Predictive validity is a form of criterion validity that tests whether a measure's scores forecast an outcome that hasn't happened yet.
A classic example: a company gives a conscientiousness assessment to job applicants during hiring, then checks a year later whether the people who scored higher actually turned in stronger performance reviews. If they did, the test has predictive validity for that outcome. What makes this distinct from concurrent validity is the gap in time — the test comes first, and the real world has to play out before you find out whether the score meant anything.
This is often the more valuable, and more difficult, form of criterion validity to establish. It's one thing for a test to line up with how someone is doing right now; it's a stronger claim to say a test taken today tells you something real about a year or a decade from now. That's exactly why predictive validity is the gold standard for high-stakes assessments like hiring tools, college admissions tests, and clinical risk screenings — the whole point of using them is to know something about the future before it arrives.
Studying predictive validity properly takes patience: researchers have to test people, wait for the outcome to actually happen, and then check the correlation, which is part of why it takes years to build a solid validity case for any serious assessment.