This responsive piece was originally posted to the Teams channel for the Technology Enhanced Learning session on our CAPITAL course for lecturers.
When discussing the pedagogical value of online annotation activities, a concern that is often always raised is the potential for student misbehaviour. While our online polling service (Vevox) filters a comprehensive, truly international and frankly eye-opening lexicon of banned words, our social annotation service (Talis Elevate) does not. This means that there is the risk of a user posting offensive language. A profanity filter, however, would not stop offensive views from being shared (see Hansard, Question Time, most of the UK press).
We need to view the risks through the lens of the benefits social annotation activities offer:
- Social annotation supports rich disciplinary practices useful for students throughout their academic and career journeys, e.g., reading comprehension, close reading, textual analysis, language development and argumentation – practices that are essential, and complimentary to AI skills.
- Students value using social annotation in their courses because it can help them feel more connected to the texts they are reading and to the peers with whom they are learning.
- Asynchronous group-based online learning activities offer flexibility to an increasingly diverse and pressurised community of learners; and an alternative opportunity to contribute to and demonstrate understanding
- Teachers can better support student learning through the use of social annotation, in large part because it lets them better understand what, how and when their students are reading
The risks of individuals posting offensive comments is real. The likelihood of it happening is low but the impact can be high. Students and staff can be affected by this, and the purpose of the learning activities undermined. How can we mitigate against this in order to retain the benefits?
I am a firm believer in providing people with a chance to learn and to change. I do not expect I am alone here. There are parallels with the Gen AI conversations here. As a University we should be providing our students (and staff) to experience new ways of working, communicating and collaborating. Removing opportunities to engage in social learning or build resilience in an AI-dominated world is a dereliction of that duty. We should employ enabling tactics to counter and change poor behaviour.
My own research and practice in the use of online discussion fora showed that within a group of learners there can be a diversity of skills and wills. Understanding and harnessing these can underpin successful activities. I was interested in tackling the ‘free rider calculus’, in which non-engaged learners slowly demotivate highly active learners by benefitting from the latter’s effort and contributing nothing – or worse – in return. One way is to provide learners with specific roles in an activity. A moderator or ‘traffic manager’ role can not only maintain the enthusiasm of an engaged student but can act as an early warning system in the event of misbehaviour. With regular activities, this role can be assigned to different students. A recent research article from ALT identified three types of online learner: Pioneer, compliant & sceptic, which complements my findings and experience. The article also points to the desire among students for the co-construction of knowledge, which Talis Elevate clearly facilitates.
On a more practical level, we can weigh up the benefits and risks of offering anonymity. While this may encourage less confident students to post, and perhaps all students to post what they may think are ‘stupid questions’, it can embolden bad players. While we cannot disable anonymity, any student can flag any comment as inappropriate, and staff (or students with elevated permissions) can easily hide it.
Beyond this, whilst students can post anonymously, Talis retains the technical ability to identify students if required. If there is a comment made on a resource that needs formal follow up, or a student continuously abuses the anonymity, Talis can identify the student if requested. This is the same model used in our FutureLearn courses, which have thousands of learners worldwide, and in which our self-policing community has contributed to the maintenance of supportive and positive learning environments.
