跨模态、自适应与人机协同：新托福写作教学的创新实践与路径探索——基于双重刺激理论的托福邮件写作题型动态脚手架设计

Multimodality, Adaptivity, and Human-AI Collaboration: Innovative Practices and Pathway Exploration in New TOEFL Writing Instruction—Dynamic Scaffolding Design for TOEFL Email Writing Tasks Based on Double Stimulation Theory

摘要: 新托福考试的写作板块对考生的时间管理能力、多模态信息处理能力及学术和职业场景适应性提出了更高要求。而传统单一文本输入与滞后反馈的教学模式难以应对这一挑战。研究以双重刺激为机制，基于真实的师生线下一对一教学实录，结合转化能动性理论，深入剖析了教师在面对学生动机不足与准备度缺失的矛盾时，如何通过跨模态输入脚手架、自适应学习路径决策与人机协同反馈闭环重构课堂生态。结果显示，这套做法能减轻学生的认知负担，增强学习信心和动力，教学重点从改正语法错误转向训练逻辑论证。文章说明了教学设计过程、实际效果、应用价值和推广路径，同时指出了规模化应用的难点。

Abstract: The new TOEFL iBT (2026) writing section places higher demands on students’ time management skills, multimodal information processing abilities, and adaptability to academic and professional tasks. Traditional text-based instruction with delayed feedback proves inadequate for these requirements. Grounding its mechanism in dual stimulation theory and drawing upon authentic records of one-on-one, face-to-face teacher-student instruction, this study integrates transformative agency theory to conduct an in-depth analysis of how teachers reconstruct the classroom ecology when confronting the contradiction between students’ insufficient motivation and lack of readiness. Specifically, it examines the deployment of cross-modal input scaffolding, adaptive learning pathway decision-making, and a human-machine collaborative feedback loop. Findings show this combination lowers students’ cognitive load, builds their confidence, and moves teachers’ focus from fixing grammar errors to building logical arguments. This paper details the instructional design, empirical outcomes, core values, and replication pathways, while addressing limitations for large-scale implementation.