IMPACT OF AI-INTEGRATED PEDAGOGY ON ORAL AND WRITTEN PROFICIENCY OF ESL LEARNERS

Abstract

Learning English as a foreign language in Pakistani public schools has never been straightforward. Overcrowded classrooms, limited one-on-one contact time, and little opportunity for spoken practice have combined to leave many Grades 8 students struggling with the language skills particularly speaking fluency and written accuracy, which are formally assessed at the end of the school cycle. This study asked whether three AI-based language tools (ELSA Speak, Duolingo, and Grammarly), integrated into ten weeks of regular English lessons, could make a genuine, measurable difference to those outcomes. A quasi-experimental pretest-posttest design was used, with thirty students receiving AI-supported instruction and thirty taught through conventional methods. Thirty experimental group students also completed a 20-item questionnaire on their awareness, usage, and perceptions of the tools, and forty practicing English teachers shared their views on AI integration through a mixed-methods questionnaire. The quantitative data were analyzed through SPSS 26 and the teacher qualitative responses through thematic analysis. Results were clear: students who worked with the AI tools made substantially larger gains on every measured sub-skill than their peers in the control group, they reported strong engagement with the tools, and their motivation climbed across the intervention period. Teachers were broadly supportive of the direction, even as they pointed honestly to the infrastructure gaps and professional development needs that currently stand in the way of full implementation.