type 1 error

Definition

Statistical probability in hypothesis testing that the test sample supports a conclusion that a value is misstated when, in fact, the value is correctly stated. Defined also as an incorrect decision to reject something that should have been accepted, it is the mirror image of type 2 error. If the level of significance (called alpha level) for a test result is set too high, the possibility of type 1 error is reduced but the possibility of type 2 error is raised by the same factor, and vice versa. Also called alpha error or alpha risk. Written also as type I error.