Effect Size Calculator

Effect size measures the magnitude or practical significance of a relationship or difference, independent of sample size. While p-values tell whether an effect exists (statistical significance), effect sizes tell how large the effect is (practical significance). With large samples, even trivial differences can be statistically significant (p < .05). Effect sizes help interpret whether findings matter in real-world contexts. They enable meta-analysis by standardizing results across studies with different metrics. Report effect sizes with all statistical tests to give readers complete information about your findings' magnitude and importance.

Cohen's d conventions (differences between means in standard deviation units): d = 0.20 is small (subtle difference), d = 0.50 is medium (noticeable difference), d = 0.80 is large (substantial difference). However, these are rough guidelines - interpretation depends on context. In educational interventions, d = 0.20 may be meaningful if affecting thousands of students. In clinical psychology, d = 0.80 may be modest for intensive therapy. Consider practical significance in your field, costs of implementation, and comparison to existing interventions when interpreting effect sizes.

Yes! Effect sizes should be reported for all analyses, significant or not. Non-significant results with moderate effect sizes suggest inadequate statistical power (sample too small) rather than no effect. Non-significant results with small effect sizes indicate truly minimal effects. Reporting effect sizes for non-significant findings helps prevent publication bias and enables meta-analysts to include your data. It provides more complete information than p-values alone. Always report effect sizes with confidence intervals to show precision of estimates.