Approximate Maximum Product Spacings Estimation for the Parameter of a Rayleigh Distribution under Doubly Hybrid Censoring

In most of the life testing and reliability experiments, the experimenter is often, unable to observe life time of all items put on test and the data available to the experimenter is censored data. Especially, hybrid censoring scheme is a mixture of types I and II censoring. Maximum product spacings method is most suitable method, especially to those cases where one of the parameter have an unknown shifted origin. In this paper, we consider the maximum product spacings estimator of the scale parameter when the samples are doubly hybrid censored samples. However, maximum likelihood estimator cannot be obtained in a closed form. Therefore, the second is to consider the approximate maximum product spacings estimator of the scale parameter when the samples are doubly hybrid censored samples. The different estimators are compared by performing the Monte Carlo simulation is presented. Also, real data example based on doubly hybrid censoring have been also analysed.