**Definitions:** There are multiple perspectives on the p-value concept.
1. Assuming that $H_0$ is true, how unusual is the data we have observed. The p-value is the probability of observing the current value or something more extreme.
2. For of a particular [[Statistical Test]] $\psi$, the p-value is the smallest significance level $\alpha$ at which $\psi$ would reject $H_0$. This means that if the p-value is less than our chosen significance level, we reject $H_0$.
**Evidence scale:**
| p-value | Evidence against $H_0$ |
| --------- | ---------------------- |
|
gt;10\%$ | almost none |
| $5-10\%$ | weak |
| $1-5\%$ | strong |
| $0.1-1\%$ | very strong |
| lt;0.1\%$ | undisputable |