返回Google圖書搜尋

Comparing Detection Methods for Software Requirements Inspections

其他書名	A Replicated Experiment
出版	University of Maryland, 1994
URL	http://books.google.com.hk/books?id=HMw4GwAACAAJ&hl=&source=gbs_api

註釋Abstract: "Software requirements specifications (SRS) are usually validated by inspections, in which several reviewers independently analyze all or part of the specification and search for defects. These defects are then collected at a meeting of the reviewers and author(s). Usually, reviewers use Ad Hoc or Checklist methods to uncover defects. These methods force all reviewers to rely on nonsystematic techniques to search for a wide variety of defects. We hypothesize that a Scenario-based method, in which each reviewer uses different, systematic techniques to search for different, specific classes of defects, will have a significantly higher success rate. We evaluated this hypothesis using a 3 x 24 partial factorial, randomized experimental design. Forty eight graduate students in computer science participated in the experiment. They were assembled into sixteen, three-person teams. Each team inspected two SRS using some combination of Ad Hoc, Checklist, or Scenario methods. For each inspection we performed four measurements: (1) individual defect detection rate, (2) team defect detection rate, (3) percentage of defects first identified at the collection meeting (meeting gain rate), and (4) percentage of defects first identified by an individual, but never reported at the collection meeting (meeting loss rate). The experimental results show that (1) the Scenario method has a higher defect detection rate than either Ad Hoc or Checklist methods, (2) Scenario reviewers are more effective at detecting the defects their scenarios are designed to uncover, and are no less effective at detecting other defects, (3) Checklist reviewers were no more effective than Ad Hoc reviewers, and (4) Collection meetings produce no net improvement in the defect detection rate -- meeting gains are offset by meeting losses."