Part of the Maths & Statistics suite · 235 calculators

Sample Proportion Calculator

The proportion in a sample, its standard error and a confidence interval — with the Wilson interval, which behaves properly when the count is small.

The sample proportion p̂ = x ÷ n estimates the population proportion.

Results update as you type
Results
Sample proportion
22.5%
Standard error
Margin of error
Wilson interval — lower
Wilson interval — upper
Normal (Wald) — lower
Normal (Wald) — upper
Odds
Reviewed September 2026. Pure mathematics: the result does not depend on where you are. Terminology follows the national curriculum (maths, brackets, decimal point).
No account required · Google Analytics off unless allowedCalculator arithmetic runs in your browserResults update as you type
All calculations run 100% in your browser. The calculator code does not submit your figures to GlobalCalc to obtain a result.
About sample proportion

How the sample proportion calculator works

The sample proportion p̂ = x ÷ n estimates the population proportion. Its standard error is √(p̂(1 − p̂)/n), and the familiar interval is p̂ ± z × SE.

That familiar interval fails badly near 0 and 1 — it can even run below zero. The Wilson score interval solves it by inverting the test rather than approximating it, and is what the calculator leads with for small samples.

Formula: p̂ = x/n; SE = √(p̂(1 − p̂)/n)

Worked examples

InputsSample proportionNote
45 of 20022.5%22.5%, ±5.8 points
A clean run: 0 of 300%Wilson still gives a sensible upper bound
95 of 10095%95%, and the normal interval overshoots

Frequently asked questions

What is a sample proportion?

The share of a sample with some property — 45 of 200 is 22.5%.

Why is the Wilson interval better?

The usual normal interval assumes symmetry that does not hold near 0 or 1. At 0 successes it gives a nonsensical interval of exactly [0, 0]; Wilson gives a proper upper bound.

What is the standard error?

The typical distance between the sample proportion and the true one — √(p(1−p)/n). It shrinks with the square root of n, so quadrupling the sample halves it.

How many do I need for ±3 points?

About 1,067 at 95% confidence, for a proportion near 50%.

What if I saw zero successes?

The rule of three gives a rough upper bound of 3/n. With 30 trials and no events, the true rate could still be as high as about 10%.

Where these figures come from

Last checked: September 2026. Formulas are fixed by mathematics and do not change with tax years or regulations.