Face validity is whether a test appears, on the surface, to measure what it claims to measure — regardless of whether it actually does.
It's the weakest and least technical form of validity, because it's based on appearance rather than evidence. A horoscope has high face validity to a believer: the description feels personal and on-target. But feeling accurate and being accurate are different claims, and face validity only speaks to the first one. This is closely related to why the Barnum effect works so well — vague, flattering statements can feel face-valid to almost everyone.
Face validity isn't worthless, though. If a job applicant is taking a test that looks completely unrelated to the job, they may disengage or feel the process is unfair, even if the test is statistically sound. So face validity matters for user experience and buy-in, just not as evidence of accuracy.
The gap between face validity and real validity is one of the more persistent traps in the personality-test industry: tests that feel insightful sell well, whether or not they're backed by actual evidence.