Effect Size Type
Visual Representation
Effect Size Reference Guide
| Measure |
Small |
Medium |
Large |
| Cohen's d |
0.2 |
0.5 |
0.8 |
| Pearson's r |
0.1 |
0.3 |
0.5 |
| Eta-squared (η²) |
0.01 |
0.06 |
0.14 |
| Cramér's V (df=1) |
0.1 |
0.3 |
0.5 |
Note: These are Cohen's conventional benchmarks. Always consider context and field-specific norms when interpreting effect sizes.
Why Effect Sizes Matter
- Statistical vs Practical Significance: A p-value tells you if an effect exists, but effect size tells you if it matters
- Sample Size Independence: Unlike p-values, effect sizes don't inflate with larger samples
- Comparison Across Studies: Effect sizes enable meta-analysis and cross-study comparisons
- Power Analysis: Effect sizes are essential for determining required sample sizes
- Real-World Impact: They quantify the practical importance of findings