18 March 2020
Innovative steps in tackling contract cheating at the University of Northampton
Dr Robin Crockett
Reader in Data Analysis and University Lead on Contract-Cheating, University of Northampton
I started as an Academic Integrity Officer at the University of Northampton in 2012 and had something of a baptism of fire: my second investigation was a contract-cheating case which necessitated a self-taught crash-course in assignment forensics.
That marked the start of a voyage of discovery, allowing me to develop expertise outwith the familiar (for me, at least) territory of time-series and Fourier analysis. That expertise, plus the increasing scale of the problem, led to me taking on the role as University Lead on Contract-Cheating, along with my research readership, in September 2019.
Contract-cheating is not new, nor is wider academic cheating: in the Ming and Qing dynasties, students carried cheat-sheets into civil service exams. In the era before word processors and the Internet, ghost-writers advertised their services, honest or otherwise, via word of mouth or small ads in newspapers, for example, and students commissioning assignments would likely have had to re-handwrite from purchased documents, requiring time and effort. Since the advent of word processors and the Internet, everything has become so much simpler, cheaper and, sadly, more commonplace and sophisticated...
So, what’s my role? In a nutshell, I investigate and co-ordinate investigations into allegations of contract-cheating and mentor and train colleagues at Northampton and the Midlands Integrity Group. I also have a research brief – new techniques to help identify commissioned assignments.
That’s where my background comes in – my research focus here is stylometry as applicable to student assignments, forensic stylometry if you will. In some ways, it’s not much different to time-series analysis: some of the datasets stylometry looks at are essentially sequences of numbers, as are time-series. The outputs and consequences are, of course, very different... Yes, Northampton has Turnitin software, and it’s good, but it can’t – and never can – do everything. We (the sector) need to engage with and contribute to the QAA and its comprehensive advice and adapt, develop expertise, train and – importantly – resource specialists to do the heavy-lifting that’s often required.
So, what’s stylometry and how is it useful? Stylometry can be defined as the statistical analysis of writing style. In this context, if stylometric analysis reveals two or more inexplicably different writing styles among a student’s submitted assignments, how likely is it that the student in question wrote everything? We all have individual writing styles, some more distinct from others. Sometimes it’s vocabulary (‘outwith’...), sometimes spelling convention, sometimes it’s more complex patterns of word usage.
Stylometry can be powerful: in 2013 Patrick Juola at Duquesne University in the USA used JGAAP, software he co-wrote, to identify JK Rowling as the likely author of The Cuckoo Calling. A high-profile example, perhaps, but it illustrates the types of tools, skills and expertise the sector needs. Personally, I use the R statistical computing software that’s been my mathematics-statistics software of choice for many years – and it’s free and open-source (FOSS).
We (the University of Northampton) don’t have a ‘statute of limitations’: we reserve the right to investigate suspected contract-cheating at whatever point evidence comes to light, including after graduation, and in that eventuality rescind awards if contract-cheating is judged to have occurred. Indeed, it was an increasing awareness of that possibility that played a part in the University’s decision to appoint someone – me – to lead and co-ordinate its policy and practice. Academic Integrity Officers are common within the sector but I think I am one of the first, if not the first, to be given a specific role – and time – to lead on contract-cheating. It’s a start, not an end, but it reflects the seriousness with which we regard the problem and, I think, serves as a pointer for other universities.
On a closing duty-of-care note, there’s another, even darker side to contract-cheating. Once a student is on a database somewhere as having commissioned an assignment, that information is potentially available to extortioners and blackmailers – forever. Ideally, every student could be deterred from cheating but, realistically, it’s inevitable that some students will be tempted – sometimes misled, deceived – by essay-mill or ghost-writer adverts. I don’t want to see any students drawn into cheating, let alone put themselves at risk of extortion or blackmail. Thus, it’s important to detect contract-cheating sooner rather than later, and get students back into studying with integrity before cheating becomes a habit.