Exploring the Impact on AI writing tools on student revision practices in secondary ELA classrooms

Abstract

The growing availability of AI writing tools presents both opportunities and challenges for how secondary students revise their work. While tools such as Grammarly and ChatGPT offer immediate feedback that can scaffold surface-level accuracy, concerns remain about their potential to discourage deeper, conceptual revision. Guided by socio-cognitive writing theory, this mixed-methods empirical study examines the impact of AI writing tools on revision practices in secondary English Language Arts (ELA) classrooms. Participants revised argumentative essays with or without access to AI tools, using data from writing samples, surveys, and think-aloud protocols. Quantitative analyses indicate that AI-assisted students produced significantly more revisions overall, particularly at the surface level, while qualitative findings reveal that engagement with global revision was strongly mediated by teacher guidance and students’ digital literacy. The results suggest that AI writing tools can function as effective cognitive scaffolds when embedded within intentional pedagogical frameworks. Implications are discussed for classroom instruction, teacher professional development, and the responsible integration of AI in secondary education.