---

Note that the probability of comitting a type I error (a, false positive) never changes. However, as the difference between the null hypothesis and the actual population increases, the power of the test (1-b) also increases; and the probability of committing a type II error (b, false negative) decreases.