(The) effects of the structural positions in review networks on the course's popularity and satisfaction in MOOC online community

Abstract

In this paper, we apply network theory to understand students' behavior and relationships between courses in MOOCs. With prominent online review community data, we study the associations between courses based on student enrollments behavior. We find that all of the course networks have `small-world' property and that it increases over time, showing the distance between the courses becoming closer. Further, we study the impact of individual network properties on performance outcomes from the perspectives of satisfaction and popularity. We find that courses have gained popularity and satisfaction where they are positioned in community overlaps and when they bridge communities with weak ties. Moreover, strong connections between courses have a negative impact on their performance outcomes. We also find that in the ways of measuring brokerage, community overlaps and shortest paths based betweenness distinctively capture the different characteristics in these complex systems. From the results, we believe that our study can expand the research on performance analyses with network effects as well as the research on MOOCs from the perspective of the structural position of courses overall. Additionally, we expect that our study can be utilized by those who devise policy directions for future higher education systems and for the development of a course recommendation system.