Two Studies in Social Networks and Education: Empirical Investigations
Join Zoom Meeting
https://nus-sg.zoom.us/j/6153715329?pwd=alh6enAwSWdZbHhNZ3g1eFpKM0ZHQT09
Meeting ID: 615 371 5329
Password: 362666
Abstract:
Social networks play a critical role in determining educational outcomes. The rapid development of social learning technologies has fundamentally changed the way student learn, collaborate, and communicate. While social interactions through these technologies are prevalent in higher education, it is still unclear about the effects of online social interactions on academic performance and real-world communication. This dissertation seeks to contribute towards understanding (i) the impact of peer effects through online social interactions on academic performance, and (ii) the interplay between online and offline communication networks.
In my first study, I propose a three-stage framework to assess peer effects in a college setting by modeling the joint dynamic process of course selection, individual interactions, and academic performance. Specifically, I leverage a stochastic actor-oriented co-evolution model to measure the peer effects and homophily simultaneously and extend the model by introducing a prior selection stage. I apply the proposed model to a proprietary, de-identified student-level dataset spanning four years from a major Asian university. The empirical results provide support for the presence of significant and positive peer effects on academic performance over a four-year program. In addition, I uncover the heterogeneity in peer effects, in that the peer effects are more significantly positive among high-achieving students than mid-achieving and low-achieving ones. Findings of this study offer useful and prescriptive insights into the prevalence and importance of peer effects on learning outcomes in higher education and thus bear important implications for both social networks and education researchers.
While an increasing number of users adopt multiple communication media to meet their composite needs, little is known about how they create or maintain their interpersonal connections across these media in the extant research. In my second study, I adopt a multiplex network framework to understand the structure of and the interplay between online and offline communication networks. Using a longitudinal dataset collected through a major university in Asia, I construct an offline face-to-face communication network and two specific types of online communication networks, i.e., an online discussion forum and emails. In particular, I employ a stochastic actor-based co-evolution approach to model the joint dynamic process of online and offline communication networks at both individual and dyadic levels. I additionally examine the asymmetric effects between online and offline networks by distinguishing the directions of their mutual effects. Taken together, the empirical results in this study suggest that online ties complement offline connections through positive associations, whereas offline ties substitute online ties through negative associations. I also find that students are more likely to reciprocate within one network, but less likely to reciprocate across networks. Findings from this study hold critical managerial implications for information systems and social science professionals, especially those involved in the design of effective communication tools and interventions.
Through this dissertation, I attempt to not only answer important questions about the socio-economic value of student interactions on education outcomes and real-world behaviors, but also emphasize how large-scale digital traces and advanced computational methods can complement traditional social science approaches in modeling complex networked connections.