High School Students’ Perceptions of Using ChatGPT with the RACE Framework to Support Autonomous Learning in Writing English

Abstract

This study explores high school students’ perceptions of using ChatGPT with the RACE Framework (Role, Action, Context, Execute) to support autonomous learning in English writing. Employing a qualitative case study design, data were collected from class XI-7 students at SMAN 1 Mojosari through classroom observations and semi-structured interviews. Analysis followed Miles, Huberman, and Saldaña’s thematic approach, focusing on cognitive, affective, and conative domains within Bloom’s Taxonomy. Findings reveal that ChatGPT, when integrated with the RACE Framework, facilitates progressive development in all three domains. Cognitively, students advanced from remembering vocabulary to creating original works that synthesize AI-generated content with personal ideas. Affectively, they moved from curiosity toward AI-assisted learning to internalizing the value of independent learning, demonstrating increased motivation and self-confidence. Conatively, students progressed from imitating teacher demonstrations to achieving naturalization, using AI tools automatically, strategically, and reflectively. The results highlight the RACE Framework’s role in guiding purposeful AI use, promoting critical thinking, creativity, and independence rather than over-reliance on technology. Theoretically, the study enriches AI-assisted language learning literature by providing domain-specific insights at the secondary school level. Practically, it offers educators a structured model for integrating AI into writing instruction to strengthen learner autonomy.