The Use of Models in Learning Design and Learning Analytics

Abstract

The practice of Learning Design (LD) and Learning Analytics (LA) is analysed using two lenses. Firstly, it is argued that both LD and LA involve the making of models. All models involve idealisation, i.e. the simplification of something complicated to make it more tractable. Various educational actors generate idealisations of the system they are working in, but only a subset of these idealisations is embodied in LD and LA implementations. Secondly, LD and LA both change the way that the actors in education handle the complexity which they are faced with in their practice. This is analysed in terms of the cybernetic concepts of variety and black boxes. It is then argued that LD and LA implementations share a tendency to shift control to higher levels in the hierarchy, and that this has consequences for the autonomy of teachers and learners, and for the acceptability to users of LD and LA. Finally, some conclusions are offered which can help LD and LA to be implemented while addressing the common problems that have been identified.

Learning analytics driven improvements in learning design in higher education: A systematic literature review

“…In 2017, the 'Interaction Design and Interaction(s)' journal released a special issue dedicated to LA and LD (No. 33), presenting various experiences of employing LA for LD, albeit some of the academics noted that such attempts were still uncommon (Griffiths, 2017).…”

“…LA and LD have great potential for mutual reinforcement, as they ‘both create and implement models of learning processes: they identify desired outcomes for a designated learning process and propose ‘if–then’ rules about how those outcomes are best achieved’ (Griffiths, 2017, p. 115). LA can help instructors gain insights into learning behaviour during the course and make improvements to the course design as soon as needed.…”