A Theoretical Framework for User Mental Models in Learning Analytics Systems

Learning analytics systems are increasingly being utilized in teaching and learning environments to support data-driven feedback provision and learning outcome improvement. However, the effectiveness of these systems depends not only on analytical accuracy but also on how users perceive and utilize the provided information. Existing research has pointed out that the diversity in learners' ways of perceiving and utilizing systems can lead to misalignment between system functionality and user expectations. Nevertheless, systematic research on how learners construct mental models of learning analytics systems and how these can be utilized for user-centered design remains insufficient. This study aims to explore the user mental models formed during the process of using the TPIS system and to propose a framework that can systematically analyze and construct them. The proposed framework classifies cognitive elements that appear in user-system interactions and provides a theoretical foundation for systematizing mental model research in the context of learning analytics systems. Through a case study of the TPIS system, this study derived a conceptual framework that classifies users' mental models in learning analytics systems along two axes: Individual–Universal and Observable–Unobservable. Based on this framework, four mental model types (Operational, Situational, Strategic, and Interpretive) are presented to explain various aspects of user–system interaction. The proposed framework explains user cognitive processes in the TPIS system context while simultaneously providing a practical foundation for reflecting user experience in future learning analytics system design and improvement processes. Through this, the present study expands the discussion on mental models in learning analytics system research and presents the possibility of developing user-centered design strategies.