What is validity in hiring tests?
Validity is the evidence that a test score supports the decision you use it for, such as who to interview. See the main types and how to check that a test fits.
The short answer
Validity is the evidence that test results support a particular use, such as deciding who to interview for a specific job. It is not a fixed label a test carries everywhere: a reasoning test can suit choosing trainee analysts and be weak for choosing warehouse staff. Employers judge validity by asking whether the test measures what the job really requires.
Common types of validity evidence
Content validity asks whether the test covers tasks and skills that are a real part of the job, and it is the evidence most small employers can check themselves. Criterion-related validity asks whether scores relate to later job performance, which needs data on people hired. Construct validity asks whether the test measures the underlying ability it claims to, such as numerical reasoning. The US Uniform Guidelines on Employee Selection Procedures recognise all three kinds of evidence for a procedure that has adverse impact.
A hypothetical example
Hypothetical example: a logistics company uses a long written English test for warehouse pickers. Many otherwise suitable applicants fail it, yet the job involves reading short labels and a handheld scanner screen. The test is consistent, but it is weak evidence for this job. The company replaces it with a short attention to detail test and a practical picking task, both of which mirror the actual work.
How employers check validity
Start with a job analysis: list the tasks and the skills each one needs. For every test, write one sentence that links it to a task. Ask providers what validity evidence exists, and for which roles and groups. After a few rounds, compare test scores with how hires actually perform, even informally. Keep these notes, because they explain why each step is used if a decision is ever questioned.
Common mistakes
- Calling a test valid in general. Validity belongs to a use, not to the test alone.
- Confusing it with reliability. A consistent score can still be irrelevant; see reliability.
- Skipping the job analysis. Without a list of real tasks, you cannot show a test is job-related.
Related terms
See also selection rate, cognitive ability test and work sample, or browse the full hiring glossary.
Sources
- Uniform Guidelines, 29 CFR 1607.4 (Cornell LII text), checked 2026-10-02
AI can polish an answer.
Hire for the ability behind it.
Explore advanced hiring assessments and integrity tools built for the AI era.