Assurance for sample size determination in binomial reliability demonstration testing
K.J. Wilson and M. Farrow
https://doi.org/10.19124/ima.2018.001.30
Abstract
Frequently,manufacturers are required to demonstrate that productsmeet reliability targets. A typical way of doing this is to use reliability demonstration tests (RDTs), in which a number of products are put on test and the test is passed if a critical number or fewer fail. There are various methods for determining the sample size for such tests. Traditionally this was based on the size of a hypothesis test following the RDT. More recently, Bayesian approaches have been proposed based on the idea of risk criteria. However, these approaches do not lead to a single, optimal sample size and conflate the choice of sample size for the test and
the analysis to be undertaken once the test has been conducted. In this paper we offer an alternative approach to sample size determination based on the idea of assurance. This approach can overcome each of these issues with risk criteria. Assurance chooses the sample size to answer the question: “What is the probability that the RDT will result in a successful outcome?”We demonstrate the use of assurance for sample size calculations in binomial RDTs and discuss the specification of prior distributions for the design and analysis of the test.