Skip to content
Toolden

TOOLDENCALCULATORSON-DEVICE

Sample Size Calculator

How many people do you need to survey? Work it out from your population, confidence level and margin of error.

Enter the population and margin of error, then pick a level. The required sample updates live in the results.
Enter the parameters, or load the sample.

No sample size yet. Enter the parameters to begin.

HOW TO USE

  1. 1Enter the size of the population you are sampling, the total group your survey represents.
  2. 2Pick the confidence level: 90%, 95% or 99%. The higher the confidence, the more people you need.
  3. 3Set the margin of error you can tolerate, e.g. 5% means results are within ±5 percentage points of the true value.
  4. 4The required sample size uses the worst-case 50/50 split, which gives the largest possible sample. Halving the margin of error roughly quadruples the sample you need.

EXAMPLE

INPUT

Population 1,000, 95% confidence, 5% margin of error

OUTPUT

278 people

FAQ

What formula does the calculator use?
It starts from the standard sample-size formula for proportions with the worst-case 50/50 split: n₀ = z²·0.25 ÷ E², then applies the finite population correction for small populations. For a population of 1,000 at 95% confidence with 5% margin of error, you need 278 people.
Why do smaller populations still need so many people?
The sample size depends mainly on the desired precision, not the population. The finite population correction only kicks in noticeably once the sample would exceed roughly 5% of the population, a population of 10,000 needs 370 people, barely more than 385 for an effectively infinite one.
What does the 50/50 split assumption mean?
When estimating a proportion (like support for a candidate), the most uncertain case is a 50/50 split. Using it guarantees the sample is large enough for any actual result. If you expect a lopsided split you can use a smaller sample, but this calculator always stays safe.
Why does my sample sometimes equal the whole population?
With very small populations or very tight precision requirements, the formula can push the sample to the full population size, at that point you simply survey everyone.
Is my data sent anywhere?
No. All calculations run entirely in your browser, your numbers never leave your device.

RELATED TOOLS