Integrating Artificial Intelligence into Task-Based Teaching: An Action Research on Improving College English Speaking Skills

Author : Huang, Liangyan

Abstract : This research employs a two-cycle action research methodology to examine the efficacy of AI-assisted task-based language teaching (TBLT) in improving tertiary students’ oral English speaking proficiency. The participants were 12 second-year dance majors from a university in China. The research process was conducted over two four-week cycles. he research started with semi-structured interviews to identify their English backgrounds and needs. Data were triangulated through pre-research semi-structured interviews, classroom observations, teacher and student reflections, and student assignments. The results demonstrate that the AI-assisted TBLT model significantly enhanced learners’ oral fluency, lexical precision, and class engagement, attributable to the positive role of AI in task design, instant feedback, and fostering learning autonomy. Furthermore, pedagogical reflections underscored that the activity authenticity and interactivity of the AI-assisted tasks were pivotal to student success. Consequently, the research furnishes empirical insights for integrating AI into oral English pedagogy and suggests refinements for task-based instruction.

Keywords : AI-assisted TBLT, oral English proficiency, action research, tertiary education, learner autonomy

Conference Name : International Conference on Higher Education and Future Skills Development (ICHEFSD - 26)

Conference Place : Kuala Lumpur, Malaysia

Conference Date : 7th Feb 2026