Statistical Power & Sample Size (t-test)

Find the sample size per group needed for a two-sample t-test, or the power you achieve with a given sample size, free online.

Solve for

Effect size (Cohen's d)

Significance level α

Desired power (0–1) — for sample size (optional)

Sample size per group — for power (optional)

About this calculator

For a two-sided, two-sample comparison with effect size d (Cohen's d): n per group ≈ 2(z₁₋α/₂ + z_power)² ÷ d² + z₁₋α/₂² ÷ 4. Power for a given n is Φ(d√(n/2) − z₁₋α/₂) plus a small term for the opposite tail. Aim for 80% power or higher.