Choose the true mean, run test after test, and see how often a hypothesis test makes a Type I error or a Type II error.