SYNERGIZING PROBLEM-BASED LEARNING AND AI TOOLS: A LONGITUDINAL STUDY ON SPEAKING PROFICIENCY AT THE B2 LEVEL

Abstract

This research examines the long-term effects of combining Problem-Based Learning (PBL) with Artificial Intelligence (AI) tools on enhancing speaking skills at the B2 level in English as a Foreign Language (EFL) settings. Carried out over nine months, the study included 92 undergraduate students separated into experimental and control groups. The experimental group participated in PBL activities enhanced by AI-powered applications, incorporating speech recognition technology, chatbots, and automated feedback mechanisms, whereas the control group underwent conventional communicative teaching. A mixed-methods strategy was utilized, combining CEFR-aligned speaking evaluations, discourse analysis, learner surveys, and interviews. The results show that the experimental group achieved notably greater improvements in fluency, interactional competence, lexical variety, and self-regulation than the control group. The findings also indicate that AI tools improved learner independence and diminished speaking anxiety. The research adds to the developing area of AI-supported language education and provides teaching perspectives on combining technology with learner-focused approaches to promote improved speaking skills.