Analysis of Learning Records to Detect Student Cheating on Online Exams: Case Study during COVID-19 Pandemic

Abstract

In March 2020, due to the Covid19 pandemic, higher education had to switch from face-to-face to exclusively virtual mode overnight. In this unexpected scenario, supervisors also had to adapt the assessment procedures, including the exams. This caused a significant controversy, as, according to many students, supervisors were more concerned about how to prevent students from cheating, than actually measuring their learning. This paper introduces an experience that implemented several of the students' requests in an online exam and conducts a comprehensive analysis of students’ behavior according to the virtual learning environment records. Different existing software tools are used for the analysis, complemented with a Python application ad-hoc developed. The objective indicators gathered provide evidence that some students cheated and invite focusing on evidence-based assessment.