Information as Interpretation: Measuring Learning Behavior for Knowledge Insight

Traditional Learning Analytics (LA) has been primarily focused on the behavioral aspects of learners’ learning due to its data-driven nature, often lacking analysis of the knowledge-related aspects of what learners have learned. To tackle this issue, the Information Theory on Learning Analytics and Knowledge (ITLAK) framework uses information theory to quantify the informational value of learners’ learning behavior regarding their interaction with knowledge. This paper shows the foundation of ITLAK and its case study, demonstrating its theoretical validity and practical usefulness. ITLAK was applied to the context of English Intensive Reading (IR) specialized in English grammar learning, and a field experiment was conducted with Japanese junior high school third-year students using an IR system. The results showed the possibility that the information content calculated by ITLAK is an indicator that can capture the behavioral and knowledge-related aspects of learning. In particular, it was suggested that the information content metric functions in immediate feedback and captures aspects of learning distinct from the number of contacts with knowledge. This finding can indicate that ITLAK provides a theoretical foundation for advancing Knowledge-Aware Learning Analytics (KALA) and opens new possibilities for LA. Future research will involve revalidating the findings with large-scale data and designing learning support models.