Evaluating an electronic plagiarism detection service: the importance of trust and the difficulty of proving students don’t cheat

Plagiarism by students is seen as a problem for universities. The fear is that students will increasingly use the internet and related technologies to obtain analysis, interpretation or even complete assignments and then submit these as their own work. Academics and markers face the problem of trying to identify this work in order to safeguard the reputation of their institutions and fairly reward those students who have completed their assessments fairly and honestly. Within this context, electronic plagiarism detection services provide a resource that might screen out some forms of unfair practice but in doing so create additional work in certifying the absence of plagiarism. This paper reports the results of an evaluation of one such service as used within an interdisciplinary school of social sciences. The paper describes how the system works, the experiences of staff and students in using the service together with an evaluation of the data generated by the process. The key findings are that the service did identify examples of poor scholarship and unfair practice that had not been detected under the usual marking system but that rigorously checking every ‘low-risk’ script for plagiarism will overwhelm the marking process. Trust and student honesty thus remain central to a successful academic system