Dynamics of Emotions and Network Structures in a Course forum: An Empirical Investigation in the Last four Weeks Before the Exam

Abstract

In recent years, a growing number of educational researchers are keen to utilize social network analysis (SNA) and emotion detection for exploring collective learning states. Students' emotions and interaction characteristics before the exam typically suggest some significant traces of learning states. In this study, data from the discussion forum of Chinese legal history course in a university learning platform was used to investigate evolutionary trends of students' network characteristics and emotion densities (EDs) in the last four weeks before the final exam, as well as visualized the distribution of the high-EDs (including positivity, negativity and confusion) students in the weekly network. Empirical analyses suggested that, as the exam approaches, learners' network structure and emotional densities are constantly changing. After experiencing a smooth change in the first two weeks, the average degree centrality reached a peak in the third week, and confusion emotional density far exceeded positive and negative emotional density in the last week, which may help in identifying the potential academic losers and providing timely interventions.