Delta: The True Clinically Significant Difference to be Detected

Abstract

Sample size determination is a very important aspect of planning a clinical trial. The actual calculation is usually the responsibility of the project statistician, but fruitful communications with the clinical monitor are required. When the variable of interest follows a normal distribution, the statistician must have specified values of the error variance and a difference. Some consequences of misspecification of these values, with emphasis on the difference, are presented. Some discussion of the role of communication between the project statistician and the clinical monitor is also presented.