基于大数据学习分析的在线学习绩效预警因素及干预对策的实证研究

Empirical Research of Predictive Factors and Intervention Countermeasures of Online Learning Performance on Big Data-based Learning Analytics

利用学习分析技术挖掘教育大数据,能够发现潜在价值,并使其转换成有意义的教学信息,进而优化学习过程、提高教学效果,成为教师、学生及教育研究者的共同诉求。从学习分析技术的教育应用视角,追踪、积累并筛选在线学习行为数据,利用多元回归分析法判定影响学生学习绩效的预警因素,在此基础上构建了干预模型,将其应用于教学实践中,对产生的学习行为数据进行二元Logistic回归分析,并结合问卷调查和访谈法对该模型在学习活动、知识习得等方面的有效性进行验证。结果表明,通过学习过程中实施的干预模型识别出存在学习危机的学生,及时向其发出预警信号并提供个性化干预对策,有利于增强学习动机,培养学习毅力,提高学习质量。

To Mine educational data by learning analytics will be helpful to discover the potential value of the data and then transfer them into meaningful information, optimize learning process and improve teaching effects, which becomes the common demands of teachers, students and educational researchers. From the perspective of educational application of learning analyties, an intervention model has been constructed based on tracing, accumulating and selecting the online learning behavior data in Moodle, and using multiple regression to determine predictive factors affecting students＂ learning performance. After that, this model is applied to teaching practice, and the learning behavior data produced is analyzed by binary logistic regression. The validity of the model in learning activities, knowledge acquisition etc. has been verified through questionnaire and interview. The results indicate that through the intervention model, the students with learning crises can be identified. A timely warning sign and personalized intervention can be helpful to enhance students＂ learning motivations, to cultivate their learning perseverance and improve their quality of learning as well.