For employers

Pre-employment test validation: what does an employer have to prove to show that a hiring test is job related?

Validating a pre-employment test means collecting evidence that scores on it predict or represent performance on the job. The Uniform Guidelines on Employee Selection Procedures, 29 CFR part 1607, accept three strategies: criterion-related, content and construct validity. The Guidelines call for that evidence when a test has adverse impact on a race, sex or ethnic group, and each strategy starts with information about the job.

Hands assemble a plain metal bracket while a clipboard lies face down

What does it mean to validate a pre-employment test?

To prove with data that the test measures what the job requires. Under 29 CFR 1607.5 the data come from a criterion-related, content or construct validity study that meets the technical standards of section 1607.14.

Section 1607.5(B) sets one standard for each strategy. In a criterion-related study the procedure should be "predictive of or significantly correlated with important elements of job performance." In a content study the content of the procedure should represent important aspects of performance on the job. In a construct study the procedure should measure identifiable characteristics that are important for successful job performance.

The Questions and Answers that the federal agencies published in the Federal Register on March 2, 1979 explain the choice in answer 54. A user should pick a strategy that is "(1) appropriate for the type of selection procedure, the job, and the employment situation, and (2) technically and administratively feasible." The same answer states: "Whatever method of validation is used, the basic logic is one of prediction." Performance on the test is presumed, on the average, to indicate performance on the job after selection.

Section 1607.5(C) ties the Guidelines to generally accepted professional standards, among them the 1974 Standards for Educational and Psychological Tests of the American Psychological Association. Section 1607.5(E) asks that the study and its report be adequate and accurate and that the test be administered and scored under standardized conditions.

Does every pre-employment test have to be validated?

No. In answer 9 the agencies say the Guidelines require evidence of validity only when the procedure adversely affects the opportunities of a race, sex or ethnic group for hire, promotion or another employment decision.

Question 9 asks: "Do the Guidelines require that only validated selection procedures be used?" The answer begins "No" and continues: validation is desirable in personnel management, but the Guidelines require users "to produce evidence of validity only when the selection procedure adversely affects the opportunities" of a race, sex, or ethnic group for hire, transfer, promotion, retention or other employment decision.

Where there is adverse impact, section 1607.3(A) applies. The use of such a procedure "will be considered to be discriminatory and inconsistent with these guidelines" unless "the procedure has been validated in accordance with these guidelines, or the provisions of section 6 below are satisfied." Under 1607.3(B), when two or more procedures serve the employer's legitimate interest and are substantially equally valid, the user should choose the one with the lesser adverse impact.

The answers also limit the duty. Under answer 13, when the whole selection process has no adverse impact, its components generally need not be examined or validated, with exceptions in answer 25. Under answer 21, when the numbers are too small to be meaningful, the agencies generally do not require validity evidence without other evidence.

The Supreme Court addressed validation studies in a Title VII case in 1988. In Watson v. Fort Worth Bank & Trust, decided June 29, 1988, Justice O'Connor wrote for a plurality of four: "Our cases make it clear that employers are not required, even when defending standardized or objective tests, to introduce formal 'validation studies' showing that particular criteria predict actual on-the-job performance." Watson concerned proof in a lawsuit. The Guidelines and the 1979 answers describe what the enforcement agencies expect.

Why does every validation study start with a job analysis?

Section 1607.14(A) says validity studies should be based on review of information about the job. The job analysis names the work behaviors that the test items or the performance measures have to match.

Section 1607.14 repeats the requirement for each strategy. For a criterion study, "There should be a review of job information" to find the measures of work performance that matter. For a content study, "There should be a job analysis" of the important work behaviors and their relative importance. For a construct study, "Each construct should be named and defined" so that it can be told apart from other constructs.

Section 1607.9(B) encourages professional supervision of selection work but adds that it "is not a substitute for documented evidence of validity." The agencies consider whether a thorough job analysis was done and whether the procedure was developed and used under professional standards.

For a written skills test the job analysis is a document. It lists the duties of the position, the knowledge and skill each duty uses and how often and how much each duty counts. Every test item is then linked to a line of that list, which answer 79 requires for job knowledge tests on an item by item basis.

When does content validity fit a skills test or a work sample?

When the test samples real work behaviors or a defined body of job knowledge. Section 1607.14(C)(1) rules content validity out for traits such as intelligence, aptitude, personality, commonsense, judgment, leadership and spatial ability.

The list in 1607.14(C)(1) reads "intelligence, aptitude, personality, commonsense, judgment, leadership, and spatial ability." For a test of knowledge, skill or ability, the attribute measured has to be a "necessary prerequisite to successful job performance." Paragraph (C)(4) gives the scale: "The closer the content and the context of the selection procedure are to work samples or work behaviors," the stronger the basis for content validity.

Answer 79 sets the test for a job knowledge test. "There must be a defined, well recognized body of information, and knowledge of the information must be prerequisite to performance of the required work behaviors." The next sentence: "The work behavior(s) to which each knowledge is related should be identified on an item by item basis."

Section 1607.5(F) adds a limit. Employers should generally avoid decisions based on knowledge, skills or abilities that are normally learned in a brief orientation period, where such measures have adverse impact.

What went wrong with the validation study in Albemarle Paper Co. v. Moody?

The study was built months before trial, used supervisors' subjective rankings and tested experienced workers near the top of the lines of progression, only four of them Black. On June 25, 1975 the Supreme Court found it defective.

In Albemarle Paper Co. v. Moody, 422 U.S. 405, the company hired an industrial psychologist about four months before the 1971 trial. He spent half a day at the plant and designed a concurrent validation study, which plant officials carried out without his supervision. The study covered the Beta Examination and Forms A and B of the Wonderlic Personnel Test.

The sample came from ten job groupings near the top of nine lines of progression: 105 employees, four of them Black. Two supervisors ranked each employee against coworkers in the same grouping and were asked which did a better job "irrespective of the job that they were actually doing." The Beta Examination correlated significantly in three groupings, one Wonderlic form in seven and the combined battery in two. The District Court found the tests job related.

The Supreme Court disagreed on four grounds. The results were patchy and did not validate the battery for every skilled line where it was used. The criterion was "subjective supervisorial rankings," and nobody could say what job performance standards the supervisors had applied. A sample of workers near the top of a line of progression did not show whether the test measures the qualifications of new entrants to lower jobs. The tests were given to new applicants, younger, less experienced and often nonwhite, while the sample was experienced and almost all white.

The Court measured the study against the EEOC guidelines then in force and called them "entitled to great deference." The tests had to be "predictive of or significantly correlated with important elements of work behavior." The Court vacated the judgment of the Court of Appeals for the Fourth Circuit and remanded the case to the District Court.

Can an employer rely on the test vendor's validation study?

Only within section 1607.7: the study has to meet the Guidelines, job analyses must show that the jobs involve substantially the same major work behaviors, and a test fairness study is expected. Promotional literature does not count.

Under section 1607.7(A) users may rely on validity studies by other users, test publishers or distributors "under certain circumstances," and they remain responsible for compliance. Paragraph (B) sets three conditions for criterion-related evidence from another source. The studies have to show clearly that the procedure is valid. Incumbents in the user's job and in the studied job must "perform substantially the same major work behaviors," as shown by job analyses of both. The studies should include "a study of test fairness for each race, sex, and ethnic group" that is a significant factor in the user's labor market. Where a study has no such investigation and the user cannot technically run its own, paragraph (B)(3) lets the user rely on the study until studies elsewhere show unfairness or an internal study becomes feasible. If other variables are likely to affect validity significantly, paragraph (D) says the user "may not rely upon such studies."

Section 1607.9(A) names what cannot replace evidence of validity: the "general reputation of a test" and "casual reports of its validity." It specifically rules out "assumptions of validity based on a procedure's name or descriptive labels," "all forms of promotional literature," "data bearing on the frequency of a procedure's usage," "testimonial statements and credentials of sellers, users, or consultants" and "other nonempirical or anecdotal accounts of selection practices or selection outcomes."

Answer 35 asks whether validity reports in commercial test manuals can help. "They may be. However, it is the user's responsibility to determine that the validity evidence is adequate to meet the Guidelines." Answer 43 explains why a vendor's study may not transfer: "a selection procedure found to have validity in one situation does not necessarily have validity in different circumstances."

How should an employer set a cutoff score, and may it rank candidates?

Under section 1607.5(H) a cutoff should normally be reasonable and consistent with normal expectations of acceptable proficiency in the work force. Ranking needs its own evidence, because evidence for pass/fail use may not support it.

Section 1607.5(G) deals with the method of use. If one method has greater adverse impact than another, the validity and utility evidence must support the method actually chosen. "Evidence which may be sufficient to support the use of a selection procedure on a pass/fail (screening) basis" may be insufficient to support ranking.

For content-valid tests, 1607.14(C)(9) permits ranking only where a job analysis or other evidence indicates that "a higher score on a content valid selection procedure is likely to result in better job performance." Under 1607.5(H) a cutoff above the level of acceptable proficiency may be appropriate where candidates are ranked on a validated procedure, and the adverse impact of that cutoff has to be considered.

Two more paragraphs of 1607.5 apply to hiring tests. Paragraph (I) lets an employer treat candidates as evaluated for a higher job only where most employees progress to it within a reasonable time, seldom more than 5 years. Paragraph (J) permits interim use of a procedure while a study is under way, if the user has substantial evidence of validity; if the study fails to support validity, the interim use is no defense.

What records should an employer keep for a test it has validated?

Where the procedure has adverse impact, the documentation described in section 1607.15. The agencies' answers add applicant records at every stage of the process and a validity report complete enough for an enforcement agency to evaluate.

Section 1607.5(D) requires the documentation of section 1607.15 where a procedure is part of a selection process with adverse impact and the procedure itself has adverse impact. Answer 15 adds: "Records should be kept for persons who were applicants or candidates at any stage of the process."

On the study itself, answer 89 says that reports of validity studies "should contain all the information necessary to permit an enforcement agency" to evaluate the study. Answer 90 covers the source data behind it: "The Guidelines require the maintenance of these records in some form." Under answer 34, an agency will not review validity evidence "without evidence of how the selection procedure is used."

Small users get simpler rules. Answer 82 asks whether the Guidelines simplify recordkeeping for small users, and the answer is "Yes."

How does the IAE written exam fit into an employer's validation file?

IAE publishes a program with sections and weights for each exam field and reports rubric scores for each task. Whether the exam is valid for a given job is the user's question under answer 35 and section 1607.7.

The exam is one sitting of up to 2 hours: 5 questions of 10 minutes, 2 cases of 25 minutes and a 10-minute question about the candidate's own project. It cannot be paused. Pasting is disabled, the system logs how each answer was typed, each candidate gets different figures in the cases, and a doubtful answer brings one more written question with 5 minutes to answer. Before the exam the candidate uploads a photo of an ID document and takes a snapshot, and the Secretariat checks them manually.

Answers are scored against the rubric of the field, and an examiner with no ties to the company or the candidate approves the result within 5 business days. The company sees the status, the total score out of 100, the score on each rubric criterion (correctness, reasoning, risks and clarity) in each task, and the certificate PDF if the exam is passed. It does not see the questionnaire, the CV, the answers or the reason for refusal.

Each field has an open program with sections and weights, version 1.0. An employer that plans to use the exam in hiring can compare the published sections and sample tasks with its own job analysis and keep its own applicant records by group. The exam uses standard IAE tasks, and the candidate does no work for the company. At registration a company can also mark that it needs tasks for its own roles and list those roles. The company decides on the interview and on hiring.

Companies register on the employer page and receive a link to the company cabinet within one business day. Companies that register by December 31, 2026 test up to 5 people free in the pilot. Outside the pilot the company pays for each candidate on an invoice, with no subscription.

Sources

Read also

All articles

Pilot until December 31, 2026

Register your company

Free for up to 5 people. A link to the company cabinet arrives by email within one business day.

Company registration

Received

Cabinet link within one business day

Employees5
FieldSoftware and data
Pilotuntil Dec 31, 2026