In hypothesis testing, the probability of rejecting a false null hypothesis, i.e. of concluding that a treatment is effective when a treatment effect of a given magnitude truly exists, for example.