Adaptive Learning Platforms in English Education: Enhancing Comprehensive Skills and Social Interaction in the Digital Classroom

Abstract

Advances in computer-assisted‑ language learning (CALL) promise personalized instruction, yet most adaptive systems focus on isolated skills or vocabulary. This study reports a 12‑week quasi‑experimental mixed‑methods evaluation of an adaptive platform that integrates the Vygotsky zone of proximal development (ZPD), Bayesian Knowledge Tracing (BKT), and Item Response Theory (IRT) to foster comprehensive proficiency across listening, speaking, reading, and writing. ANCOVA controlling for pre‑test scores indicated significant gains in all skills for the adaptive group (F = 19.30–44.65, p ≤ 1.82 × 10⁻⁵); effect sizes ranged from moderate for writing (d = 0.62) too large for listening, speaking, and reading (d = 0.84–0.93). Learners reported higher perceived ease of use, usefulness, and enjoyment, and these constructs correlated with proficiency gains. Qualitative data revealed enhanced motivation, autonomy, and reduced anxiety. Framing the design as a controlled short-term‑ intervention, the study highlights the platform’s pedagogical affordances and psychometric calibration while acknowledging the limitations of duration, sample homogeneity, and anxiety measurement. These findings support the feasibility of integrating cognitive scaffolding with latent mastery estimation to deliver balanced, adaptive English instruction.