Methodology for Unique Questions Generation to Prevent Plagiarism in Lab Assignments

In this paper, we describe our operational experience in preventing plagiarism in computer science lab assignments while minimizing student disputes. Plagiarism is a prevalent issue in take-home lab assignments. The effort dedicated to plagiarism prevention is relatively minor compared to that of plagiarism detection. Most existing literature on plagiarism prevention only briefly outlines the necessary steps, with very few providing detailed, actionable guidelines. Additionally, some of the methods discussed are tailored for online tests rather than take-home laboratory assignments. This paper presents multiple strategies for generating unique questions for students. We also address various challenges related to question assignment, generation, grading, and future enhancements of questions. The generated topics in computer science encompass digital systems, microprocessor assembly language, operating systems, and computer networks. Furthermore, we outline best practices to mitigate plagiarism. In stateful question assignment, the questions are evenly distributed to all students. Our results show that when the number of questions is less than number of students, stateless question assignment has a better plagiarism control than stateful question assignment. Finally, we discuss the implications of unique question generation for student grading and propose several workarounds.