【作 者】：
毛 刚， 周跃良
【关 键 词 】：
习惯性行为； 元认知； 学习投入； 学习分析； 教育大数据
【栏 目】：
网络教育
【中文摘要】：
以学习行为数据为基础的教育大数据分析是当前研究的热点，然而，当前研究更多地关注学习规律和特征发现、学习行为模式挖掘和学习风险预测，缺乏对学习者认知与元认知能力特征的深入分析和讨论。针对这一问题，研究聚焦于学习者习惯性学习行为，采用文献研究方法对学习分析相关实践和认知心理研究进行理论分析，阐明习惯性行为与元认知之间的关系。研究围绕元认知投入这一核心概念，从元认知投入的方式、力度和调控三个方面建立学习分析框架，并从元认知激活和保持两个维度提出量化分析方法。研究以“现代教育技术应用”课程为实践场景，分析了学习者习惯性行为数据，发现学习者元认知投入方式具有相对稳定性；教育资源发布策略会对学习者元认知投入的力度和调控产生影响；与男生相比，女生在元认知投入的各个方面表现更优。研究证实，该分析框架和方法能够有效地支持对学习者习惯性行为的深度分析，揭示行为背后隐含的元认知投入特征。
【英文摘要】：
Educational big data analysis based on learning behavior data is a hot topic in current research. However, current researches attach more importance on discovering learning rules and characteristics, mining learning behavior patterns and predicting learning risks, which lacks in-depth analysis and discussion on the characteristics of learners' cognitive and metacognitive abilities. In response to that problem, this study focuses on learners' habitual learning behaviors, and adopts literature research method to carry out a theoretical analysis on related practices of learning analytics and cognitive psychological research so as to clarify the relationship between habitual behaviors and metacognition. Based on the core concept of metacognitive engagement, this study establishes a learning analysis framework from three aspects of metacognitive engagement, namely, the manner, strength and regulation, and proposes a quantitative analysis method from metacognitive activation and maintenance. Taking the course of "Modern Educational Technology Application" as a practice scenario, this paper analyzes the data of learners' habitual behaviors and finds that the manner of learners' metacognitive engagement is relatively stable. The strategies of releasing educational resources will affect the strength and regulation of learners' metacognitive engagement. Compared with male students, female students perform better in all aspects of metacognitive engagement. This study proves that the analytical framework and method could effectively support the in-depth analysis of learners' habitual behaviors and reveal the hidden characteristics of metacognitive engagement behind those behaviors.