P-Value Calculator

The p-value calculator converts a Z-score (a standard normal test statistic) into its one-tailed and two-tailed p-values and tells you whether the result is statistically significant at the common 95% and 99% confidence levels. The p-value is the probability of observing a test statistic at least as extreme as the one you got, assuming the null hypothesis is true. A small p-value means your data would be surprising under the null hypothesis, which is treated as evidence against it. This tool is for students taking statistics courses, researchers and analysts running hypothesis tests, and anyone who has computed a Z-score and needs to translate it into a probability they can compare against a significance threshold.

What the p-value actually measures

The null hypothesis (H₀) is the default position of "no effect" or "no difference" — for example, "the new method changes nothing". You collect data, compute a test statistic, and ask: if H₀ really were true, how often would chance alone produce a result this extreme or more so? That probability is the p-value. It is not the probability that the null hypothesis is true, and it is not the probability that your finding happened by chance — it is a statement about the data given the assumption of H₀. The calculation here uses the standard normal distribution: the calculator evaluates the cumulative distribution function Φ at your Z-score, takes the smaller tail as the one-tailed p-value, and doubles it for the two-tailed p-value.

Significance and α = 0.05

Before testing you choose a significance level, α — most commonly α = 0.05, corresponding to 95% confidence. The rule is simple: if the p-value is less than α, you reject the null hypothesis and call the result statistically significant; if it is greater than or equal to α, you fail to reject H₀. This calculator flags significance at 95% (p < 0.05) and at 99% (p < 0.01) automatically. Stricter fields use α = 0.01 or even α = 0.001 (0.1%) to guard against false positives; the threshold should be fixed in advance based on how costly a wrong conclusion would be.

One-tailed vs. two-tailed

One-tailed (one-sided): tests whether the parameter is specifically greater than (or specifically less than) a value, using just one tail of the distribution. Use it only when you have a strong, pre-registered reason to expect the effect in one direction.

Two-tailed (two-sided): tests whether the parameter merely differs from a value in either direction, so it counts both tails — which is why its p-value is twice the one-tailed value. It is more conservative and is the default choice in most research because it does not presume which way the effect will go.

Z-tests and t-tests

A Z-score is the right statistic when the population standard deviation is known or the sample is large. For small samples with an estimated standard deviation, the t-distribution is technically correct: it has heavier tails, so for the same statistic a t-test gives a slightly larger (more cautious) p-value. As the sample size grows beyond roughly n = 30 the t-distribution converges on the normal curve, so for large samples the Z-based p-values this calculator returns are very close to the t-test result.

Worked example

Enter a Z-score of 1.96. The standard normal curve places about 2.5% of its area beyond +1.96 in the upper tail, so the one-tailed p-value is roughly 0.025, and the two-tailed p-value is about 2 × 0.025 = 0.05. That is the famous boundary: a Z of 1.96 sits right at the edge of significance at the 5% level. A larger Z, say 2.58, pushes the two-tailed p-value down to about 0.0099 — significant at both the 95% and 99% levels — while a Z of 0 gives a two-tailed p-value of 1, meaning no evidence against H₀ whatsoever.

Tips and common mistakes

Tips: the sign of the Z-score does not affect the two-tailed p-value, since both tails are symmetric; |z| is what matters. Decide one- vs two-tailed and your α before you look at the data, not after.

Common mistakes: a low p-value is not the same as a large or important effect — with a big enough sample even a trivial difference can be "significant", so always read effect size alongside it. Equally, p ≥ 0.05 does not prove the null hypothesis is true; it only means you lack the evidence to reject it. And switching to a one-tailed test after seeing the data simply to squeeze under 0.05 is a misuse that inflates false positives.