P-value calculator
Enter a test statistic — a z, t, or chi-square value — and get the exact p-value. Everything runs in your browser; nothing is uploaded.
—
How to use it
Pick the distribution your test produced a statistic from, type the value, and choose one- or two-tailed. The p-value updates as you type.
- z (normal) — large-sample tests, proportion tests, a standardized z-score.
- t (Student) — one- and two-sample t-tests and regression coefficients. You'll also need the degrees of freedom (roughly your sample size minus the number of estimated means).
- χ² (chi-square) — tests of independence and goodness of fit. Degrees of freedom come from the table dimensions; chi-square is read in the upper tail, so the "tails" control doesn't apply.
What "two-tailed" means. A two-tailed p-value counts results as extreme in either direction — the usual default when your hypothesis is just "there's a difference." Use one-tailed only when you committed to a direction before seeing the data. For a symmetric z or t, the one-tailed p is exactly half the two-tailed p.
How to read the result
By the common 0.05 convention, p < 0.05 is called statistically significant. But the p-value is only half the story — it tells you whether an effect is probably real, not how big it is. Read the full picture on the p-value glossary page, and always pair it with an effect size and a confidence interval.
Skip the lookup entirely. In Stratum you don't compute the statistic by hand and look up a p-value — load your data and the t-test, ANOVA, correlation, chi-square, and regression all report exact p-values directly, with the matching chart. Download Stratum →
See also
- Other calculators
- Sample-size calculator · Effect-size calculator
- Learn the concept
- What is a p-value? · Confidence interval
- Do it on real data
- Run a t-test