E-assessment system to eliminate e-learning cheating
Although online learning offers opportunities for people unable to benefit from onsite opportunities – due to distance, access or disability – institutions remain somewhat sceptical. One reason is the perceived lack of reliable e-assessment systems (authenticating identity and authorship) which can be integrated into current Virtual Learning Environments (VLEs) in a scalable and cost-efficient way. The EU-supported Innovation action project, TeSLA, developed an e-assessment system, complemented by development frameworks, which meets European educational and technological requirements, as well as quality assurance, privacy and ethical commitments. TeSLA utilised instruments like facial and voice recognition, keystroke dynamics, forensic analysis and plagiarism tools. Frameworks were developed for: education (with seven learning scenarios), legal and ethical concerns (including FAQs and data security protocols), and quality assurance. Authentication and authorship TeSLA evolved from a Proof of Concept developed by the Open University of Catalonia into a system offering expanded functionalities and an architecture designed for supporting many students and for integration into a range of VLEs. After students sign a consent form, TeSLA gathers, stores and analyses their personal data and obtains their biometric profile for authentication purposes. Students proceed with regular learning activities, with the system automatically collecting data for comparison with their enrolment details. “TeSLA is a unique education solution because it uses multiple authentication and authorship technologies simultaneously,” says project coordinator Dr Ana-Elena Guerrero. The system has three main alerts: green indicating no concerns, orange alerting for difficult cases requiring human intervention, and red for potential cheating, when evidence is sent directly to teachers. TeSLA allows teachers to select the most appropriate functionalities for each learning activity. For example, if an activity includes video, they can activate the facial and voice recognition tool to capture data. Analysis of the collated data can be shown to teachers through a dashboard visualisation tool as a safeguard against cheating. The system can also provide benchmarking information for classrooms of the same course. TeSLA was developed for integration into VLEs regardless of their underlying technology. Where this isn’t possible, TeSLA comes with a Learning Tools Interoperability (LTI) module, meaning that it can be used as an external application. The developers also created a Moodle plugin which retains all the TeSLA features. Seven large-scale pilots have been conducted in higher education institutions, together with six academic and technical third-party pilots. These involved 22 941 learners (including 861 students with special educational needs) and 457 teachers, with 532 activities within 310 courses. “TeSLA was found to be scalable and robust. Students reported being satisfied and didn’t find it too intrusive, saying it wasn’t only for institutions, but also students, showcasing their ability. Teachers affirmed TeSLA’s reliability against fraud,” says Dr Guerrero. Flexible and lifelong learning By preventing plagiarism and fraud, TeSLA not only promotes academic rigour, but contributes to a more diverse European higher education sector, creating new opportunities for flexible and lifelong learning. The open source version of TeSLA is currently available and comes with most of the features of the commercial version (some have intellectual property limitations). The team are working to make the cloud-based commercial version ready, based on FOSS licences, which will be deployable within VLEs using a web interface. Longer-term, the team will integrate more VLEs into the system, as well as adopt cutting-edge supporting technologies, such as blockchain and artificial intelligence, to improve the veracity of e-assessment biometric instruments.
Keywords
TeSLA, e-learning, e-assessment, authentication, authorship, cheating, fraud, online learning, Virtual Learning Environments, learning, Higher Education