Title: The Landscape of Causal Inference: Perspectives from the Citation Networks
Date: Oct 17, 2014
Time: 12:30pm – 1:30pm
Abstract: Causal inference is a relatively new but increasingly popular field in statistics. Using papers and their citations that include selected key terminologies in causal inference, we construct citation networks for this particular field. By examining the features of the citation networks and their temporal changes, we are able to present the evolution of the field and its structures from new perspectives. In particular, we study whether the citation dynamics reflects status differences and social processes such as reciprocity, transitivity, and preferential attachment by using Exponential Random Graph Models (ERGMs). As fitting ERGMs on big networks has been unusual and challenging, we explore several new strategies for doing so.
Bio: Dr. Weihua An is assistant professor in the departments of statistics and sociology at Indiana University, Bloomington. His main research areas are social network analysis and causal inference. With a broad academic background, he also has strong interests in organizations, health and social policy, development, etc. His work has been published in premier journals and influential handbooks in the field including Sociological Methodology, Social Forces, Social Networks, Sociological Methods and Research, Social Science Research, Handbook of Causal Analysis for Social Research, and The Sage Handbook of Social Network Analysis. He is also author of two statistical packages in R: “LARF” for instrumental variables estimation of causal effects through local average response functions and “IUPS” for incorporating uncertainties in propensity scores for drawing valid causal inference. For more information about Dr. An, please visit his website.