AI-enhanced peer collaboration in the digital writing classroom: A case from Indonesia

Abstract

Despite growing interest in peer collaboration and AI-assisted language learning, existing studies often examine these approaches in separate ways. This study responds to this gap by investigating the effectiveness of AI integration into peer collaborative practices in influencing students’ writing and their experiences in authentic digital writing classrooms. Students’ writing performance was examined after AI-enhanced peer collaboration, along with their experiences during the implementation of the digital writing classroom. Comparisons between utilisation in pair-structured writing collaboration and unstructured AI-enhanced peer collaboration practices were also measured. A mixed-methods study was employed, involving 30 undergraduate EFL students who participated in AI-structured peer collaboration throughout one semester, while another 30 students were assigned to unstructured strategy implementation. The quantitative data collected were analysed using the SPSS parametric test, and the qualitative data were analysed using thematic analysis. The results of the independent t-test and effect size tests revealed that students’ writing performance has notably improved (p<.005, d=2.27; large effect size). Students also expressed positive perspectives on AI-enhanced peer collaboration, indicating support from SEL, interactional process, and ecological perspectives. This study highlights the pedagogical value of integrating AI into EFL writing instruction to enhance students’ writing performance and collaborative skills.