Design, Development, And Effect Analysis Of A Generative AI-Integrated Team-Based Writing Module : A Stepwise Approach Using Prompt Templates And Reflective Journals

This study aims to design and develop a stepwise writing module (Weeks 3-6; four weeks) that integrates generative AI into a team-based social-issue report assignment in the general education required course Writing and Expression, and to analyze the outcomes and remaining challenges of an instructional design that positions AI not as a substitute for drafting but as a collaborative tool for checking, revising, and refining texts. The participants were 69 students at C University in the Seoul metropolitan area (Section A: 35, Section B: 34), who produced an outline, introduction, body, and conclusion in stages as a team. At each stage, prompt templates were applied to review structure, audience orientation, evidence, and style; teams then engaged in discussion to implement the principle of “selective adoption.” Data included reflective journals and pre-/post-tests along with open-ended surveys, which were interpreted through descriptive statistics and descriptive content analysis. In the pre-survey, AI use experience was high (Section A: 81.3%, Section B: 74.3%), whereas awareness of ethics and source use was low (Section A: 40.6%, Section B: 54.3%). In the post-survey, perceived naturalness of AI use (A: 87.5%, B: 88.2%), learning efficiency (A: 90.6%, B: 88.2%), and class satisfaction and intention for future use (A: 96.9%, B: 94.1%) were reported as high, while ethics understanding and practice remained relatively low (A: 53.1%, B: 60.6%). Reflective-journal analyses showed that AI functions shifted from ideation and structuring in the outline stage to genre-requirement checks in the introduction, evidence and argumentation reinforcement in the body, and summarization and style editing in the conclusion; overall, partial adoption of AI suggestions predominated, but both high reliance and minimal use were observed in the final stage.