Inference for Spillovers in Networks

Topic: 
Inference for Spillovers in Networks
Date & Time: 
Wednesday, March 7, 2018 -
17:00 to 18:30
Speaker: 
Dean Eckles, Assistant Professor at MIT Sloan School of Management
Location: 
Room 502, NYU Shanghai

Social and behavioral scientists are interested in testing of hypotheses about spillovers (i.e. interference, exogenous peer effects) in social networks; and similar questions may arise in other settings (e.g., biological and computer networks). However, when there is a single network, this is complicated by lack of independent observations. We explore Fisherian randomization inference as an approach to exact finite-population inference, where the main problem is that the relevant hypotheses are non-sharp null hypotheses. Fisherian randomization inference can be used to test these hypotheses either by (a) making the hypotheses sharp by assuming a model for direct effects or (b) conducting conditional randomization inference such that the hypotheses are sharp. I present both of these approaches, the latter of which is developed in Aronow (2012) and our paper (Athey, Eckles & Imbens, 2017). This usually involves selecting some vertices to be “focal” and conditioning on their treatment assignment and/or the assignment of some of all of their network neighbors. The selection of this set can present interesting algorithmic questions; we, for example, make use of greedy methods for finding maximal independent sets. Dean Eckles illustrates these methods with application to a large voter turnout experiment on Facebook (Jones et al., 2017).

Dean Eckles is a social scientist, statistician, and faculty at the Massachusetts Institute of Technology (MIT). Dean is the KDD Career Development Professor in Communications and Technology, an assistant professor in the MIT Sloan School of Management, and affiliated faculty at the MIT Institute for Data, Systems & Society. He was previously a member of the Core Data Science team at Facebook. Much of his research examines how interactive technologies affect human behavior by mediating, amplifying, and directing social influence — and statistical methods to study these processes. Dean’s empirical work uses large field experiments and observational studies. His research appears in the Proceedings of the National Academy of Sciences and other peer-reviewed journals and proceedings in statistics, computer science, and marketing. Dean holds degrees from Stanford University in philosophy (BA), cognitive science (BS, MS), statistics (MS), and communication (PhD).

Introduction and moderation of the Q&A by Jeff Lee, Visiting Assistant Professor of Marketing.

The talk is co-sponsored by NYU Shanghai Center for Data Science & Center for Business Education and Research.

Please click here to RSVP.