P-value from z-score calculator

This p-value from z-score calculator converts a finite z statistic into the probability associated with the selected tail of the standard normal distribution. Choose a left tail for unusually low outcomes, a right tail for unusually high outcomes, or a two-sided tail when departures in either direction matter. The result is deterministic and requires no lookup table, network request, or simulation, making it suitable for quick checks as well as repeatable automated analysis.

Choose the tail before interpreting the result

A p-value is not determined by the z statistic alone; it also depends on the alternative hypothesis. Select left when increasingly negative values provide evidence against the null hypothesis, such as testing whether a process mean has fallen below a benchmark. Select right when increasingly positive values are the relevant evidence, such as testing whether a conversion rate has improved.

Select two-sided when either a positive or negative departure of the same magnitude would count as evidence. That final choice doubles the probability beyond the absolute observed z magnitude, with symmetry supplied by the standard normal distribution. Make the choice from the study question, ideally before looking at the data.

Switching from two-sided to one-sided merely because the observed direction is convenient changes the test after the fact and can make the reported significance misleading. Enter z as a signed number: the sign is essential for left- and right-tail calculations, while a two-sided result depends on its absolute magnitude.

Understand what the calculator computes

The calculation uses the cumulative distribution function of a standard normal random variable, which has mean zero and standard deviation one. For a left-tailed test, the returned p-value is the probability of observing a value less than or equal to the supplied z statistic. For a right-tailed test, it is the complementary probability above that statistic.

For a two-sided test, it is twice the smaller of those two tail probabilities, capped within the valid probability interval from zero to one. The implementation evaluates a deterministic complementary-error-function approximation, so identical inputs always produce identical numeric output and no external statistical service is involved. Very large magnitudes may round to zero or one at ordinary JavaScript floating-point precision; that reflects numeric representation at extreme tails rather than an assertion that an event is logically impossible.

For routine hypothesis tests, the output provides substantially more precision than is normally shown in a report. You can retain the full value for computation and round only when presenting the result.

Report the p-value with the test context

Use the result alongside the test definition, significance threshold, assumptions, and effect estimate. A small p-value says that the observed statistic, or something more extreme in the chosen tail, would be unusual under the null model. It does not give the probability that the null hypothesis is true, measure the practical importance of an effect, or prove that a research claim is correct.

Before using this calculator, confirm that a standard normal reference distribution is justified. That may follow from a known population variance, a large-sample approximation, or a statistic explicitly defined to have a standard normal null distribution. Smaller samples with an estimated standard deviation may require a t distribution instead.

In a report, state the z statistic, whether the test was left-tailed, right-tailed, or two-sided, and the resulting p-value. Compare it with a threshold chosen in advance, while also reporting confidence intervals or effect sizes when available.