Sample Size Calculator
Calculates the sample size needed for a two-group comparison, with an optional Bonferroni multiple-testing correction. Works for clinical trials, biomarker discovery and RNA-seq DE analysis.
0.10.5 (Medium)0.8 (Large)1.2 (Very large)2.0
Large — a strong difference
70%80% (convention)90%99%
A single comparison, e.g. a clinical trial primary endpoint
1:12:13:15:1
Required sample size
26
Group 1
+
26
Group 2
=
52
Total
two-sided t-test (exact power) · adjusted α = 5.00e-2 · power = 80%
Quick presets
📌 Assumptions and limitations
- Exact power of the two-sample t-test (equal SDs), the same calculation as G*Power; earlier versions used a z-approximation that came out about one sample per group too small
- Assumes the two groups have equal variances; if the variances differ a lot, you need Welch's t correction
- Cohen's d = (μ₁ − μ₂) / σ_pooled, for a two-sided test
- Bonferroni is a conservative correction; using FDR (Benjamini-Hochberg) can lower the sample size
- Survival and binary outcomes need separate formulas (Schoenfeld, χ²); this tool is for continuous variables only
- In a real study, increase the sample size further to allow for dropout (~10-20%)