Flipped AI-Supported Prompt-Based Scaffolding: An Innovative Approach to Enhancing EFL Descriptive Writing

Aim/Purpose: This study investigates the effectiveness of a flipped AI-supported writing model that integrates structured prompt-based scaffolding to enhance EFL learners’ ability to produce descriptive statistical writing. Background: While artificial intelligence is increasingly used in writing instruction, prior research has largely focused on text generation and automated feedback. Limited attention has been given to AI as a guided learning partner in flipped instructional designs, particularly for genre-specific writing tasks such as static bar chart descriptions. Methodology: A quasi-experimental one-group pre-test/post-test design was employed with 25 B1–B2 EFL learners at a public university in Oman. The four-week intervention combined AI-assisted pre-class preparation, a teacher-guided “write-with” in-class phase, and AI-supported revision. Writing performance was assessed using parallel tasks evaluated with an IELTS-based rubric, and learner perceptions were collected through a post-intervention questionnaire. Contribution: This study contributes to AI-mediated writing pedagogy by proposing and empirically testing a flipped, prompt-based scaffolding model for EFL descriptive writing. It contributes to emerging research on prompt-based scaffolding, AI literacy, and technology-integrated writing pedagogy in EFL contexts. Findings: Results revealed a statistically significant improvement in writing performance following the intervention, with large effect-size gains across all four assessed writing sub-skills (task achievement, coherence and cohesion, lexical resource, and grammatical range and accuracy). Questionnaire findings indicated positive learner perceptions, highlighting increased confidence and improved understanding of responsible AI use. Recommendations for Practitioners: Educators should integrate structured AI prompts within flipped learning designs to support pre-class preparation and scaffold in-class writing. Emphasis should be placed on training learners to use AI for guided learning rather than full-text generation. Recommendation for Researchers: Researchers should consider incorporating structured prompt-based scaffolding when examining AI-supported writing instruction. Future studies should adopt more robust experimental designs, include diverse learner populations, and examine how AI can be integrated as pedagogical support rather than as a text-generation tool. Impact on Society: The model offers a practical and scalable approach to integrating AI into language classrooms, supporting ethical AI use while enhancing learners’ academic writing skills. Future Research: Future research should examine the effectiveness of the Flipped AI Writing Model using control groups, longer intervention periods, and across different writing genres. Longitudinal studies are also needed to explore the development of AI literacy and self-regulated learning over time.