🧮  P-Value (probability value)

P-Value (probability value)#

The probability of data at least as extreme as observed, assuming the null is true.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

The p-value is the probability of observing data at least as extreme as what you saw, assuming the null hypothesis \(H_0\) is true. It is a measure of evidence against \(H_0\): a low p-value means the data would be surprising under \(H_0\) (stronger evidence against it), a high one means the data sit comfortably with \(H_0\).

Reading it#

By convention \(p < 0.01\) is very strong evidence, \(p < 0.05\) is “significant”, \(p < 0.10\) is weak, and \(p > 0.10\) is not significant. The decision rule pairs it with a threshold: if \(p \le \alpha\), reject \(H_0\); if \(p > \alpha\), fail to reject.

How it’s computed#

Form the test statistic (z, t, \(\chi^2\)…) and find the tail probability of a value that extreme under its distribution — one tail for a one-sided test, both for two-sided. For a two-tailed t-test,

\[p = 2 \times P\!\left(T \ge |t_{\text{observed}}|\right).\]

For example, 17 heads in 20 tosses gives a binomial p of about 0.003 against a fair coin (reject); a two-sample test with \(t = 2.1, df = 28\) gives \(p \approx 0.045\) (reject at 0.05).

What it is not#

Three persistent misconceptions: the p-value is not the probability that \(H_0\) is true; it is not an effect size (a tiny p means surprising, not large); and it is sample-size sensitive — huge samples make trivial effects significant, tiny samples can miss real ones. Always read it alongside an effect size and a confidence interval.


Theme: Statistical Inference & Power  ·  All terminology



See also

Source article Adapted (context, re-expressed) in our own words from: P-Value (probability value) (insightful-data-lab.com).

Tags: purpose: reference topic: terminology level: beginner