Sample Size Calculator

You need four inputs: the minimum effect size you want to detect, the expected variability, your significance level (alpha, typically 0.05), and your desired statistical power (typically 0.80). This calculator combines these to compute the sample size needed.

Power is the probability of detecting a real effect (1 - beta). Power of 0.80 means an 80% chance of detecting the effect if it exists. Higher power requires more samples but reduces the chance of a false negative (Type II error).

Cohen's d is a standardized effect size for t-tests: d = (mean difference) / (standard deviation). Convention: 0.2 = small, 0.5 = medium, 0.8 = large. Use it when you don't know the exact scale of your measurements but can estimate relative magnitude.

Alpha = 0.05 is the most common choice (5% chance of false positive). Use 0.01 for stricter tests (medical, safety-critical). Use 0.10 for exploratory analysis or screening experiments where false negatives are costlier.