Learning analytics in higher education – why we still have a long way to go.
Share this article
With regular prophesying of the possibilities of big data, you’re bound to come across the term ‘learning analytics’ (or ‘student data analytics’) as a key trend to look out for in EdTech. Perhaps you’re thinking that the onslaught of MOOCs (Massive Open Online Courses) will be the catalyst for the major evolution of analytics in education, well you’ll be disappointed.
Learning analytics – what is it?
According to the 1st International Conference on Learning Analytics and Knowledge, learning analytics is defined as:
“the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs.”
Basically, data scientists identify key trends in how students learn within online environments and use these trends to improve the education experience – institutions aim to create personalised education experiences based on how you learn best.
For more on the theory check out this bite-sized infographic about learning analytics from College Stats.
A loose definition.
We have a client that’s particularly interested in this area and asked us to take a deeper dig for anyone in higher education doing it well. In our research, we found a lot of universities that claim to be doing ‘learning analytics’, but the term is used pretty loosely.
“Generally, what you see is learning analytics applied to course management systems, and course management systems is basically just a fancy way of saying an online discussion board, and they are looking at how many times a student will respond, […] how quickly they respond to a discussion, how often they log on. So just very, very basic levels of data.”
Reynol Junco, Harvard’s Berkman Center for Internet and Society
From theory to practice – we’re simply not there yet.
Siloed systems, privacy issues, ethical issues, costs to set up the right systems, costs to set up staff – there are stacks of issues that need to be addressed before you can make learning analytics a reality by taking it from theory to practice, but the biggest issue of all is in identifying what data should be mined and analysed in the first place. Finding meaningful patterns to inform effective learning is difficult when there’s no strong definition of effective learning.
Using analytics at this level needs infrastructure and major leadership support
The closest example we’ve come across – although for the purpose of retention, rather than learning – is the work that University of Kentucky presented at EDUCAUSE last year. The primary data collection tool is a mobile app that students opt in to where they are served quizzes to check their stress levels etc. They also draw data from their Learning Management System (LMS – in this case, Blackboard) to check if students have hit certain milestones, like uploaded assignments, logged in to class etc. The app then sends alerts to students based on key retention signs (e.g. ‘Deadline reminder – don’t forget to upload your assignment’).
University of Kentucky’s approach is sophisticated, and they’ve gone the distance to get where they are today. They merged their entire research and business intelligent teams and hired 3 data scientists to make up a team of 15 FTE (full time staff) – that’s 15 FTE purely focused on developing and running their predictive analytics model to aid in student retention.
When many university staff are wearing 3 or more hats as it is – this is simply unrealistic for most universities.
If it sounds too good to be true, it probably is.
Be wary of any vendors promising what seems revolutionary when it comes to learning analytics – because frankly the software is only a minuscule part of the puzzle. Dig a little deeper into Universities claiming to be active in learning analytics and you’re unlikely to find anything substantial. But if you do, let us know! We’d love to know of any universities actively using learning analytics.
If you’re interested in more (and have an hour to burn) – check out this talk from Roy Pea, Professor of Education and the Learning Sciences at Standford University at EDUCAUSE last year – “Learning sciences and learning analytics: time for a marriage.”