How learning analytics will transform universities

8 September 2016

Few university graduates will recall the task of filling out course evaluation forms with nostalgia. At the end of a class, with the prospect of being free to leave as soon as the page is completed, and little personal incentive to offer constructive comments, the process is not designed to succeed. Lecturers complain the comments can be discouraging and hurtful, while some have even called for an end to the practice.

Like in so many other domains, big data techniques offer a solution to this problem. By tracking the digital footprint of students, ‘learning analytics’ gives lecturers and tutors a rich insight into the level of class engagement on a week-by-week basis; how much students use online resources, participate in forum discussions, take books out from the library, and even the amount of time spent on campus. This data is then crunched to create personalised predictions of each student’s performance, creating – as a Reform report published today highlights – a series of opportunities for university faculty and the sector as a whole.

First, education will become more personalised. Tutors will be notified of struggling students, creating the opportunity to design targeted interventions. At some universities, students have access to apps that trace their own weekly activities, contrast their performance with that of their cohort and send notifications if they are at risk of dropping out. Not only does this add a competitive element; it also gives students more realistic insights into how they are doing, if they are putting in enough effort and what areas of work they may want to increase.

Whether data is supplied to faculty or fed back to the students themselves, insights from learning analytics have delivered results. A learning analytics pilot run by the Open University saw retention rates increase by 2.1 per cent, adding an estimated £1.8 million in additional income; a study from Nottingham Trent University found 27 per cent of students that were given access to their learning data changed their behaviour, for example by increasing their attendance.

Second, learning analytics will help policymakers in their ambition to evaluate teaching excellence.  By measuring how much students access resources, they can understand the extent to which lecturers are inspiring their students to engage with the course material, and more generally, the quality of tuition. If matched with demographic data, these tools can also be used by universities to understand the challenges faced by students from disadvantaged backgrounds. In its report on the Teaching Excellence Framework, the Business, Innovation and Skills Committee recommended the adoption of a metric that could highlight exactly this aspect of university performance.

Without doubt, universities that successfully embrace these innovations will be best placed to meet the current, challenging climate. The bottom line of the higher education sector is under threat from the dwindling number of 18-year-olds, as well as the emergence of high-value apprenticeships. Those at the bottom of the league table have most to fear, with the Universities Minister Jo Johnson MP suggesting last month that underperforming institutions will not be able to offset future reductions in the teaching grant through further rises in tuition fees.

Students are increasingly sceptical about the value for money represented by tuition fees, so the least policymakers can do is ensure universities are offering genuine quality. If learning analytics became a default part of measuring university performance, policymakers would be one step closer to achieving this goal – as well as confining those tiresome evaluation forms to history.

Emilie Sundorph, Researcher, Reform

Comments

Comments

No comments yet.