Reporting Effect Sizes in Practice
You'll learn how to communicate both statistical and practical significance by reporting effect sizes alongside p-values.
Why Report Effect Sizes?
A p-value tells you whether an effect is statistically significant—that is, unlikely to be due to chance. But it doesn't tell you whether the effect is big enough to matter. With large sample sizes, even tiny, trivial differences can yield p < 0.05. Effect sizes answer the "so what?" question: how meaningful is this finding in real-world terms?
Best Practices for Reporting
Always report both p-value and effect size together. A complete result statement looks like this:
"The treatment group scored significantly higher than the control group (t(98) = 3.45, p = 0.001, Cohen's d = 0.68, 95% CI [0.28, 1.08])."
This tells your audience:
- Statistical significance (p = 0.001): The effect is unlikely due to chance
- Practical significance (d = 0.68): The effect is medium-to-large in magnitude
- Precision (95% CI): We're reasonably confident the true effect is between 0.28 and 1.08
What to Include
- The effect size value (e.g., Cohen's d = 0.68)
- Confidence interval for the effect size (shows uncertainty)
- Interpretation guideline context (small/medium/large, or domain-specific benchmarks)
- Sample size (helps readers judge generalizability)
Common Pitfalls to Avoid
- Reporting p-values alone (leaves readers guessing about importance)
- Saying "significant" without clarifying whether you mean statistical or practical
- Cherry-picking which effect sizes to report (report all primary analyses)
Key Takeaway: Report effect sizes with confidence intervals alongside p-values to give a complete picture of both whether an effect exists and whether it matters practically.