Power Analysis Calculator
Test cases replayed : 3 of 3 match · Methodology · No outside source listed
Enter the expected effect size and the group size to get the power of a two-sided test, and the group size needed for the power you want.
Result
Power with this group size
Group size needed for the target power
Estimate for planning only, not tax, legal, financial or medical advice. This calculator uses the method explained on this page.
This browser did not allow copying. The sentence is in the box below, already selected: copy it with your keyboard or the menu.
How to use this calculator
Fill in the fields in this order; each step takes you to its field.
- Effect size (Cohen's d): enter 0 or more.
- Sample size per group: enter 2 or more.
- Significance level α (two-sided): choose 0.01, 0.05 or 0.10.
- Target power (0 to 1): enter a number from 0 to 1.
- Read the result: Power with this group size; the panel also shows Group size needed for the target power.
How this calculator works
Normal approximation for two independent groups of equal size n and a two-sided test: power = Φ(d√(n/2) − z₁₋α/₂) + Φ(−d√(n/2) − z₁₋α/₂), where d is Cohen's d (difference of means ÷ standard deviation) and Φ the standard normal distribution function.
Group size for a target power 1 − β: n = 2 (z₁₋α/₂ + z₁₋β)² ÷ d², rounded up (Cohen, Statistical Power Analysis for the Behavioral Sciences).
A t-test needs one or two more participants per group than this normal approximation for small samples.
The formula
| Symbol | Meaning | Unit |
|---|---|---|
| Effect size (Cohen's d) | ||
| Sample size per group | ||
| Significance level α (two-sided) | ||
| Power with this group size | % | |
normcdf(x, m, s) | the probability that a Gaussian variable with mean m and standard deviation s is at most x (mean zero and standard deviation one without m and s). | |
norminv(p, m, s) | the value that a Gaussian variable with mean m and standard deviation s stays at or below with probability p (mean zero and standard deviation one without m and s). |
Worked example
| Effect size (Cohen's d) | 0.5 |
|---|---|
| Sample size per group | 64 |
| Significance level α (two-sided) | 0.05 |
| Target power (0 to 1) | 0.8 |
| Group size needed for the target power | 63 |
|---|
Power with this group size 80.74%
Each step uses the unrounded value of the step before it; figures on this page are rounded for display.
Frequently asked questions
How many participants per group for a medium effect?
For d = 0.5, α = 0.05 and 80% power, the normal approximation gives 63 per group; the exact t-test needs 64.
What power should a study have?
80% is the usual minimum in Cohen's convention; 90% is common when missing a real effect would be costly.