From self-regulated to self-determined learning: identifying the heutagogical gap and designing GenAI scaffolds in asynchronous higher education
Asynchronous online courses require substantial learner independence, yet structural flexibility does not necessarily lead to self-determined learning. This study examines the transition from self-regulated learning (SRL) to self-determined learning and explores how generative artificial intelligence (GenAI) may scaffold heutagogical development. We conducted a secondary qualitative analysis of 304 SRL-coded meaning units from 75 preservice teachers enrolled in an asynchronous course. The data were coded using a 0–3 heutagogical-gap scale. The resulting patterns were then translated into a GenAI prompt repository, presented as an empirically grounded design output rather than a tested intervention. The heutagogical gap was defined as the developmental space between learners' capacity to regulate learning within a predefined structure and their capacity to define learning goals, pathways, products, and evaluative criteria more independently. Overall, 86.8% of meaning units reflected some level of gap, while 37.5% showed substantive or critical gaps. Patterns included dependence on external feedback, limited knowledge transformation, help-seeking characterized by isolation or dependence, and coping without explicit emotional regulation. The findings informed an AI-Enhanced Heutagogical Cycle (AIHC) aligning SRL dimensions, heutagogical transitions, and pedagogically constrained GenAI roles. The study distinguishes effective regulation within a given structure from self-determined learning. It proposes GenAI as a differentiated scaffold for expanding learner agency rather than substituting for learners' cognitive effort, judgment, and responsibility. The framework and prompt repository require validation in future intervention studies.