Comparative Analysis of Learning Style Models for E-Learning: Validating the Felder-Silverman Framework Using Behavioral Data

Résumé

Personalization in online education requires sufficient modeling of learners’ preference and engagement behavior. As much as the theories of learning styles have been used to inform instructional design, there is still a concern on how it is used in behaviorally driven and mobile instructional environments yet to be explored in detail. The paper addresses this gap by undertaking a behavioral clustering analysis of 14,003 online learners, using the K-Means that generated six profiles of learners’ engagement. Four other common learning style models, VARK, Kolb, Honey and Mumford, and Felder-Silverman learning style model (FSLSM), among them, were tested against the clusters on the basis of tests such as behavioral alignments, system compatibilities, and correlations of the performance outcomes. Systemtraceable dimensions and statistically significant predictive power were indicated in the findings since FSLSM displays the strongest maps of behavior with a system trace, followed by high degrees of quantitative reliability. These findings have given an original but empirical basis for incorporating FSLSM into adaptive and mobile learning whereby real-time personalization is possible according to learner behavior. The research presents a data-driven framework of learner profiling that has been validated and supports intelligent mobile learning systems that adapt to learners in responsive ways.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0