Academic integrity is integral to the education system. Society relies upon the integrity of qualifications awarded to individuals during their studies as evidence of their knowledge and ability. Sadly, academic integrity is not always upheld, and for multifarious reasons, students find themselves engaging in academic misconduct, whether that is by copying exam answers from a peer, collaborating on an assignment that was meant to be individual work, copy-and-paste plagiarism, or for some, when the challenges of completing work on time become too great, turning to an essay mill to complete their work on their behalf. Approaches to ensuring academic integrity focus on prevention, detection and penalty, and these can be addressed through culture, strategy and policy. This thesis takes a holistic approach to academic integrity, by firstly exploring students’ perceptions of academic integrity and academic misconduct, and then secondly by developing a novel approach to the detection of plagiarism and contract cheating. It is this latter theme that forms the substantive part of the research, through the development of a novel detection tool, ‘Clarify’, that uses a cyber security related digital forensics approach to the detection of plagiarism, contract cheating, collusion and the much more recent problem of artificial intelligence generated essays. Existing plagiarism detection tools focus on the textual content of the document, searching databases and repositories for text matches, and more recently, analysing the stylometric ‘fingerprint’ of the document to highlight anomalies in the writing style of the submitting student. ‘Clarify’ does not use this approach, but instead unpacks the document to its component parts, exploring the metadata within and in particular, the edit mark-up or revision save identifiers (RSIDs) that remain within the metadata to help build a picture of the way the document has been created. This in turn allows the assessor to establish whether the document has been created in an authentic manner, with an extended pattern of editing throughout the document, or whether it falls outside this norm by having large sections of unedited text. Other digital artefacts are also examined including font metadata, image information, editing time and more, and these details are reported back to the assessor via a web-based interface. The author has published a number of papers as a result of this work, one on student perceptions of academic integrity, two relating to digital forensics which detail the approaches used in developing ‘Clarify’, and a further two (one published and one in print) on the use of digital forensics and cyber security techniques to assist in the prevention and detection of academic misconduct, all of which have international reach. In particular, ‘Clarify’ as a tool has received international interest and as such is an important contribution to the field of academic integrity.
Date of Award | 2023 |
---|
Original language | English |
---|
Supervisor | Mike Reddy (Supervisor), Andrew Ware (Supervisor) & Susan Haywood (Supervisor) |
---|
Understanding academic integrity and plagiarism in the digital age: Can digital forensics techniques help prevent and detect academic misconduct?
Johnson, C. (Author). 2023
Student thesis: Doctoral Thesis