Doing what I do, I talk a reasonable amount about moocs, OER, learning analytics, the sorts of opportunity they provide and how we might do research on them. I’ve also done a reasonable amount of work on the Wikipedia pages – which I strongly encourage more people to edit, especially: * [http://en.wikipedia.org/wiki/Massive_open_online_course]1
- [http://en.wikipedia.org/wiki/Learning_analytics]2 Anyway, one thing we could ask in this ‘new’ area – especially moocs – is **what would the ideal mooc research institute look like? ** A not entirely flippant response is: “well, I work there”. The OU’s history in mass distance education, and their current work on social and semantic web technology, OER, LA, and recent step into the mooc space place it exceptionally well on a global stage for research in this area. Of course, another answer is – go check out the [lytics lab at Stanford]3. In both cases I really respect the work going on, and hope we (and the lytics group) continue to grow in our interdisciplinary, edu-tech work. Of course the [Society for Learning Analytics Research]4 is also working in a related area – although not specifically on moocs, and George Siemens (one of the people behind SoLAR) of course was also one of the first (if not the first) people (with Stephen Downes) to run a cmooc. I’m posting this as a ‘thinking aloud’ (thinking allowed) piece – I’ve missed things, my emphasis might be wrong, I’ll have neglected examples, and of course – the concept itself is probably flawed (better many good labs, unlikely to have one lab to rule them…), nonetheless, it was interesting thinking about how I’d most like to work (and with whom), and the sorts of questions I think are/are not being addressed at the moment, so if it helps others thinking about those issues so much the better. # Who? No one is quite sure what’ll come out of moocs – whether they “flop” or not will probably depend on how we measure their success, and the success of whatever comes next. Personally I’d like to see [remixable OER based moocs]5 opening up life long learning with local centres (and communities) engaging in study together, alongside the continued existence of universities with strong (pedagogically solid) teaching and research…it’s a bit of a dream but hey – it shouldn’t be. In any case, we’ll need theorists to be engaged here – whether in mooc-centre labs, or affiliated groups engaged in its work – thinking about models, social impact, sector-wide impact, etc. We’ll also want to play with data, with materials and structures, with participants. We’ll want to look at the people, the materials and content, and the networks – and the rich interactions going on. So who would we want for that? Well maybe the following (who else do people think?): 1. Data scientists – whether it makes sense to talk about ‘data scientists’ as a particular group I’m not sure, but broadly I mean people (and teams) with mathematical/statistical, and computing skills who are adept at finding and processing rich data sources, building visualisations of this data and tools for end users to process/visualise the data in meaningful ways. They should understand the data they’re processing, and the output to some extent but that’s probably largely driven by the following… 2. Learning scientists and psychologists – here I’m talking about the sort of CSCL people and psychology in education people who can contribute to understanding system design, technology supported learning, learning processes, etc. I don’t think everyone needs to be “techie”… 3. Technologists – data sources and learning environments in moocs are diverse and rich, bringing this together is in part about data science, but it’s also about interoperability, designing systems to display information from various sources, and perhaps designing new environments too (as has been common in CSCL) 4. Educators from disciplines – different faculties have different ways of working, different skills and knowledge, and established resources, obviously we need individual faculty members to run courses, but I’d think having some faculty specialists with consistent contact with mooc researchers would be useful a) for research and b) to provide guidance to institutional providers 5. Teacher educators – As a corollary to the above, some people with backgrounds in teacher education and online education would make sense 6. Policy, Philosophy and Economics – (also sociology) – as I said before, these things have impacts, we should care about that and those discussions should – to some extent at least – be happening within platform-research groups, not just by externals ‘looking in’ (although that should also be happening. It needs to be not just multi-disciplinary, but multi-vocal, and _inter_disciplinary. I know some labs where the tech people are somewhat separate from the others and not as included in discussions (multi-disciplinary), others where voices are heard from multiple perspectives and often productively so- but the teams basically work as two distinct groups. I don’t think a mooc research lab can operate like that. It should also be feeding back in, not only to mooc development and deployment but also offline teaching and learning – for that it needs to be integrated into the wider structures. # What? What sort of research would we do? Play with data, a/b testing, we might do face to face research (both of the ‘useability/HCI lab’ type, and in lecture/traditional delivery contexts) & apply to online. We might want to explore blended options, flipped classrooms, and improving f2f delivery too (particularly given the quality of at least some lectures…). We might want to look at: 1. Interactions between students. and between students and faculty, and the quality of these interactions 2. Paradata – while learning analytics is student centred, paradata is resource centred, and gives us some indication of how learning objects (OER, activities, articles, etc.) are used, how often they’re used, who by, what combinations they’re used in, what people are saying about them (and tags they’re adding), etc. 3. Dynamic understanding of knowledge, as not just static facts but flowing, constructed, connected 4. Data on how learner’s understand their own learning – whether we have formal assessment or not, this seems like a pretty important facet if we want to understand what students are making of this free educational experience 5. Theory – around learning, policy, philosophy, assessment (and lifelong learning), etc. # How? A lot of cool things are being done on learning analytics at the moment, but there’s also lots of work (e.g. at [CSCW]6) which is of relevance to learning analytics (and moocs) but not currently being deployed by researchers in the area. So to give a brief overview of some potential methods, sorts of analysis, etc., I’d expect to see: 1. [Discourse Centric Learning Analytics]7 – analytics that goes beyond simple counts of posts, to explore the nature and quality of interaction 2. [Dispositional analytics]8 – analytics on the characteristics of effective learners, how to support diverse sets of students, and how to spot particular ways of working in context 3. [Sensemaking]9 (including at multiple levels – from individual students right the way up to institutional) – research on supporting students (and course designers, etc.) to make best sense of data, to advance further and to support themselves. 4. Generally, process data, e.g. on self regulated learning, promisingness, epistemic behaviour, metacognition, working in particular subject-contexts, etc.
Where? – Lytics locals The issue of what a lab would literally
look like is an interesting one. I’d suggest that labs of this sort would need to be pretty well connected for livecasting and global collaboration to work with colleagues from a wide range of stakeholders. I also wonder, though, whether such a lab would best benefit from being located within a mooc-office, working with stakeholders from students to course providers. Perhaps one model is to have a slightly distributed lab, with “lytics locals” with course providers, and the platform, and coming together in labs to share and collaborate. I’d be interested to know what other people do at the moment and how well they think it works. # Issues Of course, such ideas have issues. Including: 1. What is the short/long term commitment of institutions to moocs, and if this is in question, how do they recruit people to work on mooc-research? How can we ensure long term impact of any research done in this context? 2. How do we ensure we’re not swamped by data (and that we’re working ethically with that data)? 3. How could such a research centre deal with processes such as the REF (or whatever comes next) which may mitigate against interdisciplinary research which can’t clearly be tied to one particular disciplinary-submission category
Footnotes
-
http://en.wikipedia.org/wiki/Learning_analytics ”http://en.wikipedia.org/wiki/Learning_analytics” ↩
-
http://sjgknight.com/finding-knowledge/2013/01/open-remixable-moocs-what-would-it-look-like-what-could-we-do-with-it/ “Open (remixable) moocs – what would it look like, what could we do with it?” ↩
-
http://sjgknight.com/finding-knowledge/2013/03/cscw2013-2-workshop-papers-texan-fun/ “CSCW2013 – 2 workshop papers & Texan fun” ↩
-
http://www.solaresearch.org/events/lak/lak13/dcla13/ “DCLA at LAK13” ↩