Does Emotional Interaction Affect the Effectiveness of Online Collaborative Conversations?

Collaborative learning has become the predominant method of learning today. This study explores the impact of emotional interaction on the effectiveness of online collaborative conversations, using freshmen at a university as subjects. Participants were randomly divided into three groups—positive, neutral, and negative — using quasi-experimental methods to support their completion of collaborative tasks with emotional interaction tools. Texts were analyzed based on social networks and cognitive networks. The results show that the network density and out-degree centrality in the positive emotion group were significantly higher than those in other groups, indicating that members formed tight collaborative networks through reciprocal interactions, with balanced distribution during the cognitive phase. In contrast, the neutral emotion group had a loose network with the lowest out-degree and in-degree centrality, where members invested more during the information accumulation and verification phases but had limited overall collaboration efficiency. The negative emotion group had the lowest network density and a significantly lower proportion in the integration phase compared to other phases. The findings provide valuable insights for designing online collaborative activities.