On Step-Stress Partially Accelerated Life Testing with Competing Risks Under Progressive Type-II Censoring

The article discusses the application of step-stress partially accelerated life testing (SSPALT) with competing risks under progressive Type-II censoring for the Nadarajah–Haghighi (NH) distribution. It introduces a new method for statistical inference in reliability testing, focusing on the NH distribution due to its ability to produce decreasing and unimodal failure rates. The paper presents maximum likelihood estimation (MLE) and Bayesian estimation (BE) methods, including approximate confidence intervals and credible intervals for model parameters. Simulation studies are conducted to compare the performance of MLEs and BEs under different loss functions. The results highlight the advantages of Bayesian estimation in terms of mean square error and average estimate, particularly under the LINEX loss function. The article concludes by emphasizing the practical implications of the proposed method in reliability engineering and the potential for further research in this area.