基于多种统计方法的慕课学习行为分析——以云南大学《雷达气象学》线上课程为例

Analyzing MOOC Learning Behaviors Using Multiple Statistical Methods—A Case Study of the Online Course “Radar Meteorology” at Yunnan University

摘要: 随着教育信息化的快速发展，线上教育已成为高等教育中不可或缺的组成部分，慕课(MOOC)累积了海量学生学习数据，涵盖了线上学习过程的不同学习模块和各个环节。本文通过分析校内慕课平台学生的学习数据，利用相关分析挖掘了学生出勤率、成绩与线上学习行为之间的关系，通过K-means聚类方法和随机森林模型对学生不同线上学习模块的学习拖延行为进行了聚类分析，分析了学生对不同线上学习模块的偏好和学习习惯，对不同类型的学习者提出了有针对性的教学建议，同时可根据学生的学习行为偏好和倾向优化线上学习模块的教学设计，从而提升学生的学习效果和成绩。

Abstract: With the rapid development of educational informatization, online education has become an indispensable component of higher education. Massive Open Online Courses (MOOCs) have accumulated vast amounts of student learning data, encompassing various learning modules and stages within the online learning process. By analyzing student learning data from an institutional MOOC platform, this study employed correlation analysis to uncover relationships between student attendance rates, academic performance, and online learning behaviors. Utilizing K-means clustering and a random forest model, we conducted a cluster analysis of learning procrastination behaviors across different online learning modules. The analysis revealed student preferences and learning habits pertaining to distinct online modules. Based on these findings, targeted teaching strategies and recommendations are proposed for different types of learners. Furthermore, the instructional design of online learning modules can be optimized according to students’ learning behavior preferences and inclinations, thereby enhancing learning outcomes and academic performance.