EXPLORING ESSENTIAL CONCEPTS IN THE AUTOMATION OF STUDENT PLAGIARISM MANAGEMENT - A CASE STUDY

Objectives: The primary objective of this case study was to arrive at heuristics for the automation of the student plagiarism management process, which now peremptorily includes contract cheating using Large-Language Model (LLM) Artificial Intelligence (AI) tools, such as ChatGPT. Theoretical Framework: The essential core components of the academic integrity phenomenon, such as a formal institutional regulatory framework, imperative training, and common approaches for addressing transgressions, were extensively explored. The efficacy of the automated system used by the Private Higher Education Institution (PHEI) in the case study was investigated in relation to deterrance, by analysing the incident data captured on the system. The Technology Adoption Model (TAM) was applied for assessing usefulness/usability and ease-of-use perceptions of the system, both via questionnaires and using usage data captured on the system. Method: The PHEI’s database allows for quantitative analysis of patterns and trends in the occurrence of reported plagiarism (including contract cheating). The adoption of the automated system was assessed by the trends in the number of reported cases and in the percentage of repeat offenders. Results and Discussion: For plagiarism in general, the number of repeat offenders was consistently less than half of first offenders, with third offenders limited to between 0% and 1% in each subsequent year tested. The broader academic integrity system – incorporating the automated system - proved to be effective as a plagiarism deterrent. The results obtained also revealed that the (ostensible) prevalence of AI contract cheating was limited to about 7% of the total reported cases of plagiarism. Regarding the adoption of the automated system, both ease-of-use and usefulness TAM ratings were high overall. Research Implications: The study highlights possible system parameters, and possible implications and relationships implicit in automated systems related to plagiarism, as heuristics for further studies. Originality/Value: The potential and the constraints of enhancing efficacy by applying automated means for detecting and deterring plagiarism are highlighted. This study also elucidates the issues on the emerging spectrum of perspectives on the use of AI tools in academic research. This is evidenced by the complications of identifying plagiarism in AI-generated verbiage, as well as the academic value (or deprecation thereof) of incorporating AI tools in formal academic research.