This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.

Abstract Details

Activity Number: 427
Type: Contributed
Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
Sponsor: IMS
Abstract - #308582
Title: Causal Inference in Observational Studies with Missing Data: A Bayesian Nonparametric Approach
Author(s): Jennifer Hill+ and Jose Zubizarreta*+
Companies: New York University and The Wharton School, University of Pennsylvania
Address: , , NY, 10003, , , PA, 19104,
Keywords: Bayesian backfitting ; Causal inference ; Missing data ; Observational studies ; Sum-of-trees model
Abstract:

This poster presents a new approach to draw causal inferences from observational studies with missing data. In the line of Hill (2010), we modify the Bayesian Additive Regression Trees model (BART; Chipman et al., 2009) to handle missing covariate data and in their presence flexibly model the surface of the potential outcomes. Our modification of BART consists of the implementation of two methods in the algorithm: a proximity matrix, and the incorporation of missingness as a different category in the covariates in the tree-growing process. Both methods perform well in the presence of high levels of selective missing covariate data, as it is shown in simulations with the five dimensional test function used by Friedman (1991) and in the estimation of average treatment effects with the classic LaLonde (1986) example with induced missingness.


The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2010 program




2010 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.