Comparison of Four Nonparametric Tests of Symmetry

Many statistical procedures assume symmetrically distributed data (examples are linear rank tests, estimation of location parameters, two-tailed p-values, and confidence intervalls). Thus, it is important check the assumption of symmetry. Using a simulation study this paper compares the performance of four different tests on symmetry, namely the Randles Fligner Policello Wolfe Test, the Cabilio Masaro Test, the Mira Test, and the Miao Gel Gastwirth Test.