Randomized Controlled Trials with Minimal Data Retention

Amidst rising appreciation for privacy and data usage rights, researchers have increasingly recognized the principle of data minimization, which holds that the accessibility, collection, and retention of subjects' data should be kept to the minimum necessary to answer focused research questions. Applying this principle to randomized controlled trials (RCTs), this paper presents algorithms for drawing precise inferences from RCTs under stringent data retention and anonymization policies. In particular, we show how to use recursive algorithms to construct running estimates of treatment effects in RCTs, thereby allowing individualized records to be deleted or anonymized shortly after collection. Devoting special attention to the case of non-i.i.d. data, we further demonstrate how to draw robust inferences from RCTs by combining recursive algorithms with bootstrap and federated strategies.