Bayesian Model Averaging for Propensity Score Analysis

Abstract

This article considers Bayesian model averaging as a means of addressing uncertainty in the selection of variables in the propensity score equation. We investigate an approximate Bayesian model averaging approach based on the model-averaged propensity score estimates produced by the R package BMA but that ignores uncertainty in the propensity score. We also provide a fully Bayesian model averaging approach via Markov chain Monte Carlo sampling (MCMC) to account for uncertainty in both parameters and models. A detailed study of our approach examines the differences in the causal estimate when incorporating noninformative versus informative priors in the model averaging stage. We examine these approaches under common methods of propensity score implementation. In addition, we evaluate the impact of changing the size of Occam’s window used to narrow down the range of possible models. We also assess the predictive performance of both Bayesian model averaging propensity score approaches and compare it with the case without Bayesian model averaging. Overall, results show that both Bayesian model averaging propensity score approaches recover the treatment effect estimates well and generally provide larger uncertainty estimates, as expected. Both Bayesian model averaging approaches offer slightly better prediction of the propensity score compared with the Bayesian approach with a single propensity score equation. Covariate balance checks for the case study show that both Bayesian model averaging approaches offer good balance. The fully Bayesian model averaging approach also provides posterior probability intervals of the balance indices.

Derivation and validation of a universal vital assessment (UVA) score: a tool for predicting mortality in adult hospitalised patients in sub-Saharan Africa

“…We used the Bayesian model averaging package in R to select the clinical variables that produced the model with the minimal Bayesian information criterion. 18 We then used the rpart package in R to create decision trees for each selected clinical variable on in-hospital mortality. For the decision tree analysis, we upsampled in-hospital mortality to adjust for the class imbalance of mortality.…”

Bayesian estimation of the average treatment effect on the treated using inverse weighting

“…Following the works of Kaplan and Chen and Zigler, we now consider a more complex structure for the treatment model. More specifically, we assume Z i ∼ Bern ( e i ), where , where X 1 ∼ N (1,1), X 2 ∼ Poisson (2), X 3 ∼ Bernoulli (0.5), X 4 ∼ N (0,1), X 5 ∼ N (1,1), X 6 ∼ N (0,1), and | X 4 | represents the absolute value of X 4 .…”

“…Second, it was shown that the point estimator does not possess good small‐sample properties . Kaplan and Chen also examined the differences in the causal estimate when incorporating noninformative versus informative priors in a model averaging stage . Zigler et al developed a Bayesian strategy that augments the outcome model with additional individual covariates .…”

Introductory Chapter: Spatial Analysis, Modelling, and Planning

“…Instead of selecting the only best model and accepting that it properly defines the data generation process, a group of models analyze all the possible models to be resultant from the existing variables set and combine the results through a mul tiplicity of techniques, for example, bootstrap aggregation, bagging, boosting, support vector machines, neural networks, genetic algorithms, and Bayesian model averaging [21,22]. The subsequent group of models constantly achieves better results than the designated best model making higher accurate predictions across a wide group of domains and techniques [23,24].…”