Power of a Statistical Test

The power of a test is the probability of rejecting $H_0$ when the alternative is true:

$$\text{Power} = 1 - \beta.$$

Power depends on the effect size, the variability of the data, the sample size, and the Type I error level $\alpha$. In practice, higher power means a better chance of detecting scientifically meaningful effects.

---

Related

• STA305
• Sample Size Determination
• Power of the One-Sample z-Test
• Power of the One-Sample t-Test
• Power of the Two-Sample t-Test