A Proposal on How to Use Software Reliability Growth Model to Build Confidence in Dashboard Testing

Abstract

Software testing as an integral part of software development leads to the question of when the software (or application/dashboard) can be released and how confident that most defects/bugs/faults have been discovered. This paper discussed a relatively new but simple and practical proposal that can be used to build confidence for releasing software (or applications/dashboards). Instead of contrasting various software reliability growth models (SRGM) and choosing which one is the best, we use them to collaborate to help make decisions. We demonstrate our proposal with 18 real-life datasets that are publicly available in the literature. We use three widely used SRGMs, namely: Bass, Gompertz, and Logistic in our proposal to identify when we can stop testing. It turns out that when the testing has found most defects, most (if not all) of the SRGMs will converge to similar value for the maximum potential defects in the system.