Propensity score estimation: machine learning and classification methods as alternatives to logistic regression

By D. Westreich, Justin Lessler, M. J. Funk. 2010

479 citations recorded by the source corpus.

Summary Objective Propensity scores for the analysis of observational data are typically estimated using logistic regression. Our objective in this Review was to assess machine learning alternatives to logistic regression which may accomplish the same goals but with fewer assumptions or greater accuracy. Study Design and Setting We identified alternative methods for propensity score estimation and/or classification from the public health, biostatistics, discrete mathematics, and computer science literature, and evaluated these algorithms for applicability to the problem of propensity score estimation, potential advantages over logistic regression, and ease of use. Results We identified four techniques as alternatives to logistic regression: neural networks, support vector machines, decision trees (CART), and meta-classifiers (in particular, boosting). Conclusion While the assumptions of logistic regression are well understood, those assumptions are frequently ignored. All four alternatives have advantages and disadvantages compared with logistic regression. Boosting (meta-classifiers) and to a lesser extent decision trees (particularly CART) appear to be most promising for use in the context of propensity score analysis, but extensive simulation studies are needed to establish their utility in practice.

Subfields: Artificial Intelligence, Statistics and Probability, Causal Inference & Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

In Causal Inference & Discovery

Citations in the corpus

Deep Neural Networks for Estimation and Inference: Application to Causal Effects and Other Semiparametric Estimands (2018, background): “Deep neural networks have been argued (experimentally) to outperform the previous state-of-the-art (Westreich, Lessler, and Funk (2010), Shalit, Johansson, and Sontag (2017), Hart-ford et al. (2017)).”
Approximate residual balancing: debiased inference of average treatment effects in high dimensions (2016, neutral): “(Some of these methods assume that the propensity scores can be consistently estimated by using a sparse logistic model, whereas others allow for the use of more flexible modelling strategies following, for example, McCaffrey et al. (2004), Van der Laan et al. (2007) or Westreich et al. (2010).)”
Balance diagnostics after propensity score matching. (2019, methodology): “Furthermore, machine learning algorithms such as classification and regression trees (CART), random forest and neural networks can be employed to improve the specification of the propensity score model (15).”
Doubly robust estimation of causal effects. (2011, methodology): “(Alternatively one could use, for example, a probit model or a machine learning approach (9, 10)).”
The role of prediction modeling in propensity score estimation: an evaluation of logistic regression, bCART, and the covariate-balancing propensity score. (2014, result): “Previous studies have shown bCART to perform particularly well for PS estimation in certain settings compared with other nonparametric methods and misspecified logistic PS models (10, 19, 20).”