Sample Size Estimation for Trials and Other Two Group Comparisons

Abstract

When an investigator is designing a study to compare two groups with a continuous outcome, an essential step is the calculation of sample sizes that will allow a reasonable chance (power) of detecting a predetermined target difference (effect size), if it truly exists, at a given level of statistical significance. Four key components are required to facilitate a suitable sample size calculation. These are the target difference or effect size, the standard deviation of the outcome, and the Type I and Type II errors. This case study will illustrate the process and methods for determining the required sample size using a real case study as an example—the SELF randomized controlled trial to compare a self-managed single exercise program versus usual physiotherapy treatment for rotator cuff tendinopathy.

locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles