Improving Software Quality through Instrumented Formal Technical Review

点击翻译按钮获取中文摘要

Formal technical review (FTR) encompasses a range of structured, collaborative methods employing technical personnel to improve the quality of a software development artifact. Research shows that FTR, when appropriately practiced, provides significant benefits. However, research also shows that FTR introduces substantial new overhead, and that its cost-effectiveness is easily compromised by a variety of group process obstacles. In addition, current FTR practice is difficult and costly to investigate effectively, and thus many important questions about its nature and comparative effectiveness remain unanswered. This project uses CSRS, a client-server, instrumented, collaborative software review system to investigate issues in research and practice of FTR. Using CSRS, controlled empirical studies provide insight into factors (such as the presence or absence of meetings) influencing the cost-effectiveness of review. The project is also enhancing CSRS to support its public release and use in case-study research in industrial settings. These efforts provide insight into technology transfer issues arising in the shift from manual to computer-supported software review. They also support evaluation of novel forms of FTR made possible by computer support, such as software review involving 20 or more participants, and software review where participants are geographically distributed.