Effect Size Calculator

Cohen's d measures the standardized difference between two group means, expressed in standard deviation units. Jacob Cohen's conventional benchmarks classify d = 0.2 as a small effect, d = 0.5 as a medium effect, and d = 0.8 as a large effect. These are guidelines — what counts as meaningful depends on the research context.

Both measure the standardized mean difference, but Hedges' g applies a small-sample correction factor (also called J) that reduces bias when group sizes are small (n < 20). For large samples, the two values are nearly identical. Hedges' g is generally preferred when sample sizes are small or unequal.

Glass's delta uses only the control group's standard deviation as the denominator rather than a pooled SD. It is appropriate when the two groups have substantially different variances, particularly in experimental designs where the control group SD is the best reference for the population.

The effect-size correlation is derived from d using the formula r = d / √(d² + 4). It expresses the effect as a Pearson-type correlation coefficient, ranging from −1 to +1. Values of r ≈ 0.10, 0.24, and 0.37 correspond approximately to Cohen's small, medium, and large benchmarks.

Yes. If you have the t-statistic and degrees of freedom from an independent samples t-test, you can use the formula d = 2t / √(df). This calculator supports both methods — select 'T-Value & Degrees of Freedom' from the Calculation Method selector.

This calculator uses the simple pooled SD formula: s_pooled = √[(SD₁² + SD₂²) / 2], which treats both group SDs equally. When sample sizes are unequal and provided, the Hedges' g calculation uses the sample-size-weighted pooled SD for greater accuracy.

By Cohen's benchmarks, d ≥ 0.8 is considered large, meaning the two group means differ by 0.8 standard deviations — a difference visible to the naked eye in most distributions. However, in fields like social psychology even d = 0.4 may be practically significant, while in precision sciences a much smaller effect might matter greatly.

A p-value tells you whether an effect is likely real (not due to chance), but it does not tell you how large or meaningful it is. With a very large sample, even a trivially small difference can reach statistical significance. Effect sizes provide a scale-independent measure of the magnitude of a finding, making results comparable across different studies and sample sizes.