e‐R esearch and learning theory: What do sequence and process mining methods contribute?

Abstract This paper discusses the fundamental question of how data‐intensive e‐research methods could contribute to the development of learning theories. Using methodological developments in research on self‐regulated learning as an example, it argues that current applications of data‐driven analytical techniques, such as educational data mining and its branch process mining, are deeply grounded in an event‐focused, ontologically flat view of learning phenomena. These techniques provide descriptive accounts of the regularities of events, but have limited power to generate theoretical explanations. Building on the philosophical views of critical realism, the paper argues that educational e‐research needs to adopt more nuanced ways for investigating and theorising learning phenomena that could provide an account of the mechanisms and contexts in which those mechanisms are realised. It proposes that future methodological extensions should include three main aspects: (1) stratified ontological frameworks, (2) multimodal data collection and (3) dynamic analytical methods.