Analysing interactions in online discussions through social network analysis

Document Type


Publication Date



Centers and Institutes


Background: Online discussion is one of the commonly used tools to enhance students' interactions and engagements in online courses, but it is not clear how social presence in online discussions impacts students' learning and what kinds of interactions we should encourage. Social network analysis provides a new methodology to investigate how social interactions in online discussions influence learning. Objectives: The study aims to identify students' online discussion interaction patterns and investigate how they are related to students' learning performance in an asynchronous course. Methods: Students who enrolled in an asynchronous undergraduate course were invited to participate in this study. Participants' online discussion participation data were collected from six online discussions from an asynchronous online course. Data were analysed using social network analysis methods as well as correlation and regression analysis. Results and Conclusions: We found that statistically significant positive correlations existed between learning performance and out-degree and closeness, respectively, which provided evidence to support that social interactions in online discussions have positive correlations with learning. We also found that out-closeness contributes to students' learning performance. The results of the study imply that knowledge construction occurs mainly in the centre of the outgoing network. By interpreting this study's results with the social construction of knowledge model, we conclude that knowledge construction mainly happens in reading and commenting on others' posts by internalizing and integrating external useful information from others. Implications: While numerous studies have emphasized the importance of interactions in online discussions or collaborative learning, there are limited studies on what kind of interactions we should encourage and what role students play in knowledge construction. The results of this study provide evidence that knowledge construction happens when students are reading and commenting on others' posts, which includes a hidden process of integrating external useful information into their own understanding. Practical strategies are provided to encourage students to be expansive and actively reach out to other students.

Publication Title

Journal of Computer Assisted Learning