Learning analytics progressively finds itself at the intersections between institutional financial viability, the provision of effective, supported learning and an explicit respect for student privacy and agency. Such vulnerabilities should be acknowledged, together with the risks associated with issues such as the (over) collection and misinterpretation of data, the lack of algorithmic accountability, and the increasingly skewed quantification of learning. Set against this, of course, is the positive potential offered by learning analytics to support more appropriate connections between students, their aspirations and learning trajectories and the structuring and support of their learning in effective and resourceful ways.
Learning analytics allows for the allocation of institutional resources to design and deliver effective learning opportunities using a range of student data in order to optimise institutional efficiency and student support. Given the finite resources available to deliver proactive student support, there are growing questions around the practical issues around appropriate resource allocation and the need to put in place decision-making processes which also reflect institutional priorities and principles. The classification of individuals according to risk and their potential for recovery is well-established in medicine, and the notion of educational triage (Authors, 2014) provides a useful and critical lens on the potential, risk and ethical dilemmas inherent in the analysis and differential use of student data to drive support.
This paper explores the epistemological and ontological assumptions underlying and informing the connection between learning analytics and educational triage. We provide a conceptual framework and guiding principles for educational triage as useful heuristic tool to connect strategies to ensure educational sustainability and the moral obligation to support student success. The framework draws on existing policy and codes of practice (e.g., Open University, 2014). It recognises and reflects the juxtaposition and tensions between the (assessed) potential to positively influence student success, the number of students flagged as requiring (additional) resources and care, the scope of care required, and the cost and resources available for data-informed interventions.
Open University 2014, ‘Policy on ethical use of student data for learning analytics’, viewed 24 March, 2016, https://learn3.open.ac.uk/mod/url/view.php?id=85812.
Authors 2013, ‘Learning analytics: Ethical Issues and dilemmas, American Behavioral Scientist, vol. 57, no. 1, pp. 1509–1528.
Authors, 2014, ‘Educational triage in open distance learning: Walking a moral tightrope’, IRRODL, vol 15, no. 4, pp. 306-331.