Research on the mechanism of knowledge diffusion in the MOOC learning forum using ERGMs
Bing Wu,Cancan Wu
DOI: https://doi.org/10.1016/j.compedu.2021.104295
2021-11-01
Abstract:PurposeThe popularity of MOOCs (massive open online courses) has been increasing rapidly around the world, but there is a lack of research on knowledge diffusion in the MOOC learning forum. This article uses ERGMs (exponential random graph models) to synthesize user attributes from the perspective of social networks, social support and social capital to systematically explore the mechanism of knowledge diffusion in the MOOC learning forum.Design/methodology/approachFirst, the literature related to knowledge diffusion is reviewed to define two forms of knowledge diffusion in the MOOC learning forum, i.e., knowledge transferring and knowledge sharing. Second, based on social network theory, research hypotheses related to centrality, reciprocity, and transitivity are proposed; based on social support theory, research hypotheses related to emotional support, information support and sense of belonging support are proposed; and based on social capital theory, research hypotheses related to the number of posts and number of being liked are proposed. Third, "machine learning", the most popular course on the Coursera platform, is selected as the research object. The data from March 2016 to December 2019 is obtained, and then, ERGMs are used to simulate the network and test the research hypotheses. Finally, based on the conclusions of this study, corresponding suggestions are proposed from the perspectives of MOOCs platform providers, MOOCs providers and MOOCs learners, to promote the development of MOOCs.Findings– (1) In the MOOC learning forum, users with high degree centrality contribute more, actively answering questions from other users, and transferring knowledge to others, but knowledge sharing between two users with high degree centrality is unlikely to occur. (2) Significant reciprocity features exist in knowledge transfer in the MOOC learning forum. (3) There is no significant triangular relationship structure in the knowledge sharing network of the MOOC learning forum, which means that multiple dyad sharing partners are more likely to be formed in the MOOC learning forum. (4) Users with emotion tendency tend to transfer knowledge to other users, and knowledge sharing is likely to occur between individuals with emotion tendency. (5) There is no tendency that users with low learning progress receive knowledge transferred by other users. (6) Teaching assistants, as users with higher knowledge potential in the MOOC learning forum, are prone to transfer knowledge to other users. (7) The trend of knowledge sharing between users in the same region is not obvious. (8) Users with more posts and users with more being liked are more likely to receive knowledge transfer by the other users, and thus, they are easier to obtain answers from. A possible explanation for this result could be that users with more posts and users with more being liked have a higher contribution to the MOOC learning forum, and therefore, questions they raised are more likely to attract attention.Theoretical implicationsFirst, current research on the MOOC learning forum focuses on the discussion of user behavior, but there is a lack of research on knowledge diffusion. Therefore, in terms of the research perspective, from the aspect of knowledge diffusion, this article defines two forms of knowledge diffusion: knowledge transfer and knowledge sharing. Then, knowledge transfer network and knowledge sharing network are constructed to explore user posting and replying behaviors, to expand the user behavior research in the MOOC learning forum. Second, current research on the user interaction network in the MOOC learning forum lacks comprehensive research on the network structure and user attributes. The micro logic embedded in the network structure must be systematically analyzed. Therefore, the network structure and user attributes are synthesized to explore the influencing factor of knowledge diffusion, which enriches the research content of the MOOC learning forum. Third, ERGMs are suitable for exploring potential factors of network formation, as well as the impact of multiple features simultaneously on network formation. Currently, ERGMs are widely used to research citation network formation mechanisms and social media network formation mechanisms. However, there is still a lack of research on user interaction networks in the MOOC learning forum while applying ERGMs. Therefore, this article introduces ERGMs into the knowledge diffusion research in the MOOC learning forum to provide effective research methods for modeling and simulating the knowledge diffusion networks.Practical implications( -Abstract Truncated-
education & educational research,computer science, interdisciplinary applications