ABSTRACT
This chapter discusses the sample size calculations for clinical trials where the expectation is that the data will take a plausibly Normal form. It highlights that the variance used in the sample size calculation should reflect the planned analysis. This is because using an inappropriate variance could lead to a substantial over-(or under-)estimate of the variance and consequently the sample size.
The chapter describes how to undertake sensitivity analyses around the sample size calculations when designing a trial and also how to account for the imprecision in the variance estimate when estimating the sample size.
