PLAN · TOOL 5 OF 13

Sample size calculator

Prevalence, two proportions or two means, with design effect and non-response. Gives a ready Methods sentence.

What is your main objective?

Expected prevalence (%)From the literature or a pilot; use 50 if unknownAbsolute precision (± %)Often 5Population sizeOptional, for a small finite population

Significance (α, two-sided)

Design effect1 for simple random; often 1.5–2 for cluster samplingExpected non-response (%)

REQUIRED SAMPLE SIZE

359participants

Sample size calculation (single proportion / prevalence)
n = Z²·p(1−p)/d² = 1.96² × 0.3 × 0.70 / 0.05²
Two-sided α = 0.05
Base sample size: 323
Adjusted for 10% non-response: 359
Suggested Methods sentence: “Assuming an expected prevalence of 30% and an absolute precision of 5%, with 95% confidence and 10% non-response, the required sample size was 359.”

Standard formulas for planning. For complex designs (cluster trials, survival or regression models), confirm with a statistician or software such as G*Power or OpenEpi.

NEXT TOOLReporting guideline finder →