I wanted to gather some of my thoughts here on some of the interesting observations from the conference – they’ll be quite unconnected, but I hope interesting!
Justin Reich talked about how learners engage in political MOOCS, particularly looking at how groups of learners with different political leanings might engage with each other. While the talk itself was very interesting, I found the discussion and questions afterwards most interesting. Justin shared an informal result from a project of one of his students, who was looking at how learners are instructed to engage with each other, finding that across 25 MOOCS, no course provided explicit instructions on how learners should communicate with each other, or how they should engage with their learning community. Given that many of our MOOC learners has had little or no prior experience with online learning, this is an extraordinary result, but reflects our face-to-face learning environment assumptions, where we have had little prior requirement to direct learners in this way.
Justin also talked about one of his research goals,which is one that we share – the idea of recommender systems designed to help learners find a cohort with common learning goals within the larger MOOC comfort. Our work on personalised learning and discussion forum analysis aims to support this – our initial results document our work in automatically analysing discussion forum data to group discussion threads into topics, but then to do further analysis of course content to automatically label those topics in a meaningful way. From this work, we have created visualisations that help learners navigate these topics. These visualisations can also be used by academics to explore what topics are being discussed by learners, and how topics are changing over time.
Stephanie Gottwald and Tinsley Galyean talked about their work in exploring literacy development at scale – distributing tablets filled with apps that build fundamental literacy skills to areas with local learning challenges. They discussed 3 case studies, a school in Alabama with little access to pre-school development, a community in Ethiopia, with no formal schooling, and a further school in South Africa, with limited teacher training and support. It was fascinating to see the students in each community explore the apps, each quickly building technical competency, and highly engaged with such increased access to learning support. Some of the interesting observations that they made, included that often the spread of learning was instigated by a critical group of high performing learners, able to quickly navigate the materials, and facilitators, possibly other learners who take a role in helping to spread information. Time in app is also correlated with learning, and so designing apps that keep learner attention, and having enough engaging and correct content to keep them engaged over time, is critical.
There were some interesting questions related to whether this kind of system is in conflict with what we know about the development of self-regulated learning, and whether the degree of scaffolding in these systems is too high. While only briefly addressed due to time, it was noted that this kind of system forms only part of a learning environment, but that is also targeted at a specific environment where there was little to no formal support.
Ben Gelman from the EECL Lab at George Mason shared work exploring an analysis of subgroup clustering in a Scratch community, identifying that clusters formed amongst the users, and that they tended to move further away over time. They also identified a temporal aspect, showing that users in older clusters tended to no longer active, and that new users tended not to follow projects from older or inactive users. Most users stay within the cluster that they initially identified with, which was usually theme based. Is this a consequence of the way that Scratch works? I.e. That when you first engage with Scratch, you are shown the most recent projects developed by others?
Ben did identify something that we have also seen in our teacher development MOOCS, in that it is important to provide spaces for new learners to create their own persona, and to establish themselves as part of the new community. An interesting question asked for views on the purpose of following, based on the idea that learners will follow new people in bursts, associated with when they want to learn something new. This made me think that navigating communities takes two forms, when wanting to connect with those in your sub community, and follow the general community, and when wanting to pursue new knowledge, which is unlikely to be met by your existing community. How can we support both patterns of engagement, and the idea of switching between the two?