Propensity score matching for multiple treatment levels: A CODA-based contribution

This study proposes a simple technique for propensity score matching for multiple treatment levels under the strong unconfoundedness assumption with the help of the Aitchison distance proposed in the field of compositional data analysis (CODA).