Applying Text Analysis for Detecting Academic Misconduct on a Statistics Exam

Abstract

This practitioner paper presents an example of using text similarity analysis (specifically using Levenshtein similarity) as one component of an investigation into incidents of academic dishonesty in an online assessment at a Canadian university. The paper begins with an overview of the Levenshtein similarity method followed by a description of the academic offences it was used to provide evidence for. The paper concludes with reflections on challenges and opportunities the use of text similarity analysis affords.