Instructors: Kaitlin Torphy & Jiliang Tang
Since the last presidential election, an emerging awareness of social influence within social media, misinformation, and the relative power of big data have played out in public space a la the Mueller Report, televised special counsel investigations, and 24-hour news coverage. Regardless of partisan affiliation, gone is the notion that information diffused within virtual space, and individuals’ engagement therein is irrelevant. Within the 21st century, how individuals relate within virtual space is a window into their thinking and often behavior on and offline. Part 1 of this course considers how the emergent field of computational social science may extend traditional approaches to causal estimation to incorporate computer science applications with big data generated continuously within virtual space. Pairing data generated on and offline, researchers may dive down into individuals’ behavior identifying patterns relative to policy reform or contextual variation. To frame the potential to extend social science research, we will present techniques from the Teachers in Social Media Project. The Teachers in Social Media Project uses interdisciplinary approaches to study teacher engagement online and its impact on classroom practices and policy reform. Today, a majority of teachers engage with social media for professional purposes. Pinterest, a prominent social media platform, enables teachers to curate instructional resources and share them with an embedded network of colleagues. Building off previous NSF support (NSF REAL–1420532), we find 90% of 310 sampled teachers, across career stage, use Pinterest. At the end of Part 1 in this course, Fellows will be able to describe computational social science research and explain the affordances and cautions when conducting this type of research. In Part 2 of this course, we will present a variety of research papers employing traditional social science and computational approaches to examine questions of inquiry. Fellows will be asked to examine research and identify where computational social science was employed, why, and how it supported causal estimation. They will also learn how to navigate different languages and cultures between computation and social science. Fellows will be given the opportunity to generate their own research proposals, working individually, or in teams. Proposals should include the research question of interest, data to be collected or used, and proposed analysis. The course will end with an opportunity to present ideas and receive feedback from the instructors and peers.