In [[null hypothesis]] significance testing, the p-value is the probability of obtaining test results at least as extreme as the result actually observed, under the assumption that the null [[hypothesis]] is correct.