Developing fine-tuned prompt literacy in EAP: Graduate students’ experiences with customized GPT for academic inquiry

초록

Given the limited research on AI interaction competencies in advanced academic contexts, this qualitative case study examines how graduate students develop fine-tuned prompt literacy—the ability to strategically craft and refine prompts—while using customized GPT-based tools (MyGPT) for academic writing and research. Four graduate students in an English for Academic Purposes (EAP) context participated in iterative interactions with personalized MyGPT tools tailored to discipline-specific tasks. Data came from semi-structured interviews, open-ended survey responses, prompt-writing practices, and AI-generated outputs. The findings reveal four themes: (1) the development of prompt literacy through iterative experimentation and reflection, (2) the strategic customization of MyGPT to support research workflows and academic writing, (3) critical engagement with AI outputs, including the identification of limitations, inaccuracies, and ethical concerns, and (4) shifts in learner identity and cognitive-emotional responses, including increased confidence, agency, and ambivalence. The study positions fine-tuned prompt literacy as a sustained, discipline-sensitive, and evaluative form of academic AI literacy in contemporary L2 contexts. Pedagogically, it suggests that EAP instruction should scaffold sequenced prompting, prompt diagnosis, and verification routines to support reflective human-AI collaboration. Future research should examine broader samples and longer-term development. Ultimately, when guided by explicit scaffolding and ethical awareness, customized AI tools like MyGPT can support not only academic writing development but also the cultivation of critical digital literacies among graduate EAP learners.

키워드