Optimal step stress accelerated life testing for the Marshall-Olkin inverse log-logistic cumulative exposure model

In this article an attempt has been made to analyze the Marshall-Olkin inverse log-logistic simple step-stress accelerated life test under the cumulative exposure model. The scale parameter of the Marshall-Olkin inverse log-logistic distribution is assumed to be a log-linear function of the stress level. A comprehensive simulation study is conducted for various sample sizes at different stress levels with the optimum time to assess the performance of the maximum likelihood estimators via mean squared error, asymptotic variance, coverage probability, and confidence interval width. It is also concluded that increasing the difference between stress levels improves estimation performance when one stress level is fixed. However, the variations in parameter estimates among the examined stress combinations are not significant. For illustration purposes, one real data set is also examined.