Designing programs that are less likely to hide faults

Abstract An important motivation for software testing is to increase confidence that the software no longer contains faults. In this paper we explain a technique for using fewer tests to gain an equivalent confidence in software. Our techniques complement random black-box testing. To be able to use fewer tests and gain equivalent confidence, we must either 1. 1) isolate and remove software characteristics that discourage software from revealing faults during testing, or 2. 2) find a method of selecting tests that have a greater ability to reveal the existence of any existing faults. The first of these two alternatives is the subject of this article. We present a conjecture concerning “testability,” a software characteristic that frequently increases the likelihood that faults are detected during random black-box testing. We propose design measures to increases testability.