s_pooled = √(((n₁−1)s₁² + (n₂−1)s₂²)/(n₁+n₂−2)); d = (M₁−M₂)/s_pooled; Hedges' g applies a small-sample correction.
About this tool
The Cohen's d Effect Size Calculator measures how far apart two group means are in standard-deviation units — the standardized mean difference that tells you whether a statistically significant result is also practically large. Enter summary statistics (each group's mean, standard deviation and sample size) or switch to raw-data mode and paste the two samples; the calculation happens entirely in your browser.
The pooled standard deviation combines both groups' spread as s_pooled = √(((n₁−1)s₁² + (n₂−1)s₂²) / (n₁+n₂−2)), and Cohen's d = (M₁ − M₂) / s_pooled. Because d is slightly biased upward in small samples, the tool also reports Hedges' g = d·(1 − 3/(4(n₁+n₂) − 9)), a bias-corrected version that is preferable for small n. The sign simply reflects which mean is larger; the magnitude is what matters.
Cohen's benchmarks classify the absolute effect as small (0.2), medium (0.5) or large (0.8), and the tool labels your result on that scale. In raw-data mode each group needs at least two values; in summary mode each sample size must be at least two and the standard deviations non-negative, so the pooled SD is defined.
Frequently asked questions
How is the pooled standard deviation found?
It is the square root of the weighted average of the two variances: s_pooled = √(((n₁−1)s₁² + (n₂−1)s₂²)/(n₁+n₂−2)). This assumes the two groups have roughly equal variance, the standard assumption behind Cohen's d.
How do I interpret the size of d?
Using Cohen's conventions on the absolute value: about 0.2 is small, 0.5 medium and 0.8 large. Below 0.2 is negligible. These are rules of thumb — what is meaningful depends on your field and the cost of the effect.
What is Hedges' g and when should I use it?
Hedges' g is Cohen's d multiplied by a small-sample correction factor, 1 − 3/(4(n₁+n₂)−9). It removes the slight upward bias of d and is recommended whenever the combined sample size is small (roughly under 20).
Why is my d negative?
The sign only shows direction: d is negative when the first group's mean is smaller than the second's, since d = (M₁ − M₂)/s_pooled. The effect size is the magnitude — swap the group order to flip the sign without changing how large the effect is.