Progress and Pitfalls in Personalisation

One of the strategic objectives of the university that I work for specifically addresses the need for personalisation. I'll admit that at first I was a little perplexed. What exactly do they mean when they say 'personalisation?' Aren't we supposed to treat all students equally? Turns out, that personalisation has more to do with customising the 'learning experience' of each student to reflect their different needs, goals, and abilities. But it much more than that. In order to personalise, you need to know a bit about the person, yes? So a large part of the university's strategy in achieving an excellent personalised experience is to "record, support and grow student participation in cocurricular activities."

Over the last couple of years a lot of discussion has been derived from the topic of personal data and how it is used by governments and organisations, both responsibly and otherwise. For example, as information professionals, the "technologically neutral" Snoopers Charter would have us collect and retain the private information of our students and visitors under the catch-all of national security. A little more recently, the Higher Education Commission released their report on the wider use of student data within institutions. The report From Bricks to Clicks: the Potential of Data and Analytics in Higher Education, does much to highlight the benefits of using student data to its full potential. The field of Learning Analytics is described as being replete with transformative potential, as a driver for engagement, and as an important tool in the ever-present quest to improve the student experience. The Open University does just this, with Professor John Domingue explaining that "predictive analytics can identify which students may not complete their course on time or even hand in individual assignments," claiming that OU has already implemented a system that identifies students who are at risk of falling behind in their programmes. What steps are taken once at-risk students have been identified remains unclear.

The University of London admits that "It is also important for institutions to be honest about their objectives in the use of learning analytics," Something that perhaps not all HEIs have fully grasped. This is especially important because "Some goals will be student-centric while others may be more institution focused." Or in other words, the institution potentially derives more benefit from the use of student data than the student does. Learning Analytics is big business. Paul Feldman, Cheif Executive of JISC raphsodises that "The prize here is enormous." And well he would; JISC has recently invested over £1m in projects to leverage Learning Analytics. The resultant service will operate a 'freemium' version of the Student Insight system.

The HEC report does highlight and recommend the need for clear ethical policies and codes of practice and JISC developed a Code of Practice for Learning Analytics in June last year. Citing Kant's Categorical Imperative Pistilli and Willis posit that we have duty to act if Learning Analytics can reasonably predict the need for an intervention. Willis later asks if inaction isn't more dangerous; "Is it unethical for an institution not to readily offer support when it can identify students who might benefit from various resources?"

Not everyone agrees that Learning Analytics will improve either student experience or institutional effectiveness. For example, Prinsloo, Slade, and Galpin explain that just like any other project in Higher Education there is a "danger that learning analytics might become part of the broader bureaucratisation of student learning." And I can't help but wonder what it feels like to be a student at OU (or anywhere else for that matter) who is contacted and offered additional services that they haven't asked for based on information they didn't know was being collected. Researchers at Northern Arizona University refer to this as "intrusive advising" whereby student trust in the institution's ability to respect their privacy is eroded as a result of unsolicited interventions. Is that the kind of personalisation we want? Somehow, I don't think so.

Higher Education Commission (2016). From Bricks to Clicks - The Potential of Data and Analytics in Higher Education. London: Policy Connect. Retrieved from: http://www.policyconnect.org.uk/hec/research/report-bricks-clicks-potential-data-and-analytics-higher-education

Pistilli, M. D., & Willis, J. E. (2013). Ethics, Big Data, and Analytics: A Model for Application. Educause. Retrieved from http://er.educause.edu/articles/2013/5/ethics-big-data-and-analytics-a-model-for-application

Prinsloo, P., & Slade, S. (2013, April 1). An evaluation of policy frameworks for addressing ethical considerations in learning analytics. ACM. Retrieved from http://oro.open.ac.uk/36934/1/prinslooslade%20shortened%20paper%20Final%2010%20March.pdf

Prinsloo, P., Slade, S., & Galpin, F. (2012). Learning analytics. In Proceedings of the 2nd International Conference on Learning Analytics and Knowledge - LAK ’12 (p. 130). New York, New York, USA: ACM Press. http://doi.org/10.1145/2330601.2330635

Sclater, N. (2014). Code of practice for learning analytics: a literature review of ethical and legal issues. Retrieved from http://repository.jisc.ac.uk/5661/1/Learning_Analytics_A-_Literature_Review.pdf

Willis, J. E. (2014). Ethical Discourse: Guiding the Future of Learning Analytics. Retrieved February 3, 2016, from http://er.educause.edu/articles/2014/4/ethical-discourse-guiding-the-future-of-learning-analytics