Clinimetrics Corner: Use of Effect Sizes in Describing Data

Effect size is the name provided to a family of indices that are designed to measure magnitude of a targeted intervention. Generally, effect size measures are divided into three groups: 1) standardized difference measures 2) correlational measures, and 3) odds ratios. Effect sizes are valuable in indicating actual magnitude of selected interventions as the values are independent of sample size and statistical significance measures. Effect sizes are used to generate power requirements for a study and are the value of choice when calculating meta-analyses. Typically, effect size measures associate well with global rating of change scores for nearly all interventions.