Abstract
Software inspection is one of the methods to ensure the quality of software by finding and repairing defect early in software development process. In a software inspection process, inspectors first review software artifacts individually and then meet in a team in order to find as many defects as possible and to eliminate false positives. However, several empirical studies suggest that inspection meeting may not be necessary since an insignificant number of new defects are found as a result of the meeting.
In this paper we report on a controlled experiment with 54 undergraduate students who inspection Object-Oriented design documents. Firstly, we compare the performance of inspection teams using two reading techniques: Checklist-Based Reading (CBR) and Perspective-Based Reading (PBR). Secondly, we compare the performance of interacting and nominal teams. The results of comparison between CBR and PBR do not reveal a significant difference between these techniques. Meanwhile, the results of comparison between interacting and nominal teams show that interacting teams do not detect a significant number of new defects during a meeting, and they are less effective than nominal teams; however interacting teams reported fewer false positives as compared to nominal teams.
In this paper we report on a controlled experiment with 54 undergraduate students who inspection Object-Oriented design documents. Firstly, we compare the performance of inspection teams using two reading techniques: Checklist-Based Reading (CBR) and Perspective-Based Reading (PBR). Secondly, we compare the performance of interacting and nominal teams. The results of comparison between CBR and PBR do not reveal a significant difference between these techniques. Meanwhile, the results of comparison between interacting and nominal teams show that interacting teams do not detect a significant number of new defects during a meeting, and they are less effective than nominal teams; however interacting teams reported fewer false positives as compared to nominal teams.
Original language | English |
---|---|
Pages (from-to) | 875-886 |
Number of pages | 12 |
Journal | Information and Software Technology |
Volume | 46 |
Issue number | 13 |
DOIs | |
Publication status | Published - Oct 2004 |
Externally published | Yes |