Contrast-specific propensity scores

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