Chapter 3 Divergence measures and their application to one-shot devices with a single failure mode ⍟ ⍟ This book has a companion website hosting complementary materials. Visit this URL to access it: https://data.mendeley.com/datasets/879xmdz3d8/1.

Classical inferential methods for one-shot device testing are based on the maximum likelihood estimators (MLEs), which are known for their efficiency, but also for their lack of robustness. To tackle the robustness problem, the use of divergence-based estimators is known to work well in many statistical contexts. In this chapter, we define a family of estimators and Wald-type tests based on the density power divergence (DPD) measure for one-shot device testing with a single failure mode under constant-stress accelerated life tests (CSALTs). This inference may then be applied under the assumption of different lifetime distributions and the robustness of these estimators and tests may be established through the study of their influence function (IF) and also empirically through Monte Carlo simulations.