This proportion confidence interval calculator gives a range of plausible values from a sample. The range is an estimate of the true population proportion. Enter the number of successes x and the sample size n, or a sample percent and n, choose 90%, 95% or 99% confidence, and pick a method. The result shows the lower and upper limits, the sample proportion, the critical value, a success–failure check and the interval drawn on a 0–100% scale. A sample proportion is itself a ratio, x out of n, the same kind of part-to-whole comparison our proportion calculator works with.
How to Use the Proportion Confidence Interval Calculator
- From counts: type the number of successes x and the sample size n.
- From a percent: type a proportion p or successes count via its percent and n; the calculator rounds p × n to whole successes.
- Choose the confidence level selector value: 90%, 95% or 99%.
- Choose a method; Wilson score is the default.
- Compare methods: see all five intervals for the same data side by side.
There is no calculate button. The output display updates as you type in the number of successes input or the sample size input, and the reset button restores the example.
Proportion Confidence Interval Calculator Uses
One page covers the common ways to build an interval estimate for a share. As a Wilson score interval calculator, it uses the default method. It is also a Clopper-Pearson interval calculator. Some call that an exact binomial confidence interval calculator. The method menu turns it into an Agresti-Coull interval calculator. It can also act as a Wald interval calculator. A Jeffreys interval calculator uses a Bayesian beta prior instead; this page sticks to the four classical methods plus continuity correction.
Some people search for a proportion CI calculator. Others say binomial confidence interval calculator. Statistics courses call it a one-proportion confidence interval calculator. Others say one-sample proportion confidence interval calculator. Researchers may call it a population proportion confidence interval calculator. It is a confidence interval calculator for proportions only. For two groups, a two-sample proportion confidence interval calculator compares p₁ − p₂.
It runs as a confidence interval online tool with no sign-up. As a statistical interval calculator, it reports both limits. It doubles as a proportion statistics calculator for p̂, the standard error and z*. Students use it as a sample proportion calculator too. Behind the scenes, the algorithm handles the standard error computation and the critical value lookup, the input fields accept whole counts, and the user interface keeps all three modes on one card.
What Is a Confidence Interval for a Population Proportion?
A confidence interval for a proportion is a range, computed from a sample, that is likely to contain the true population proportion. A 95% interval means the method captures the true value in about 95 of every 100 samples drawn the same way.
The point estimate is p̂ = x ÷ n. The interval adds a margin around it to reflect sampling error: bigger samples give narrower intervals, and higher confidence gives wider ones.
Confidence Interval Formula (Population Proportion)
The textbook formula is the Wald interval. Wald methods use the normal approximation. Most courses teach this normal distribution approximation Wald interval first.
Formula
p̂ ± z* × √(p̂(1 − p̂) ÷ n) p̂ = x ÷ n is the sample proportion, n the sample size, and z* the critical value: 1.645 for 90%, 1.960 for 95%, 2.576 for 99%. The square root term is the standard error.
A Wilson score interval centres the range on an adjusted proportion and keeps it inside 0% to 100%:
Formula
[p̂ + z²/(2n) ± z √(p̂(1 − p̂)/n + z²/(4n²))] ÷ (1 + z²/n) Wilson (1927). It has better coverage than Wald for small samples and proportions near 0 or 1.
The Agresti-Coull interval adds z²/2 successes and z²/2 failures, then applies the Wald formula. The Clopper-Pearson interval inverts the binomial distribution exactly, through beta quantiles, and is the most conservative.
How to Calculate a Confidence Interval for a Proportion
Example 1: Confidence Interval for a Proportion Using Summary Data
A poll of 1,000 voters finds 52% support. With the Wald method at 95%: SE = √(0.52 × 0.48 ÷ 1000) ≈ 0.0158, margin = 1.96 × 0.0158 ≈ 0.031, so the interval is 48.9% to 55.1%.
To verify Example 1 using the calculator, choose From a percent, type 52 and 1000, and select Wald.
Example 2: Confidence Interval for a Proportion Using Count Data
54 of 120 patients respond to a treatment, so p̂ = 0.45. The 95% Wilson score interval is 36.39% to 53.92%.
To verify Example 2 using the calculator, keep From counts with x = 54 and n = 120 and the Wilson method.
How to Interpret a Confidence Interval for a Proportion
Read the result as: "We are 95% confident that the true proportion lies between the lower and the upper limit." The confidence interval says nothing about individual people in the sample. The confidence belongs to the method, not to one particular interval; any single interval either contains the true value or it does not.
If a claimed value, such as 50%, falls outside the interval, the data are evidence against it at the matching significance level. The One Proportion Z Test Calculator gives the p-value for that comparison.
When Can You Use a One-Proportion Z Confidence Interval?
- Random sample: each member of the population had a known chance of selection.
- Independence: the sample is no more than 10% of the population, or sampling is with replacement.
- Success–failure condition: at least 10 successes and 10 failures (np̂ ≥ 10 and n(1 − p̂) ≥ 10) for the Wald method.
When the success–failure condition fails, the calculator says so, and the Wilson or Clopper-Pearson method is the safer choice. In every case the calculator assumed that you sampled at random.
Alternative Scenarios
With 4 successes in 40 trials, the five methods disagree noticeably at 95%:
| Method | Lower | Upper |
|---|---|---|
| Wilson score | 3.96% | 23.05% |
| Clopper-Pearson (exact) | 2.79% | 23.66% |
| Agresti-Coull | 3.39% | 23.62% |
| Wilson with continuity correction | 3.25% | 24.60% |
| Wald | 0.70% | 19.30% |
The Wald interval is the narrowest and the least reliable here, because only 4 successes fall far short of 10. Brown, Cai and DasGupta (2001) showed that Wald coverage can drop well below the nominal 95% even for moderate samples, which is why this page defaults to Wilson.
When the sample is a large share of a finite population, multiply the standard error by √((N − n) ÷ (N − 1)), the finite population correction. For two groups, use the Two Proportion Z Test Calculator or a two-sample interval.
More Confidence Interval Examples
These confidence interval examples come from clinical research and survey work.
Prevalence
Problem: A screening study finds a condition in 36 of 600 adults. What is the 95% interval for its prevalence?
Solution: p̂ = 0.06. The Wilson interval is about 4.37% to 8.19%. A public-health study funded by the National Institutes of Health (NIH) would usually report exactly this kind of CI for proportion data.
Customer survey at 99%
Problem: 312 of 400 customers would buy again. What is the 99% interval?
Solution: p̂ = 0.78. The 99% Wilson interval is about 72.2% to 82.9%, wider than the 95% interval (73.7% to 81.8%).
To calculate confidence interval limits by hand, follow the formula section above. To estimate population proportion values quickly, use the From counts mode. A 95 confidence interval uses z* = 1.960. A 99 confidence interval uses z* = 2.576. Each one is a confidence interval for proportion data, not for means. Statisticians use three names for this range. One is the binomial proportion confidence interval. Another is the population proportion confidence interval. A third is the sample proportion confidence interval. A one proportion confidence interval needs only x and n.
The margin of error proportion term in a poll is half the width of a Wald interval. The standard error proportion term, √(p̂(1 − p̂)/n), sets that width. The critical value proportion term z* scales it. The z-score proportion link is direct: z* is the z score that leaves the chosen tail area. Most textbooks quote the Wald form as a confidence interval formula proportion students learn first. A Wald confidence interval is simple but can mislead with small counts. The Agresti Coull interval fixes that. So does the Clopper Pearson interval. A normal approximation proportion rule of thumb is np̂ ≥ 10 and n(1 − p̂) ≥ 10. The lower bound and upper bound are also called the confidence limits. The confidence coefficient is 1 − α. The sampling distribution of p̂ has standard deviation √(p(1 − p)/n).
Definitions
Sample Proportion
The sample proportion p̂ = x ÷ n is the share of the sample with the trait, such as 54 of 120 = 0.45.
Confidence Level
The confidence level is the long-run share of intervals that capture the true proportion, set by the critical value z*.
Sample Size
The sample size n is the number of observations. Quadrupling n roughly halves the width of the interval. To plan n before collecting data, use the Sample Size Calculator Proportion.
Population Size
The population size N matters only when the sample is more than about 10% of it; then the finite population correction narrows the interval.
Sample size calculators work backwards from a target width to n. The one proportion interval on this page follows simple logic. A two proportion interval for comparing groups follows the same logic. The proportion margin error shrinks with the square root of n. A higher proportion confidence level widens the range. Each success proportion is just x ÷ n. This confidence interval tool has a second job. It serves as an interval estimate calculator. Its output is a population proportion interval.
Proportion Confidence Interval FAQ
Which method should I use?
Use a Wilson score interval by default. Among the common choices, the Wilson score method has the best overall coverage. Use Clopper-Pearson when you need guaranteed coverage, and avoid Wald when successes or failures number fewer than 10.
What does 95% confidence mean?
It means 95% of intervals built this way contain the true proportion. It is not a 95% probability that one fixed interval contains it.
How do I make the interval narrower?
Increase the sample size or lower the confidence level. Four times the sample gives about half the width.
Can the interval go below 0% or above 100%?
The Wald interval can, but Wilson, Agresti-Coull and Clopper-Pearson stay inside 0% to 100%.
What if I have 0 successes?
Use Wilson or Clopper-Pearson; both give a lower limit of 0 and a sensible upper limit. For 0 of 30 at 95%, Clopper-Pearson's upper limit is about 11.6%.
Is the margin of error the same as half the interval?
Yes for Wald, which is symmetric. Wilson and exact intervals are not symmetric around p̂, so the calculator reports their half-width separately.
References
- Wilson, E. B. (1927) – Probable inference, the law of succession, and statistical inference, JASA 22(158)
- Brown, Cai & DasGupta (2001) – Interval estimation for a binomial proportion, Statistical Science 16(2)
- Agresti & Coull (1998) – Approximate is better than exact, The American Statistician 52(2)
- Clopper & Pearson (1934) – The use of confidence or fiducial limits, Biometrika 26(4)
Related Calculators
- Sample Size Calculator Proportion: sample size needed to estimate a proportion
- One Proportion Z Test Calculator: test one proportion against a claimed value
- Two Proportion Confidence Interval Calculator: interval for the difference of two proportions