Collaborative Human-AI Revision Design embedding prompt engineering across drafting, feedback, and rewriting phases¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The article proposes a "Collaborative Human-AI Revision Design" that systematically embeds prompt engineering into three writing phases. In drafting, teachers guide learners to construct "constraint-based prompts" (e.g., outlines, vocabulary matrices, counter-arguments) while prohibiting continuous prose generation. In feedback, students use "criteria-specific diagnostic prompts" treating the AI as a formative peer reviewer. In rewriting, assessment becomes process-oriented, such as "Prompt Reflection Portfolios" rewarding metacognitive quality of interaction and productive failure.

Design Implications¶

Context¶

Requirements¶

• Prompting is taught explicitly as a cognitive support strategy, not a shortcut for text generation.

Constraints¶

• The strategy is a proposal aligned with the article's correlational empirical model, not an intervention the article tested.

Target Learners¶

• University EFL learners writing with generative AI

Target Learning Goals¶

• Deep revision, critical synthesis of AI feedback, and human-AI collaboration as academic competency

Related Strategies¶

• Implement AI-peer integrated feedback as a standardized two-phase revision cycle with rubric-aligned prompting and instructor monitoring
• Integrate AI feedback as a supportive resource within process-oriented drafting cycles, with explicit guidance on critical evaluation
• Use GenAI across the writing process stages of pre-writing, drafting language support, and revision feedback, and for discussion questions, differentiated explanations, and scaffolded practice
• Train teachers in prompt engineering, evaluating AI-generated responses, ethical and privacy considerations, and integrating AI outputs into pedagogical decision-making
• Process Writing Workshop
• Peer Revision
• Calibrate GenAI feedback through calibrated prompts and pedagogical (re-)framing
• Implement an AI-Peer-Teacher hybrid integrated feedback model staged across drafting, revision, and polishing phases
• Provide explicit instructional boundary cues, process-oriented assessment, and peer modeling of verification to frame responsible GenAI use
• Design AI writing support to reduce ownership loss: fade AI interactions over time, generate incomplete text requiring rewriting, and make the chatbox editable
• Provide timely automated feedback through an AI chatbot during EFL writing practice

Examples¶

-

Key Sources¶

• Li, H. and Zhang, W. (2026). Empowerment over enforcement: unpacking the psychological drivers of AI-assisted deep revision in EFL writing. Front. Psychol. 17:1871022. https://doi.org/10.3389/fpsyg.2026.1871022