Georgia Papadogeorgou, of the Department of Biostatistics at Harvard T.H. Chan School of Public Health, will present: “Statistical Challenges in Air Pollution Research: From Spatial Confounding to Interference”.
Abstract: In the studies of air pollution we are often challenged with assumptions that are likely to not hold, such as no unmeasured confounding and no interference. However, complex climatological and atmospheric processes are known to vary spatially, have strong associations with ambient air pollution, and are often unmeasured. Furthermore, pollution from one point source can affect ambient pollution concentrations at long distances “downwind”. The causal inference methods presented are motivated by and applied to a comparative effectiveness investigation of power plant emission reduction technologies designed to reduce population exposure to ambient ozone pollution. In the first part of the talk, I will present a new method called Distance Adjusted Propensity Score Matching (DAPSm) that incorporates information on units’ spatial proximity into a propensity score matching procedure aiming to reduce bias arising from unobserved spatial confounders. In the second part of the talk, I will show that the assumption of no interference is violated in the studies of power plant interventions, and I will propose new estimands and estimators for quantifying the potential benefits of such interventions.
A social tea will be held at 3:00 p.m. in A434 Mayo. All are Welcome.