Definition
Effect size expresses how large a difference is or how strong an association is — independently of sample size. Cohen's d gives the difference of means in standard-deviation units, η² or R² the share of explained variance, the correlation coefficient the strength of a relationship. It complements the p-value, which says nothing about magnitude.
Foundations & hypothesis testing
Rough benchmarks after Cohen: d ≈ 0.2 small, 0.5 medium, 0.8 large; for correlations 0.1 / 0.3 / 0.5. Always compare them with practical relevance in your field — a 0.5 % yield improvement can be an enormous effect in chemical production.
Effect size is also the input for sample-size planning: the smaller the effect you want to detect, the larger the sample you need. Report it with a confidence interval; journals and regulators expect it today.
In Statistica
In the ANOVA/MANOVA module Statistica computes partial η²; correlation coefficients and R² are in Basic Statistics and Multiple Regression. Cohen's d is easily derived from the t-test table (difference of means divided by the pooled standard deviation), and the Power Analysis module uses it to plan sample size.
Related terms
- p-valueThe p-value is the probability of obtaining a result at least as extreme as the one observed if the null hypothesis…
- Confidence intervalA confidence interval is a range of values that, with a stated confidence (most often 95 %), covers the true value of a…
- Statistical power and sample sizeStatistical power is the probability that a test detects an effect that really exists (1 − β).
- CorrelationCorrelation measures the strength and direction of a linear relationship between two quantities with a number from −1 to 1.
Knowledgebase guides
FAQ
- Can a result be significant with a small effect?
- Yes — with large samples it is routine. That is why you should decide by the effect size and its interval, not by p alone.
- Which effect size is "large enough"?
- The one that has practical impact in your context. Cohen's benchmarks are only a fallback when no field standard exists.
Try it on your own data
Statistica free for 30 days
Full version, no credit card. Or get a pricing estimate in a minute.
Foundations & hypothesis testing
Updated: September 2026.