Modeling Learning Practices Informed by Mindset Theory for Adaptive Educational Systems

Abstract

Dweck’s Mindset Theory has been widely used to explain how learners approach challenges and effort; yet, applying this theory to guide pedagogical feedback in adaptive educational systems remains challenging. Learners’ mindset beliefs are often inferred from self-reports or survey-based reports, which may not consistently reflect how learners act during learning activities, particularly across different contexts. Moreover, commonly used indicators such as participation or engagement primarily capture how much learners interact with a system but provide limited insight into how learners respond to challenge, distribute effort, or persist during practice, key constructs of mindset theory. We propose a behavioral model of learning practice inspired by the core constructs of Dweck’s theory at the level of observable learning practice. Such a model can support adaptive educational support by using learners’ observed learning practices rather than relying solely on survey-reported mindset. Our model is derived from interaction logs in a quiz-based, gamified programming practice environment and characterizes practice patterns along observable dimensions, including challenge diversity, effort distribution, persistence, and contribution. These dimensions are used to describe four practice types inspired by the mindset theory (Strong Fixed Practice, Fixed–Growth Practice, Growth–Fixed Practice and Strong Growth Practice). This typology remains at the behavioral level and does not claim to infer learners’ underlying mindset beliefs. We show that these practice types are associated with distinct participation patterns that extend beyond engagement. We discuss how modeling learning practice in this way can support the choice of pedagogical interventions of adaptive educational systems grounded in mindset theory.