FREE TOOL
Estimate how many participants per group you need to detect a real difference between two proportions or two means, at your chosen significance level and power.
Required sample size
93 per group
186 total participants across both groups
Two proportions uses the standard normal-approximation formula: n = (zα/2√(2p̄(1−p̄)) + zβ√(p₁(1−p₁)+p₂(1−p₂)))² / (p₁−p₂)², where p̄ is the average of the two proportions.
Two means uses n = 2(zα/2+zβ)²σ² / δ², assuming equal variance in both groups and a two-sided test — the standard formula behind most sample-size tables for a two-sample t-test.
FAQ
80% power means that if a real difference of the size you specified truly exists, your study has an 80% chance of detecting it as statistically significant — a 20% chance of a false negative (Type II error).
A smaller α (e.g. 0.01 instead of 0.05) requires stronger evidence before declaring significance, which needs more data to achieve the same power.
The main result is per group — for a two-arm study you'll need that many participants in each arm, so roughly double for the total study size.
No — this is the minimum analyzable sample size. Inflate it by your expected dropout/loss-to-follow-up rate when planning actual recruitment.
Planning a systematic review instead of a new trial? EvidenceFlow doesn't need a target sample size — it pools whatever data your included studies already report.
Create free account