Study Design
Plan a study before you collect the data: how many observations you need, how much power a given sample buys you, and how big the effect actually is.
3 calculator groups Power analysis chart No signup
Study design calculators 3 groups
| # | Calculator | Description | |
|---|---|---|---|
| 1 | Sample Size Calculators | How many observations you need to reach a required margin of error or a required test power — for a confidence interval, Z-test, T-test, χ², F-test, regression, ANOVA, MANOVA and proportions. | |
| 2 | Statistical Power Calculators | The power a given sample size achieves — the chance of detecting a real effect. Each calculator also draws the power analysis chart. | |
| 3 | Effect Size Calculator | How large the difference is, not just whether it is significant: Cohen's d, Cohen's h, Phi (φ), Cramer's V, R squared and Eta squared. |
Which one do I need?
The three are the same equation solved for different unknowns, so which calculator you want depends on what you already know.
Before collecting data — you know the effect size you care about and the power you want, and you need n. Use the sample size calculators.
Sample size already fixed — the data is collected, or n is capped by budget, and you want to know what the study can actually detect. Use the power calculators.
Results in hand — the test came back significant and you need to report how big the difference is. Use the effect size calculator.