Statistical primer for athletic trainers: Using confidence intervals and effect sizes to evaluate clinical meaningfulness

Objective: To describe confidence intervals (CIs) and effect sizes and provide practical examples to assist clinicians in assessing clinical meaningfulness. Background: As discussed in our first article in 2015, which addressed the difference between statistical significance and clinical meaningfulness, evaluating the clinical meaningfulness of a research study remains a challenge to many readers. In this paper, we will build on this topic by examining CIs and effect sizes. Description: A CI is a range estimated from sample data (the data we collect) that is likely to include the population parameter (value) of interest. Conceptually, this constitutes the lower and upper limits of the sample data, which would likely include, for example, the mean from the unknown population. An effect size is the magnitude of difference between 2 means. When a statistically significant difference exists between 2 means, effect size is used to describe how large or small that difference actually is. Confidence intervals and effect sizes enhance the practical interpretation of research results. Recommendations: Along with statistical significance, the CI and effect size can assist practitioners in better understanding the clinical meaningfulness of a research study.