Propensity score methods are commonly used in studies to figure out cause and effect. This article shows how to keep this process private when dealing with sensitive data. By adding some noise to the scores, the privacy is protected without messing up the results. The researchers tested this method on different types of data and found that it works well in practice, keeping the analysis valid and accurate.