Comparison of different estimation methods for the Marshall–Olkin extended Weibull distribution

1Doctor of Philosophy, Associate Professor, Department of Statistics, Faculty of Science, Anadolu University, Eskisehir, Turkey 2Master of Statistics, Faculty of Science, Anadolu University, Eskisehir, Turkey Abstract: The aim of this paper is to determine an efficient method for estimating the parameters of Marshall–Olkin extended Weibull distribution. For this purpose, the performance of different estimation methods called maximum likelihood, least-squares and weighted least-squares methods, is compared in terms of the bias and mean square error through an extensive numerical simulation. It can be concluded from simulation results that the least-square method usually shows good performance in terms of bias for the considered cases. Moreover, with regard to the mean square error, the weighted least square method unexpectedly outperforms the other estimators for most of the considered cases.