Artificial Intelligence and Sustainable TVET Education: Enhancing English Language Proficiency Through Prepositional Error Correction

Abstract

The rapid advancement of Artificial Intelligence (AI) is reshaping educational practices by providing innovative approaches to teaching, learning, and assessment. Within Technical and Vocational Education and Training (TVET), AI-driven technologies have gained increasing attention due to their potential to support sustainable learning and enhance students’ academic and professional competencies. Among these competencies, English language proficiency remains essential for TVET graduates as industries become more globalized, digitalized, and communication-oriented. Despite its importance, many students continue to struggle with grammatical accuracy, particularly with prepositions, which affects the clarity and effectiveness of written and spoken communication. Recent developments in AI-powered language learning applications, including intelligent tutoring systems, natural language processing (NLP), and automated grammar correction tools, offer opportunities to address these challenges. By providing immediate feedback, personalized guidance, and adaptive learning experiences, these technologies can help students identify and correct language errors more effectively than conventional learning approaches.This study examines the factors influencing the effectiveness of AI-assisted prepositional error correction in improving English language proficiency among TVET students. Specifically, the study investigates the roles of personalized feedback, real-time error detection, adaptive learning capability, learning engagement, and perceived usefulness in shaping language learning outcomes. A quantitative research design was employed, and data were collected through an online questionnaire distributed to TVET students enrolled in English-related courses at selected institutions in Malaysia. A total of 350 valid responses were analyzed using Partial Least Squares Structural Equation Modelling (PLS-SEM) through SmartPLS 4.0. The analysis included measurement model assessment and structural model evaluation, covering reliability testing, convergent validity, discriminant validity, path coefficient analysis, bootstrapping procedures, and Importance-Performance Map Analysis (IPMA). The findings reveal that personalized feedback is the strongest predictor of English language proficiency enhancement, followed by real-time error detection. Adaptive learning capability, learning engagement, and perceived usefulness also demonstrate significant positive relationships with students’ language development. The results suggest that AI-assisted learning environments can improve grammatical accuracy through timely corrective feedback, individualized learning support, and flexible learning pathways. The IPMA findings further indicate that personalized feedback should remain a priority for educators and educational technology developers seeking to maximize the effectiveness of AI-supported language learning systems.In conclusion, AI-assisted prepositional error correction represents a valuable educational innovation that can strengthen English language proficiency within sustainable TVET education while supporting lifelong learning and workforce readiness among future TVET graduates.