Validated is meaningless without three qualifiers: validated to predict what, for whom, and how well.
Why it matters when the plan changes
Buyers use the word as a filter and suppliers use it as a badge, which means it filters nothing. Predictive validity, properly defined, is the extent to which a score predicts a criterion measure; the criterion, the population and the strength of that relationship are the entire claim. An instrument validated for graduate selection in one country is described in the same word as one validated for executive appointment in another, and an instrument with no evidence at all is often described the same way again.
The tension is timing. Validation for a specific use requires outcomes, which requires time and deployment, so a new instrument cannot honestly claim it however well it is built. The Uniform Guidelines on Employee Selection Procedures expect validity evidence for the specific use rather than a general claim, the same standard the word should meet. The honest position in that interval is to describe the construction, the mechanism and the validation plan, and decline the word until the evidence exists.
In practice
A procurement shortlist asks each supplier whether their instrument is validated. All three say yes. One means a peer-reviewed study on the intended use, one means an internal concurrent study on a different population, and one means the items were reviewed by a psychologist. The question did not distinguish them, and no follow-up was asked.
Evidence
Predictive validity is the extent to which a score predicts a criterion measure, which is what a validation claim has to specify.
Predictive validity, Wikipedia (2026)Selection procedures affecting employment decisions are expected to rest on validity evidence for the specific use rather than a general claim.
Uniform Guidelines on Employee Selection Procedures, 29 CFR Part 1607 (1978)
What it cannot tell you
The word validated does not say what was validated, for whom, or how strongly. It cannot tell you whether evidence covers the population and use in front of you, or only a different one. A model can be validated in one narrow sense and entirely untested in the sense that matters for the decision being made.
Questions
A real claim states the criterion, the population and the strength of the relationship. The field defines predictive validity as the extent to which a score predicts a criterion measure (Wikipedia, 2026). Each element can be checked; a claim missing any of them cannot be evaluated, and a claim missing all of them is marketing.
No. There is no licensing authority for assessment instruments in most jurisdictions. The Uniform Guidelines on Employee Selection Procedures, issued in 1978, set an expectation that selection procedures rest on validity evidence for the specific use rather than a general claim. Some professional bodies operate review schemes, which are useful signals rather than approvals.
Construct and reliability work can be done in months. Predictive validity requires waiting for outcomes to occur and be recorded, which takes as long as the outcomes take. That is why a new instrument cannot honestly claim predictive validity regardless of how carefully it was built.
What was built, on what theory, with which reliability evidence so far, what the validation plan is and when results are expected. That is an honest and informative answer. Using the word validated to cover a plan is the claim that later becomes a problem.
No. Evidence gathered on one population, in one country, for one purpose does not establish validity for another. This is the most common overreach in the field, because the transfer is invisible to a buyer who was only shown the headline figure.