis in Communications in Statistics - Simulation and Computation
- Vol. 39 (2) , 335-346
Abstract
International audienceStatistical practitioners frequently wish to know whether a variable is symmetrically distributed. There are a number of different tests available but the most commonly used one is perhaps that based on the standardised third central moment, as defined by Pearson and Fisher in the early 1900:s. While this traditional skewness measure uniquely determines the symmetry of a variable within the Pearson family, it does not uniquely determine symmetry for a general distribution. In this paper we propose a modified version of the classical skewness test which is easy to conduct and consistent against a wide family of asymmetric distributionsKeywords
This publication has 16 references indexed in Scilit:
- Robust Testing for SkewnessCommunications in Statistics - Theory and Methods, 2007
- Order Statistics2003
- Testing for symmetryBiometrika, 1987
- Skewness and Asymmetry: Measures and OrderingsThe Annals of Statistics, 1986
- Testing SymmetryJournal of the American Statistical Association, 1982
- The Mean, Median, Mode Inequality and Skewness for a Class of DensitiesAustralian Journal of Statistics, 1981
- Plots and tests for symmetryBiometrika, 1977
- Information Given by Odd MomentsThe Annals of Mathematical Statistics, 1946
- The Limits of a Measure of SkewnessThe Annals of Mathematical Statistics, 1932
- "Das Fehlergesetz und Seine Verallgemeinerungen Durch Fechner und Pearson." A RejoinderBiometrika, 1905