How to use
- Enter the success count and total trials.
- Enter the null proportion you want to test against.
- Choose two-sided, greater, or less, then read the exact p-value before moving to CI pages.
Wave 33 exact inference entry
Read an exact one-proportion p-value
The share URL stores only the selected tail and decimal places.
| Successes | Failures | Probability under H0 |
|---|
How to choose between nearby pages
Use this page when the first question is whether one observed proportion looks unusual under one null proportion. Use sample size before data collection, and move to confidence-interval pages after you already want interval width or effect-size framing.
FAQ
What does this page calculate?
It runs an exact binomial test for one observed proportion and reports two-sided, greater, and less p-values against the null proportion you enter.
When should I use a binomial test?
Use a binomial test when you have one success count out of one trial count and want an exact p-value for a null proportion, especially when the sample is small.
Does the share URL include my counts?
No. The share URL stores only light settings such as the selected alternative and decimal places. Successes and trials stay in your browser.
How is this different from a confidence interval page?
This page answers whether the observed count is unusual under one null proportion. Confidence-interval pages answer how wide a plausible range is after you already have the sample proportion.