Learner agency across scales

An integrative perspective on self-regulated learning in algorithmically mediated environments

Abstract

As digital and AI-mediated systems increasingly capture data about learners and use it to make decisions about how their learning is supported, questions of human agency become central. Self-regulated learning (SRL) provides a multidimensional account of how learners take control of their development. Yet although foundational SRL models attended to learners’ broader life concerns, the field has progressively narrowed to task-focused regulation, and its dominant operationalisation in educational technology has narrowed further still, to behavioural proxies amenable to data-driven intervention. This paper proposes an integrative perspective that extends SRL by connecting it with two adjacent theoretical perspectives: complex systems science and Freirean critical pedagogy. The perspective foregrounds three dimensions of agency: regulation, integration, and positioning. Together they describe a learner who is not merely responsive to learning environments but capable of understanding and acting on the structural conditions shaping their experience. A cross-scale scaffold is presented: a design orientation that supports learners to move between regulation within immediate tasks, integration of patterns across their educational trajectory, and positioning in relation to the structural conditions of their learning. This perspective contributes a theoretical integration across SRL, critical pedagogy, and learning analytics that has until now remained disconnected.