基于元认知策略的在线学习支架设计

Designing Online Learning Scaffolds Based on Metacognitive Strategies

摘要: 生成式人工智能(GAI)深度融入在线教育，在提升学习效率的同时，也带来认知外包、学习主体性消解与深度思维弱化等风险，构建以学习者为中心的认知防护机制已成为当务之急。本研究通过系统性文献综述，整合2020~2025年间52篇中外文献，围绕元认知策略如何抵御GAI认知异化这一问题，构建了涵盖目标规划、过程监控、策略调节与反思评估的在线学习支架。该支架以高等教育中的批判性思维训练为应用场景，设计遵循学习主体性优先、过程导向、动态适应与透明性四项原则，并通过课程嵌入、教学活动设计与过程性评价实现落地。以元认知为核心，该支架能够引导学习者在与GAI互动中保持认知自主性，为智能时代在线教育兼顾技术赋能与育人本质提供可操作的理论路径与实践参照。

Abstract: The deep integration of generative artificial intelligence (GAI) into online education improves learning efficiency while also creating risks such as cognitive offloading, the erosion of learner agency, and the weakening of deep thinking. Building a learner-centered cognitive protection mechanism has therefore become an urgent task. Through a systematic literature review of 52 Chinese and international studies published between 2020 and 2025, this study addresses how metacognitive strategies can counter GAI-induced cognitive alienation and constructs online learning scaffolds that encompass goal planning, process monitoring, strategy regulation, and reflective evaluation. This framework is applied in the context of critical thinking training within higher education. The scaffold design follows four principles: prioritizing learner agency, process orientation, dynamic adaptation, and transparency. It is implemented through curriculum embedding, instructional activity design, and process-based assessment. With metacognition at its core, the scaffolds can guide learners to maintain their cognitive autonomy when interacting with GAI, thereby providing an actionable theoretical pathway and practical reference for online education in the intelligent era to balance technological empowerment with the essence of education.