Short Definition
Statistical significance means an observed result passes a predefined test threshold under a statistical model.
Intuition
It is a signal that the result is unlikely under the null assumption, not proof that the result is important.
Technical Definition
A result is statistically significant when its test statistic or p-value meets the chosen significance level, such as alpha = 0.05.
Example
An A/B test may call a lift statistically significant if its p-value is below 0.05.
Common Misunderstandings
Significance does not measure effect size.
Repeated testing without correction can create false positives.