A Heuristic for Estimating the Impact of Lingering Defects: Can Debt Analogy Be Used as a Metric?
Shirin Akbarinasaji, Ayşe Bener, Adam Neal · 2017
Background: Due to tight scheduling and limitedbudget, it may not be possible to resolve all the existing bugsin a current release of a software product. The accumulation ofthe deferred bugs in the issue tracking system are obligations (liabilities) of the software team similar to financial analogyof 'debt'. Defect debt is known as latent defects which arenot resolved in the current release. Aim: In order to managethe defect debt, software managers need to be aware of theamount of debt (principal) as well as the price of the credit (interest) in their system. There are no studies in the literature tomeasure the principal or interest of defect debt. In this study, wepropose a novel approach to identify the interest of defect debt. Methodology: We developed a heuristic to specify the interestbased on three metrics: PageRank index, customer feedback andbug fixing duration. In order to investigate the feasibility of ourheuristic, we employ it to two datasets that are extracted fromboth open source and commercial software products. We validatethe heuristic using two metrics: the severity/priority of bugs, andthe duration of bug fixing time. Result: The results show that 24%and 18% of the deferred bugs are high and medium impact bugsin project 1 and project 2, respectively.