ABSTRACT
Making use of the idea of teaching as design, a new process model for ‘doing’ learning analytics is introduced where data are observed, analysed, reported, and acted upon. We apply a realist perspective to learning analytics in the classroom, the consequence of which is to emphasise both explanation and theory. A number of practical realist principles are presented, and these are applied to the case of a university statistics teacher looking to understand the effect of trialling some new pedagogical strategies in her courses. Throughout the chapter several educational theories are also applied to the case, including social cognitive theory and learning trajectories. The chapter offers a taxonomy of educational analytics that distinguishes between learner analytics and learning analytics, which is a potentially important consideration when designing personalisation at scale.
