Code-quality evaluation scheme for assessment of student contributions to programming projects

Abstract

Project-based learning is the most common approach to software engineering education, due to its emphasis on the teamwork skills essential to real-world collaborations. This study developed an automated programming assessment system (APAS) featuring a code-quality evaluation scheme to overcome difficulties in assessing the contribution of individual team members. Team participation is visualized on a weekly basis to provide insights pertaining to team dynamics, and metrics based on code quality allow the segmentation of students by level of contribution. Insights provided by the proposed system were also shown to facilitate interventions aimed at improving code quality. Empirical analysis based on submission data from 146 students (41 teams) demonstrated the feasibility of the proposed system in monitoring group-based learning projects at the university level, particularly in detecting free-riders.