On the analysis of censored reliability data

We develop an analysis of the reliability of reliability data following a negative exponential distribution and subject to type I censoring. Our main focus is to determine the role of the particular analytical approach in interim analysis. Accordingly, in this preliminary study, we aim to investigate, by means of a detailed simulation study, the effects of early stopping on decision-making in a comparative two group reliability study (eg, a two group randomised controlled clinical trial). In passing, we extend previous results to inverse parametrizations and type II censoring.