The Role of Generative AI and Peer Feedback in Cooperative Writing Within a Feedforward Pedagogy: A Study of Iranian IELTS Test Takers

Despite growing evidence for Artificial Intelligence (AI) in writing, little is known about combining Generative AI (GenAI) with peer feedback in L2 writing instruction. This mixed-methods study, grounded in AI-assisted language learning and collaborative learning theory, investigated the hybrid effect of GenAI and peer feedback through cooperative learning on the Writing Task 2 performance of 60 Iranian IELTS candidates. The experimental group received instruction integrating ChatGPT-5 and peer corrective feedback within cooperative activities, while the control group received parallel teacher-led instruction without GenAI feedback or structured peer collaboration. Pre- and posttests, assessed by certified examiners, showed the experimental group improved significantly in grammatical accuracy, lexical range, and discourse organization. Questionnaires and semi-structured interviews revealed that students perceived cooperative learning favorably and found GenAI feedback timely and useful for internalizing corrections, while peer collaboration reduced anxiety and enhanced idea development. Findings underscore the potential of large language models combined with peer feedforward to foster more effective and confident L2 writing, with implications for technology-enhanced instruction in high-stakes test preparation.