Investigating the Effects of Adaptive AI-Generated EFL Learning Materials on Cognitive Load, Engagement, and Retention

Abstract

This study investigates the effects of adaptive AI-generated learning materials on cognitive load, engagement, and retention among university students learning English as a Foreign Language (EFL). Using a quasi-experimental design, 120 intermediate-level undergraduates were randomly assigned to either an experimental group, which used an AI-adaptive learning platform, or a control group, which relied on traditional textbook materials, over a four-week intervention. Cognitive load was assessed via the NASA-TLX, engagement through the Student Engagement Scale, and retention using immediate and delayed post-tests. The findings showed that students in the experimental group experienced significantly less from extraneous cognitive load, which refers to the unessential mental effort that does not support learning. In contrast, these students also demonstrated significantly greater behavioral, emotional and cognitive engagement with the learning process, as well as much better retention success compared to those in a control group. Quantitative feedback revealed informative evidence that such attributes as user-tailoring, real-time feedback, and interactive design were overwhelmingly identified as significant merits of the AI system in use. These findings offer strong empirical evidence of the pedagogical value of AI-driven adaptive EFL materials, highlighting their potential benefit for cognitive load management and learners' intrinsic motivation which enhances learning. The study highlights the great importance and contribution of AI-based personalization in higher education EFL classes, as a potential tool to serve better the individual needs and cognitive difficulties students face while learning at university.