An Examination of Fault Exposure Ratio

The fault exposure ratio K is an important factor that controls the per-fault hazard rate, and hence the effectiveness of testing of software. The paper examines the variations of K with fault density which declines with testing time. Because faults get harder to find, K should decline if testing is strictly random. However, it is shown that at lower fault densities K tends to increase; we explain this by using a hypothesis: real testing is more efficient than strictly random testing especially at the end of the test phase. Data sets from several different projects (in USA and Japan) are analyzed. If we combine the two factors, e.g., shift in the detectability profile and the nonrandomness of testing, then the analysis leads us to the logarithmic model which is known to have superior predictive capability.