Inference under balanced joint progressive type-II censoring scheme

In this paper, we consider a balanced joint progressive type-II censoring scheme and develop inference procedures for populations exhibiting bathtub-shaped hazard rates. Such models are appropriate for modeling phenomena which indicate non-monotone failure pattern. We focus on finding useful inferences upon model parameters by considering the Chen distribution. Point and interval estimation are considered using maximum likelihood and Bayesian methods. The existence and uniqueness of the maximum likelihood estimators are established. Furthermore, asymptotic confidence intervals and bootstrap-based intervals are constructed for the model parameters. Bayesian estimates and corresponding highest posterior density intervals are obtained using an importance sampling technique under general prior assumptions. The performance of the Bayesian estimators is assessed and compared with classical estimates through extensive Monte Carlo simulation studies. A real data example is also presented to demonstrate the practical applicability of proposed methods.