Development of an AI-Enhanced, Student-Centred English Language Teaching (ASCELT) Model for Improving Communication Competence among Higher Education Students

Author: Malaysia

Abstract

Higher education institutions increasingly recognise that English communication competence the capacity to use language accurately, appropriately, and strategically across authentic contexts is critical to graduate employability, yet conventional teacher-centred instruction often fails to provide the individualised, sustained practice that competence development requires. This study proposes and validates an AI-Enhanced, Student-Centred English Language Teaching model (hereafter the ASCELT model) intended to strengthen university students' communication competence by integrating artificial intelligence (AI) tools within a student-centred pedagogical framework. Using a two-phase Design and Development Research (DDR) methodology, the study first employed the Nominal Group Technique (NGT) with nine language-education experts to brainstorm and prioritise candidate constructs for the model, followed by the Fuzzy Delphi Method (FDM) with seven experts to statistically validate the constructs using triangular fuzzy numbers, threshold (d) values, percentage of expert consensus, and defuzzification (alpha-cut) scores. Seven constructs were generated in the NGT phase; all seven achieved threshold values below 0.2, consensus levels above 75%, and alpha-cut values above 0.5 in the FDM phase, indicating strong expert agreement. The validated constructs were subsequently synthesised into the ASCELT model, comprising AI conversational chatbots for speaking practice, AI-personalised adaptive feedback, AI-assisted writing correction, student-centred collaborative tasks, blended/flipped classroom integration, redefinition of the teacher's facilitative role, and AI-driven formative assessment, feeding into a core integration layer that drives improved communication competence across linguistic, sociolinguistic, discourse, and strategic dimensions. The findings offer higher education stakeholders an empirically validated, expert-endorsed framework for embedding AI responsibly and pedagogically within student-centred English language teaching.