Contrast Specific Propensity Scores

TL;DR

This paper introduces contrast-specific propensity scores (CSPS) to enable balanced comparisons among multiple treatments, extending traditional propensity score methods beyond simple treatment-control contrasts.

Contribution

It proposes a novel methodology for creating balanced treatment groups for complex contrasts involving multiple treatments, improving causal inference in multifaceted settings.

Findings

CSPS effectively balance covariates for specified contrasts.

The method generalizes propensity scores to multi-treatment comparisons.

Potential for improved causal analysis in complex treatment scenarios.

Abstract

Basic propensity score methodology is designed to balance multivariate pre-treatment covariates when comparing one active treatment with one control treatment. Practical settings often involve comparing more than two treatments, where more complicated contrasts than the basic treatment-control one,(1,-1), are relevant. Here, we propose the use of contrast-specific propensity scores (CSPS). CSPS allow the creation of treatment groups of units that are balanced with respect to bifurcations of the specified contrasts and the multivariate space spanned by them.