Data-Driven Profiling of Reading Processes in Secondary Students: Learning Analytics Insights From PROLEC Assessment
Journal
Lecture Notes in Computer Science
Learning and Collaboration Technologies
Date Issued
2026
Type
Book chapter
Abstract
The integration of Learning Analytics (LA) into traditional psychometric assessments offers a transformative path for personalizing secondary education. This study proposes a “psychometric-first” analytical pipeline that converts item-level responses from the PROLEC-SE-R battery into actionable learner models for Interactive Learning Ecosystems (ILE). Using data from 256 Ecuadorian secondary students, we applied Categorical Principal Component Analysis (CATPCA) to address the ordinal nature of the psychometric data, identifying four latent dimensions of reading: lexical, syntactic, semantic, and integration processes. Subsequently, a k-means clustering algorithm identified three distinct cognitive profiles: At-Risk (12.0%), Average (56.8%), and High-Performance (31.2%). The internal validity of these clusters was confirmed through a Davies-Bouldin index of 1.206 and stability analysis via Jaccard bootstrapping. To bridge the gap between psychometric profiling and Human-Computer Interaction (HCI), we developed a rule-based mapping for UI/UX adaptations, where profiles trigger differentiated scaffolding—such as intensive audio-support for at-risk students and complex inferential challenges for high-performers. Results demonstrate that this scalable approach provides a theoretically grounded alternative to hardware-dependent analytics (e.g., eye-tracking) in low-resource contexts. This study contributes to the field of educational data mining by operationalizing latent cognitive features into dynamic instructional interventions, fostering more inclusive and adaptive digital learning environments. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.