Process improvement of peer code review and behavior analysis of its participants

Abstract

The software industry is urging on universities and colleges to cultivate software engineers who can write high quality programs. Peer code review (PCR) is accepted as an ideal way to maximize the learning outcome of students in writing quality code. Using this learning process, students improve their skills while scientific and efficient management removes the extra burden from instructors such as checking programs written by every student. In this paper, the previous PCR process was improved and the definitions of the relevant roles and documents were refined as well. After implementing this process in two academic years, some problems were found. By means of summarizing the email submissions by the students and also interviewing a few students, the behavior of all participants was preliminarily analyzed. With regards to the further quality assurance and high efficiency, a web-based management information system with a built-in blind review mechanism was discussed for solving the problems with process control, and also a game theoiy model was proposed for addressing the ethical issues in the whole PCR process.