Online Program Home
My Program

Abstract Details

Activity Number: 7 - Bayesian Nonparametrics in Causal Inference
Type: Invited
Date/Time: Sunday, July 29, 2018 : 2:00 PM to 3:50 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #326657
Title: Bayesian Longitudinal Causal Inference in the Analysis of the Public Health Impact of Air Pollution
Author(s): Chanmin Kim* and Corwin Zigler and Jason Roy and Michael Daniels
Companies: Boston University School of Public Health and Harvard T.H. Chan School of Public Health and University of Pennsylvania and University of Florida
Keywords: Bayesian Nonparametrics; Bayesian Updating Method; Time-varying exposures; Mediation; PM2.5

Pollutant emissions from coal-burning power plants have been deemed to adversely impact ambient air quality and public health conditions. In terms of the chain of accountability (HEI Accountability Working Group, 2003), the link between pollutant emissions from the power plants (SO2) and public health conditions (Cardiovascular and Respiratory diseases) with counting for changes in ambient air quality (PM2.5), we provide the first assessment of the longitudinal effects of specific pollutant emission on public health outcomes that are mediated through changes in the local ambient air quality. It is of particular interest to examine the extent to which the mediated effects differ by seasons. In this paper, we propose a Bayesian approach to estimate time-varying mediation effects. We replace the commonly invoked sequential ignorability assumption with a new set of assumptions which are sufficient to estimate the natural indirect and direct effects. Also, a Bayesian updating model is used to minimize modeling assumptions in longitudinal setting.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program