Turnitin’s annual User Summit was held on October 19th in Newcastle. They have noted that they are trying to address more of the needs of UK HE, and the keynote was by Kerr Gardiner , discussing the main themes from his consultation with a range of UK Universities (including Dundee). Then followed the obligatory technical presentation, on the directions that Turnitin are taking. This addressed most of Kerr’s points, which looks like potentially many of the things we’d like (such as more flexibility for different workflows) are on the horizon. I, and others, are hopeful that these things will appear in the future. Following, they then moved to the the main focus of the day which was the need to enhance academic integrity among the student body.
One key concern that was raised at several points during the day was that of Contract cheating, and many of the breakout sessions covered this, or related issues.
At York, Stephen Gow is the Academic Integrity Resources Manager, where his role focusses on the education of students (and staff) about what academic integrity is, rather than having a punitive role. As such, all students have to take a mandatory academic integrity online tutorial in their first semester.
Bill Loller started to look at some of the technical solutions to detect contract cheating – it’s much more difficult than text matching, as it requires more complex algorithms, and there are far more variables. For example, a focus on stylometry only works if multiple authors are being presented as the same. But, that falls down if the same author is used for multiple submissions – even if the author and the submitter aren’t the same person.
The final optional session that I attended moved away from this slightly, to look at issues of similarity in things other than text. Some are clearly easier to detect than others – Fintan & Lancaster (2004) looked at Computer code long ago Images are however, are much more difficult to work with, while tools such as Google reverse Image search attempt to find similar images (usually reasonably well), it requires a return to the similarity detection some will remember from before Turnitin was so ubiquitous – putting a phrase into a search engine of your choice (probably Alta Vista at that time!)
The final session was a series of shorter presentations looking at a range of different approaches globally. Stephen Bullock gave an overview of what the QAA are doing, in particular the recent report and guidance they have published. Irene Glendinning reflected on the work she and other have done looking at academic integrity across Europe, and, most recently, the development of a Scorecard for Academic Integrity Development. There is more detail (and a legible slide!) on a longer presentation.
Cath Ellis then introduced the work that she has done with others in Australia, mostly following the press coverage of the MyMaster cheating scandal in Sydney – though, as with all others presenting, they had been looking at this for much longer, it was the press that raised awareness – both in the general public, and, often among other academics.
One issue that Cath’s team found in the surveys they did in Australia is that students were more likely to engage the help of family members and friends to help with coursework, than a stranger – which means that the stylometric approach to detecting contract cheating may not always work. I’d agree with Graeme’s view that it was Cath’s presentation that was the most valuable of the day.
Finally, Phil Newton looked at the way that contract cheating sites operate, and their proliferation over the years.
Overall, a key theme that came out of the day for me is the need to educate – both students and staff. The former so they know what constitutes academic integrity, the latter to look at ways to create assignments that are difficult to plagiarise (and, ideally, authentic and engaging for students, so they’re not tempted to buy “another essay”).
[Apologies about the poor quality of the photos – I was quite a long way back and was trying to use a small camera … ]