MOOCs: learning about online learning, one click at a time
This article was written by Gregor Kennedy
Massive Open Online Courses, or MOOCs, took the world by storm in 2012. After years of experimentation at the fringes of higher education, prestigious universities from around the world progressively surged towards MOOCs, developing free online courses that were open to anyone, anywhere, with access to the internet.
Almost as quickly as universities climbed aboard, the backlash began. Commentators criticised the teacher-centred, content-heavy online pedagogy often seen in the design of MOOCs. Thorny questions arose, and still remain, about the certification of students’ learning in MOOCs, and how this fits with the degree-awarding business of universities. And MOOC providers themselves have yet to articulate a stable business model.
No doubt these conversations will continue, adding spice to the question of why internationally recognised universities – often elite institutions with strong and valuable brands – so quickly gravitated towards MOOCs in the first place. Answers to this question vary. Some universities wanted to be at the vanguard of a new educational movement; almost all saw the power of MOOCs in reaching students that they might otherwise never have engaged with.
But one reason institutions like Stanford University and Edinburgh University embraced MOOCs was that they provided a wonderful opportunity to learn about online learning. These institutions recognised that MOOCs were a vehicle for educational research, particularly through the use of learning analytics.
What can learning analytics tell us about online learning?
Learning analytics use the digital data trails that students leave in online learning environments to develop an understanding of students’ learning processes. Every video watched, quiz answered and comment posted can be tracked, mined and analysed to better understand how students are learning online. Researchers are able to capitalise on the big data sets generated by tens of thousands of MOOC students to uncover productive and unproductive patterns of learning behaviour.
These patterns can be related to a range of other variables such as students’ socio-economic or cultural background, their previous education and prior knowledge, and their motivation to study. They can also be used to predict when students will drop out, whether they will pass the course, or whether they will get a high distinction.
Learning analytics are also incredibly helpful in informing our design of online courses. For example, the Learning Analytics Research Group at the University of Melbourne published a paper this year that showed how different curriculum structures – linear or open – impact on the degree to which students are inclined to engage in the very useful learning strategy of revisiting and reviewing content and assessment tasks they have previously covered.
But the great promise of learning analytics lies in their ability to genuinely personalise students’ learning experiences. While educational technology has a history of “intelligent tutoring systems” that “adapt” to students, MOOCs have the potential to take this to a new level.
The massive data sets associated with MOOCs allow researchers to look for both explicit and hidden patterns in students’ interactions online that can predict both productive and unproductive learning behaviour. This understanding can then be used as the basis for real-time interventions with students. If analytics can be used to recognise complex patterns of learning behaviour, automated interventions can be triggered, which provide students with advice and support.
In its simplest form, this advice can point students to material they may have missed or might like to cover. But the holy grail for analytics researchers is to use sophisticated data analysis and interpretation techniques as the basis for advice to students on how they are using or covering learning material; their approach to it, the learning strategies they are employing.
The outcomes of this learning analytics research, and the research approaches themselves, can be transferred and translated into more traditional and mainstream online learning environments routinely used by universities. The benefits of MOOCs are not simply in the MOOCs themselves, but in how they are informing what we are doing in mainstream campus-based and online education.
Only the brave will make solid predictions about where MOOCs will end up and what they will become. Regardless, MOOCs have already shown how learning analytics research can inform our understanding of students’ engagement and learning in online environments. This knowledge and understanding will become increasingly important, as it is clear that online learning is destined to play a major role in the future of higher education.