Inference for the stress-strength reliability for the two-parameter Burr type X under Type I and Type II censoring

In this paper, asymptotic normal, approximate F-based and BCa bootstrap inferences are proposed for the stress-strength reliability model P(X<Y), with the assumption that both samples from the two-parameter Burr type X distribution under Type I and Type II censoring. A simulation study is carried out. There is no clear winner among these methods, asymptotic normal and approximate F-based inference work well for the heavy censoring case when R is small, while BCa bootstrap inference could a good choice when R is large.