Could viral marketing solve student absenteeism ?
Student absenteeism is a perennial problem in higher education. As a lecturer, I’ve frequently stood in half-filled auditoriums feeling powerless in the face of poor turnout. The issue is more pronounced in the world of online teaching. The attendance in my classes dropped from an average of 65% in pre-pandemic times to around 40% once teaching switched to remote, whilst with fully online cohorts attendance peaks at around 15%. Discussions as to a causal relationship between student attendance and performance are complex and without consensus. Nevertheless, for those of us that aspire to active learning, it is self-evident that non-attendees deprive themselves of those motivating and deep learning experiences which unlock a subject, and there is an intuition that many of those absentees are the very students who need these experiences the most.
The higher education playbook for dealing with absenteeism ranges from measures such as mandatory sessions, registration, warning emails, and reprimands to use of extrinsic motivators such as participation grades and in-class summative assessment. In times of desperation I’ve found myself conducting my own email marketing campaigns for my lectures to try and shift the stubborn pattern. It’s a lot of hard work, but the students stay at home.
Through its social media mechanics Handl offers potentially more organic, intrinsically rewarding, and effective means of getting learners to show up. In case you haven’t heard of Handl before, it’s an app for real time social learning — read more about it here. Over the past few months we’ve been conducting trials with a cohort of online students from University of London’s BSc Computer Science programme, and so we decided to see whether we could improve the attendance for one of our sessions through employing some social media mechanics in a viral marketing campaign. Now before you get too excited, I’m not about to reveal the silver bullet for absenteeism. The features we’ve developed are a naive first parse, and they’ve yet to prove themselves in the field. Nevertheless, I believe this approach has the potential to affect behavioural change and so I will describe what we did below.
One aspect of physical classes that online classes lack is the social interactions that occur outside of formal teaching. For example, the shared coffee before lecture, casual conversations whilst waiting in the hallway, or discussing homework in the lecture hall before the instructor arrives. These moments of free interaction offer many social possibilities for learners, and potentially intrinsically motivate them to attend. We therefore wanted to emulate these moments for the online classroom, and so constructed a three part campaign for our students.
Firstly we invited all learners to choose a social media avatar from a selection of 2000 images curated from flaticon.com. Once selected, the avatar becomes unavailable to other users of Handl making them a unique identity for the user. In order to strike a balance between identity and privacy, learners appear to each other as their avatars in conjunction with first names only.
The second stage was to invite learners to choose friends on the platform. Given the paucity of existing social connections within the cohort we provided recommendations based on previous shared encounters on Handl. The learners who had been friended received a notification of the event and an invite to friend back — the aim here being to create a viral spread amongst the cohort.
For the final stage we used the social graph produced by the friends feature to generate pre-session social groups. Each group was offered its own video call starting 20 minutes before the main session. Learners were invited to these via an email which showed them who else was in their group. Our hope here was that the personal connection might trigger a stronger incentive for turning up to the group.
In practice this first attempt at viral marketing didn’t significantly shift student behaviours. Attendance was slightly down from the previous sessions. Nevertheless, we saw some good learner engagement with these social features. 188 learners chose social avatars, 53 learners selected friends, and 70 learners attended pre-session socials. In the feedback, significant proportions of learners reported feeling encouraged to attend by these new features, and of our 95 attendees, 38 were attending for the first time.
As instructors, we found that some of the new features had benefits beyond raising attendance. The social avatars, livened up the look and feel of the platform and helped build familiarity with the cohort through associative memory. They were especially useful for navigating large numbers of study groups which could now be visually distinguished by their followers’ avatars. The pre-session socials seemed to loosen the students up before the session and make them more open to interaction.
We haven’t given up on achieving viral growth in attendance through Handl, and we’ve got plenty of ideas for improving our current model. The draw of pre-session socials could be significantly strengthened if invites came from other learners as opposed to the platform. This could happen before the socials start but also through providing shareable links to be sent through other platforms whilst they are in progress. The social graph yielded by the friends feature should become increasingly accurate as learners interact with it over multiple sessions. Allowing learners to add each other as friends during and immediately after study group calls would accelerate this process, and incorporating which friends a learner invites to a pre-session social or Study Group will finesse the graph. With increasing accuracy so increases the draw of the social groupings that learners are offered.
There’s still a long way to go with tackling absenteeism. However, by bringing social media mechanics into play, Handl opens up many possibilities for an altogether more pleasant approach to this most knotty of problems.