I really enjoyed this paper which summed up many of the current concerns about higher education in a time of increasing external pressures and the need to somehow provide an improved, more efficient (and more appropriate) service to students. The authors discuss the value of learning analytics in helping us to get a better grasp of what makes students drop out and a better understanding of the learning approach itself. What made this interesting for me was the recognition that there is so much information out there that we now need to think about how we manage it in new ways.
This paper discussed the phenomenon of data abundance, or ‘big data’ and the need for institutions to make best use of it to clarify intention and act in more intelligent ways. It lists a number of ways in which big data can be used to generate added value, for example, by facilitating the use of what-if scenarios to model the potential impact of changes to learning design.
So far, so straightforward. But the authors talk also about the complexities of ensuring that we have captured all of the relevant data – for example, we may think that we know how a student engages online by tracking their work within our institutional VLE, but we may be unable to track equally valid or developmental engagement that goes on outside of our networks, via Twitter or shared blogs, such as this one on the #LAK12 mooc. If we are to make fuller sense of the data out there, we’ll need to look beyond our own immediate institutional boundaries too.
Further, the authors argue, it’s not enough to use learning analytics to simply predict student behaviours. We ought to exploit what it tells us to adapt both learning materials and delivery design. The impact of this approach is potentially revolutionary, with different students experiencing differing content delivered in different formats, as the teaching system takes note of the student’s profile and on-study behaviours.
I really like the ideas presented here, although find the idea of providing almost tailored curricula to be more than a little daunting – in order to get anywhere near this, institutions would need to make sufficient efficiency gains to justify the investment as well as be able to demonstrate that any tailoring had a beneficial impact on a student’s learning and understanding. There is an explicit recognition that none of this is easy and that there are a lot of questions yet to be answered, not least our tendency to try to create overly simple models which we may then be tempted to apply beyond their useful (or reliable) lives. In trying to adapt and change our teaching and support models for the better, are we in danger of simply replacing one inappropriate design for another?