The NCASD Lab was well-represented at this year’s American Sociological Association meeting. Current and former center members presented seven talks at the Joint Japan/North America Mathematical Sociology Meeting held prior to the ASA. Lb members also presented papers at the ASA conference itself. In addition to presenting the group’s own work, Lab PI Butts served as organizer for the three mathematical sociology sessions at the ASA meeting. The members of the NCASD Lab are pleased to be able to contribute to the ongoing vitality of mathematical sociology as a field, both in the US and worldwide.
Lab PI Butts presented new work on the application of spectral theory to feedback centrality scores. Feedback centrality scores are used to measure power, status, and influence in many kinds of networks, and are the basis for applications such as Google’s ranking of web sites. Butts’s work shows how spectral theory can be used to provide a unified understanding of how these measures behave, and to explain when and why very different types of power-related processes sometimes lead to very similar outcomes.
Lab member Emma Spiro presented work on self-disclosure norms on Facebook. A key component of self-presentation is the decision to reveal or conceal information about one’s person or circumstances. The online environment provides a unique opportunity for comparative study of such privacy-related behavior, both because of its relative newness and because of the opportunities it affords for systematic behavioral measurement on a global scale. The project users data from a probability sample of over 1 million Facebook users, in order to relate online privacy awareness behaviors in 217 countries to the social, political, and economic factors that shape the context of social interaction.
Lab member Sean Fitzhugh presented work on developing more efficient link-trace methods for locating subpopulations within a network. Using a sample of nearly 130,000 individuals from data scraped from MeetUp.com, Fitzhugh and coauthors demonstrate the utility of a modified link-trace search they call a preferential link-trace. Not only does it afford substantial efficiency gains over a tradition link-trace search, but it allows for estimation of the total population size and size of the subpopulation of interest. Using the preferential link-trace, they found members associated with the Occupy Wall Street social movement much faster than with conventional link-trace searches.