FROM ASSESSMENT TO LEARNING: INTEGRATING ITEM ANALYSIS AND LEARNING ANALYTICS FOR DATA-DRIVEN INSTRUCTION IN DISTANCE EDUCATION

Abstract

Distance education has changed assessment from reporting static scores to data-rich environments, yet item analysis is still mostly used for test evaluation rather than instructional improvement. This project aims to reposition item analysis as a core component of learning analytics for data-driven online teaching and learning. A thorough literature assessment of a topic framework for synthesizing empirical and theoretical work on item analysis, learning analytics and technology-enhanced learning guided by PRISMA across major databases. The synthesis yields three important findings: (2) validity-focused approaches like Item Response Theory make fairness and
interpretability better for different online learners; and (3) formative assessment and adaptive feedback mechanisms turn analytics into actionable instructional interventions that boost engagement and learning results. Current methodologies have limitations in terms of data quality, evidence of causation, interpretability, and challenges in ethical data governance. This work proposes to
incorporate item analysis in a learning analytics framework to improve instructional decision-making by combining psychometrics and behavioral analytics. The outcomes of the assessment, especially the item analysis, can be used for the quality assessment of the test, instructional design, personalization and equitable learning in distance education