Comparing individual vs. collaborative processing of ChatGPT-generated feedback: Effects on L2 writing task improvement and learning
Da Yan
Abstract
This research examined the effects of collaborative processing of ChatGPT-generated feedback on second language (L2) writing development. The sample for the seven-week experiment consisted of 117 sophomore EFL learners and six teachers from a Chinese university. The students were divided into four groups: the control group processed ChatGPT feedback individually, while the experimental groups processed it with teacher, peer, or combined teacher-and-peer collaboration. Employing a mixed design, L2 writing task improvement (measured as the gain scores from draft writing to final products for the three during-intervention writing tasks) and learning (measured as performance for a post-intervention, and similar new writing task) of the four groups were assessed and analyzed. The findings revealed that the learners who individually processed feedback registered the most significant task improvements, whereas learners processing feedback with teacher collaboration progressed the most for subsequent learning processes. The study has pedagogical implications for the construction of a more inclusive ecology to support effective uses of latest technology in L2 classrooms.