Skip to content

Statistical Power Calculator

Find the power of a comparison between two groups at the sample size you can achieve. The calculator below compares two proportions. Use the link for two means if your outcome is a continuous measurement.

0.95 is the usual choice. It sets the two-sided significance level (alpha = 1 - confidence).

1 means equal group sizes. 2 means group 2 is twice as large as group 1.

When to use this calculator

  • Your sample size is limited, for example by a fixed number of patients, and you want to know the power you have.
  • You are checking a plan for a sensitivity analysis, to see how power changes if the effect is smaller than expected.
  • An ethics committee or supervisor has asked you to justify that the study can detect a meaningful difference.

How to use it

  1. Enter the number of participants in the first group and the expected proportions in both groups.
  2. Set the confidence level and the allocation ratio between the groups.
  3. Read the power. Try smaller differences to see how quickly it falls.
  4. If the power is too low, plan more participants with the matching sample size calculator.

How to read the result

Power is the probability that the study detects a true difference of the size you entered, using your sample size and significance level. Many studies aim for 80% or 90%.

Choose a difference that would matter in practice, not simply the one that makes the study feasible.

Power should be planned before the data are collected. Calculating power afterwards from the observed result adds nothing beyond the p-value, so report a confidence interval instead.

Frequently asked questions

What is statistical power?
The probability of detecting an effect of a stated size when the effect truly exists. It is one minus the chance of a false negative.
What is a good level of power?
Conventions of 80% or 90% are widely used. The right level depends on the consequences of missing a real effect, so justify your choice.
How can I increase power?
Increase the sample size, study a larger effect, reduce variability in the outcome, or accept a higher significance level. Each choice has costs that you should state.
Is this the same as a sample size calculation?
They use the same quantities. A sample size calculation solves for the number of participants at a chosen power. A power calculation solves for the power at a fixed number of participants.
Can I use this after my study to show it was powered?
Power calculated from the observed result is not informative. Report the effect estimate with its confidence interval instead.

Guides

Related calculators

Use this in a full research protocol

A free ToProt account lets you save calculations to a study, build your protocol and objectives, manage references and generate your data dictionary in one place.