Inference for block progressive censored competing risks data from an inverted exponentiated exponential model

In this paper, reliability estimation for a competing risks model is discussed under a block progressive censoring scheme, which improves experimental efficiency through testing items under different testing facilities. When the lifetime of units follows an inverted exponentiated exponential distribution (IEED) and taking difference in testing facilities into account, various approaches are established for estimating unknown parameters, reliability performances and the differences in different testing facilities. Maximum likelihood estimators of IEED competing risks parameters together with existence and uniqueness are established, and the reliability performances and the difference in different testing facilities are also obtained in consequence. In addition, a hierarchical Bayes approach is proposed and the Metropolis-Hastings sampling algorithm is constructed for complex posterior computation. Finally, extensive simulation studies and a real data analysis are carried out to elaborate the performance of the methods, and the numerical results show that the proposed hierarchical Bayes model outperforms than classical likelihood method under block progressive censoring.