Abstract

One sample location problem is widely studied problems in Nonparametric set up when the underlying distribution of the sample is symmetric. In this paper, testing for point of symmetry in one sample setting is considered by proposing a class of tests. The proposed class of tests is based on convex combination of two statistics, which are computed using subsamples. The Pitman asymptotic relative efficiencies of few members of the class are computed with respect to the other tests exist in the literature to evaluate its performance. The choice of subsample size is determined for different distributions to make the statistics easily usable.