Economic Analysis of Social Networks

Course Description

This is a Ph.D.-level field course taught by Prof. Chih-Sheng Hsieh. The course starts from the characterization of networks, then enters into diverse topics for networks: network interactions, static and dynamic network formation, network sampling, and community detection. We mainly discuss tens of empirical research to introduce the data type and famous datasets for network analysis and find significant evidence about how peer effects affect us. In class, we use software to arrange data and perform econometric regressions on network data and provide economic interpretations

Moreover, we also focus on the microfondation of the network by discussing pieces of classical literature such as Calvó-Armengoi’s works. My literature presentation introduces Calvó-Armengoi’s pioneer contribution to targeting players in networks. Follow-up extension includes Golub, Demange, and Borgatti.

Term Project

A rapidly growing number of network studies appear in economic and computer science fields, such as labor, industrial organization, and financial economics. In addition, with the widespread adoption of online learning platforms accelerating by COVID-19, people have experienced an online lifestyle in these three years, enabling platforms to collect more online behavior data. In the university, thousands of professors adopt online rather than in-class teaching. Students watch course videos, interact with the instructor, submit assignments, and take exams, all on the online teaching platforms.

Online platform services are embedded in daily life; however, little attentions are received to the relationship between learning behavior and learning outcomes. Do students who tend to watch course videos late at night achieve a higher grade? What is the impact of different online time distribution and different learning patterns? Does the friend share a similar online learning behavior, or are students with similar online learning behavior easier to get familiar with? Will students who procrastinate in working and submitting the assignments significantly lower grade? Do students tending to speed up the course videos have a higher grade? That’s what I want to scrutinize.

My term project aims to provide critical empirical analysis of online learning behavior. The objectives of this project contains

1. Propose a theoretical model that captures the similarity of behaviors for the online education platforms as a metric to cluster students.
2. Explore the relationship and effects between the characteristics and the behaviors and understand whether students with similar characteristics result in similar behaviors.
3. Explore whether the friendship network results in similar behavior and vice versa. For example, do the night type students tend to have a friend with the same behavior or do late-submission type students affect the early-submission type?
4. Construct an empirical approach based on the behavior data and explore the relationship between behavior and academic outcomes.
5. Provide a strategic guide based on the behavior data for learning behavior and a study plan for students and instructors. For example, teammate searching and team allocation.