The reading-to-write approach is regarded as one of the most widely used approaches for integrating reading and writing in ESL/EFL contexts. However, how both genre and technology can be effectively integrated remains under investigation. The current study explores how ChatGPT-integrated argumentative reading-to-write instruction influenced students’ writing quality and the types of prompts they used across writing stages. Employing a within-subjects design, the study was conducted with 31 students over 6 weeks of instruction at a Korean university. Pre- and post-writing samples were collected to compare writing quality in terms of language use and argument structure, and ChatGPT logs to identify prompt features. The findings revealed that text length and grammatical accuracy improved significantly, whereas lexical diversity and syntactic complexity did not. Notably, there was significant improvement in all argument elements, particularly propositions, warrants, oppositions, and rebuttals. The analyses of ChatGPT prompts revealed that they were predominantly dependent, general, and information-seeking. Pedagogical implications and limitations of the study are discussed.

Exploring ChatGPT’s Potential in ArgumentativeGenre-Based Reading-to-Write Instruction
I. Introduction
II. Theoretical Background
    1. Genre-based Argumentative Writing and Reading
    2. ChatGPT Prompts in Writing Research
III. Research Method
    1. Participants
    2. Research Design
    3. Data Collection Procedure
    4. Data Analysis
IV. Results and Discussion
    1. Change in Writing Qualities
    2. Question Types in ChatGPT Prompts
V. Conclusion
References

• Hyojung Lee(Dept. of English Linguistics and Literature, Lecturer, Gyeongsang National University)