UNIMODALITY OF DISTRIBUTION OF GENERALIZED ORDER STATISTICS

Abstract

Unimodality of the distribution of Generalized Order Statistics (GOSs) has a substantial role in many parameter and condence interval problems of statistics, actuarial science and economics. Under some restrictions on model parameters and distributions a number of authors have shown unimodality of distribution of GOSs. In this article, we present some new results on unimodality of distribution of GOSs that extend and generalize recently obtained results. A counter example showing that the conditions of the main Theorem are minimal is also provided.