Brian G. Barkley

I am a PhD student in the Department of Biostatistics at The University of North Carolina at Chapel Hill.


My dissertation research is in the field of causal inference. I am interested in drawing inference about treatment effects when interference is present, which is when one individual's treatment may affect another individual's outcome. Interference is sometimes known as spillovers, or, in the area of infectious diseases, herd immunuity or herd protection. I work with my advisor, Michael Hudgens.

Drawing inference about treatment effects when interference is present can be challenging for many reasons. The number of potential outcomes increases from 2 per individual, and the definitions of a treatment effect may change, too. Estimation can also be challenging in its own right. My work centers around data that arise from observational studies where individuals are not randomized to treatment. My projects involve matching and IPTW methods, and machine learning techniques.