How many participants do you need? Enter the smallest effect you care about detecting and the power you want, and get the required n. Runs a priori power analysis for t tests and two-proportion comparisons.
Power analysis beyond the basics? For ANOVA, regression, and more complex designs, run your study planning in ReliCheck Quanta on your Mac. Free to try. No Mac? Try the Quantitative Studio on the web.
Get QuantaAn underpowered study cannot reliably detect the effect it was designed to find, and a significant result from one is more likely to be inflated. Funders, IRBs, and journals increasingly expect an a priori justification of sample size. Deciding n after seeing the data is not a justification.
Base the expected effect on prior studies, a pilot, or the smallest effect that would matter in practice. When in doubt, power for a smaller effect than you hope for; the cost of a few extra participants is lower than the cost of an inconclusive study.
Calculations use the standard normal approximation formulas, n = 2(z1-α/2 + z1-β)²/d² for two groups, with results rounded up. These match tabled values closely; exact noncentral t solutions can differ by a participant or two for very small samples.