Predictive validity is the extent to which performance on a test is related to later performance that the test was designed to predict. For example, the SAT test is taken by high school students to predict their future performance in college (namely, their college GPA). If students who scored high on the SAT tend to have high GPAs in college, then we can say that the SAT has good predictive validity. But if there is no significant relation between SAT sores and college GPA, then we would say the SAT has low or poor predictive validity, because it did not predict what it was supposed to.