基于行为序列分析对在线学习参与模式的探索

Exploring the Online Learning Participation Behavior Pattern Based on Behavioral Sequence Analysis

Based on the Moodle log data of 2131 students from The Open University， the research explored students’ participation pattern in online learning by correlation analysis， lag sequential analysis and cluster analysis. The research findings showed that 36 behavioral sequences out of 122 behavioral sequences formed by 15 behaviors recording in Moodle log were significantly correlated with the grade. Among them， 4 behavioral sequences were moderately correlated with the grade. Lag sequential analysis revealed that six behavioral sequences reached the level of significance. Based on the significant sequences， the behavioral transition diagrams were inferred. Through cluster analysis， five clusters of sample’s behavioral sequences were found. According to the analysis on the behavioral sequences and the behavioral transition diagram of each cluster， five online learning participation behavior pattern were identified， including low engagement pattern， surface level engagement pattern， performance engagement pattern， step-by-step participation pattern and random participation pattern. In the last part of the paper， the value of the behavioral sequence analysis， the effective behavior sequences and pattern， reflections on the online learning behavior pattern found in the current study are discussed. It is expected that the research would provide beneficial research perspective and support for online learning analysis， evaluation and improvement.