Reproducible measurement of treatment effects requires studies that can reliably distinguish between systematic treatment effects and noise resulting from biological ...