What could Learning Analytics learn from HCI theory?
It’s good to share this chapter that’s been brewing for about a year now, with helpful feedback from quite a few colleagues en route, gratefully acknowledged. I go back to my roots and share some viewpoints from HCI on my current field of Learning Analytics. This preprint will appear (subject to minor production edits) in the forthcoming book:
Buckingham Shum, S. (In Press). What could Learning Analytics learn from Human-Computer Interaction theory? In: Kathryn Bartimote, Sarah Howard & Dragan Gašević (Eds.), Theory Informing and Arising from Learning Analytics. Springer Nature
Abstract: The design of Learning Analytics (LA) tools is an example of the general problem of designing interactive tools, which is the focus of Human-Computer Interaction (HCI) research and design practice. LA as a field must understand how to embed LA into organisations and the design of effective, trustworthy human-computer systems is where HCI theory and practice have much to offer. Consequently, this chapter argues that LA can learn from (i) the way that theory has evolved in HCI, (ii) the field’s methods for evaluating interactive systems at different scales, and (iii) HCI debates how established scientific theories and methods relate to design theories and methods. As a highly interdisciplinary applied field, LA (like HCI) faces the challenge of maintaining academic standards in the conduct and review of research from many disciplinary traditions. I propose that HCI offers inspiration for researchers seeking rigorous methods to design and evaluate LA in authentic contexts, including principles to maintain their intellectual rigour, which will also be of interest to LA journals and conferences seeking to maintain peer review standards.
