Using an Evolutionary Algorithm to Determine the Student-to-Proctor Ratio in Exam Scenarios
摘要
The computational simulation of students taking an exam and proctors monitoring it is relevant for educational institutions that want to design the way exams are applied. In our work, our objective is to minimize the students’ incentives to engage in dishonest behavior, and we explore this by considering alternate seating arrangements of the students as well as alternate monitoring approaches. Here, we propose a multi-objective evolutionary algorithm that finds how to arrange students in a given classroom and how many professors are required to monitor those students in order to reduce the students’ incentives to cheat. The algorithm finds all the Pareto solutions such that one cannot decrease students’ incentives to cheat without increasing the number of required professors. Using this system, our key contribution is to provide a heuristic for how many teachers should be assigned to monitor a given number of students by finding an adequate student-to-proctor ratio.