Comparative Effectiveness of AI-Assisted Feedback vs. Traditional Teacher Feedback in Enhancing English Language Proficiency amongst Pakistani Undergraduate Students

Abstract

Using a mixed-methods approach, this study investigates the comparative effectiveness of AI-enabled feedback in comparison to traditional teacher-centered practices in English language global classrooms. The study draws upon data obtained from 120 undergraduate students at two higher educational institutions in Pakistan, which suggests that AI tools such as Grammarly can dramatically improve grammatical accuracy and vocabulary retention. On the other hand, whereas the teacher feedback mechanism is limited in immediateness, it fosters cultural relevance and learners’ comprehension assisted by Vygotsky’s socio-cultural theory. As corroborated by the provided qualitative findings, teachers offer necessary guidance that benefit from scaffolding and a more profound understanding of the students’ cultural backgrounds. Nonetheless, while students generally report satisfaction with AI-informed collaboration due to time efficiency, they indicate an aspect of disconnection attributed to emotional rapport with a teacher. In addition, this study reveals what infrastructure-related obstacles impede the efficient integration of AI and exacerbate disparities amongst the less-privileged, suggesting a blend of AI and traditional practices. Further recommendations include the need for establishing context and localizing AI tools using Urdu-English corpora to eliminate cultural bias, as well as making sophisticated AI products available to educators. Ultimately, the findings imply the need for hybrid AI-centered applied linguistics frameworks that balance precision and cultural sensitivity in marginalized settings, in alignment with comparative research evidence showing the effectiveness of a blended approach.