Research Measurements
Effect Size
A measure of the magnitude of a treatment's impact, independent of sample size.
Effect size quantifies the magnitude of a difference or relationship observed in a study, independent of sample size. Common measures include Cohen's d (standardized mean difference: small=0.2, medium=0.5, large=0.8), relative risk, odds ratio, and the correlation coefficient. Unlike p-values, effect sizes indicate the practical significance of a finding. A supplement study might show a statistically significant (p < 0.05) but clinically negligible effect size. Conversely, a meaningful effect size in a small study may not reach statistical significance. Meta-analyses report pooled effect sizes across multiple studies.