The conventional threshold for a p-value to indicate statistical significance is 0.05. Under this convention, researchers commonly describe a result as statistically significant when its p-value is less than or equal to 0.05, assuming the study selected that significance level in advance.
The threshold corresponds to a 5% significance level, usually written as alpha, or α = 0.05. It means that if the null hypothesis were true and the test’s assumptions held, results at least as extreme would occur no more than about 5% of the time in the long run. Ronald Fisher helped popularize 0.05 in his 1925 book Statistical Methods for Research Workers.
A p-value is not the probability that the null hypothesis is true, nor does it measure the size or practical importance of an effect. The 0.05 boundary is a convention, not a universal law. Some fields use stricter thresholds such as 0.01 or 0.001, while researchers increasingly emphasize effect sizes, confidence intervals, study design, and replication alongside p-values.