Free tool · no sign-up needed

Sample size calculator for medical research

Calculate the sample size for a prevalence survey, a mean, or a comparison of two proportions or two means, with power, confidence level and dropout.

Study design

Expected proportion

From a similar study; 50% if unknown.

Margin of error

For a proportion, in percentage points (5 means ±5%). For a mean, in the outcome's units.

Standard deviation (σ)

From a pilot or a published study, in the outcome's units.

Proportion in group 1

e.g. outcome rate with the usual treatment.

Proportion in group 2

The rate you expect, or the smallest difference worth finding.

Difference to detect (Δ)

The smallest difference in means that matters clinically.

Population size (optional)

For a small, known population. Leave blank if large.

Confidence level

Power

Expected dropout (optional)

Questions people ask

What prevalence should I use if I don't know it?

Use a figure from a similar published study. If there is none, 50% gives the largest (most cautious) sample size.

What is the margin of error?

How close you want your estimate to be to the true value, e.g. ±5 percentage points. Halving the margin roughly quadruples the sample size.

What power should I choose?

80% is the usual minimum; 90% is common for trials. Power is the chance of detecting the difference if it really exists.

Why add for dropout?

Some participants won't complete the study. Dividing by (1 − dropout rate) means you still have enough people at the end.

Is this enough for my protocol?

It covers simple designs with normal-approximation formulas. Cluster sampling, matched designs, survival outcomes and multiple groups need other methods; ask a statistician for those.

More free tools

NEET score calculatorCorrect, wrong and unattempted per subject → your score out of 720. NEET rank and percentile estimatorYour marks → an estimated rank and percentile from past years. Citation generatorAPA, Vancouver, Harvard and MLA references, or paste a RIS/BibTeX/PubMed record. Study-hours plannerHours until exam day, split by subject and phase.