Janice Weinberg, Professor of Biostatistics at BUSPH, returns to the series to discuss how to appraise the statistical methods section in the paper, Aspirin versus anticoagulation in cervical artery dissection (TREAT-CAD): an open-label, randomized, non-inferiority trial. In this hour, she discusses the choice of a non-inferiority trial, its dependence upon the accepted standard of care, the need for definition of a non-inferiority margin, power and sample size, and special considerations for per protocol and intention to treat analyses in a non-inferiority trial. Special thanks to David Chung, Assistant Professor of Neurology in the Neurocritical Care Division for selecting the paper and providing the clinical context.