A €7?million (?6.2?million) tool designed to stamp out cheating in online assessment is preparing for launch after successful European trials.
The Adaptive Trust-based E-assessment System for Learning (Tesla), funded by the European Union, combines anti-plagiarism software with facial-, voice- and keystroke-identification technology.
More than 27,000 students at seven European universities, including the UK¡¯s Open University and the Open University of Catalonia (UOC), have now taken part in trials of the tool.
Denise Whitelock, professor of technology-enhanced assessment and learning at the UK¡¯s OU, said that Tesla had ¡°had a?good response¡± and had proved to be a robust anti-plagiarism tool.
It will be developed into a product available for purchase by universities, albeit with a ¡°basic¡± version likely to be made available free of charge, after the research project concludes in March.
A key weakness of many existing online assessment systems is that they may allow a student to enlist a substitute to do work in their name by giving that person their username and password, or may allow a student to enter their details to start an exam and then have someone else take it for them.
Using technology to verify the identity of people sitting online exams ¨C be it through their appearance, the sound of their voice, their way of typing or their manner of writing ¨C could improve the reputation of online qualifications. It could also reduce online providers¡¯ costs by sparing them the need to conduct face-to-face assessments.
Anna Elena Guerrero, the project coordinator at the UOC, said that students initially found the facial recognition software to be ¡°the most intrusive¡± element of the tool. However, those involved in more than one round of piloting often ended up changing their perceptions. ¡°In the beginning they felt that facial recognition was very intrusive, but later that decreased,¡± she said.
¡°Mainly students saw Tesla as an advantage, because they can save time on commuting or they feel it makes their degree stronger,¡± Dr Guerrero said. ¡°It will also be good for companies that are going to hire these kinds of people [who take online courses] to know that we can identify authorship and that the qualifications are robust.¡±
Professor Whitelock said that universities could eschew biometric identification in the tool, but added that it might be best to employ it while making it ¡°very clear where this data will go¡±. ¡°Students are used to sitting in an exam room; we need to educate people about the software and explain the advantages,¡± she said.
Last year, researchers at Arizona State University reported success in detecting cheating in online assessments by tracking whether students tilted their heads to read material off-screen and whether they paused before answering questions.
Another OU research project explored using webcams to monitor students¡¯ engagement and emotions via their facial expressions or eye movements, in a bid to improve retention and course design.
