The question we get most often from cautious HR leaders sounds something like: "If we deploy a custom pre-hire test, are we more or less exposed legally than if we just keep using the standardized battery we bought from a vendor?"
The answer is consistently "more defensible — if you build it right." But "if you build it right" is doing all the work in that sentence. To understand why, you need to understand the actual legal framework around hiring tests, and almost no HR team does. So let's walk through it.
The framework that matters: the Uniform Guidelines
In 1978, four federal agencies — the EEOC, the Department of Labor, the Department of Justice, and the Civil Service Commission — jointly published the Uniform Guidelines on Employee Selection Procedures. It's federal regulation, codified at 29 CFR Part 1607. Forty-seven years later, it's still the document every court reaches for when an employer's hiring test is challenged.
The Uniform Guidelines apply to any selection procedure used to make employment decisions — hires, promotions, terminations, transfers. A "selection procedure" includes formal tests, interviews, reference checks, application forms, performance evaluations, and informal scoring systems. Anything that affects who gets the job.
The Guidelines don't ban any particular method. They impose one core requirement: if your selection procedure causes adverse impact on a protected group, you have to be able to prove it's job-related and consistent with business necessity. No proof, no defense.
Two terms there matter — "adverse impact" and "job-related." Let's take them in order.
Adverse impact and the four-fifths rule
Adverse impact means a selection procedure disadvantages a protected group at a meaningfully higher rate than the majority group. Race, color, religion, sex (including pregnancy and sexual orientation), national origin, age (40+), and disability are the main federally protected categories. Many states add more.
The Uniform Guidelines give a practical rule of thumb for when adverse impact is presumed: the four-fifths rule (sometimes called the 80% rule).
If the selection rate for any protected group is less than four-fifths (80%) of the selection rate for the group with the highest selection rate, the Guidelines consider that "adverse impact" exists — and the burden shifts to the employer to prove the procedure is job-related.
Here's how the math actually works:
A company hires for a care coordinator role. Their pre-hire test runs on a recent applicant cohort:
60% is below the 80% threshold. The four-fifths rule presumes adverse impact. The employer must now prove the test is job-related, or stop using it.
Two important nuances that get missed:
The four-fifths rule isn't a safe harbor. Passing 80% doesn't mean you're definitely fine. Courts have also accepted statistical-significance tests (a "two standard deviation" rule) as evidence of adverse impact even when the four-fifths threshold is technically met. The four-fifths rule is a screening signal, not a legal guarantee.
The burden of proof shifts. Once adverse impact is shown, the employer has to demonstrate the test is "job-related and consistent with business necessity." If the employer succeeds, the burden flips back: the plaintiff can still win by showing a less-discriminatory alternative exists that the employer refused to use.
If your test causes adverse impact and you can't prove it's job-related, you don't get to keep using it. Simple as that.
"Job-related": the three forms of validity
"Job-related and consistent with business necessity" is the legal phrase. The Uniform Guidelines translate it into a technical requirement: you have to validate the procedure using one of three accepted methods.
For most hiring tests, the two practical paths to defensibility are content validity (the test is clearly drawn from observed job tasks) or criterion validity (the test scores statistically predict actual job performance).
Both require something most companies don't actually have: a real job analysis.
What "job analysis" really means (and why it's the hard part)
A job analysis is a structured documentation of what a job actually requires — the tasks performed, the knowledge needed, the skills used, the conditions of work, the judgment calls made. Not a job description. A job description is what you write to attract candidates. A job analysis is what you write to prove what the job actually demands.
The Uniform Guidelines require that any validated test — content, criterion, or construct — be tied to a job analysis. If you skip this step, your validation argument has nothing underneath it. A court will not accept "we believe this test measures the right things." They want documentation that the test was built from documented job requirements.
This is the single biggest reason most hiring tests aren't actually defensible: the company can't produce a job analysis to anchor the test to. They have a job description (which is for marketing the role), they have HR's gut sense of what matters, and they have a vendor's product brochure. None of those are job analyses.
A proper job analysis comes from observing the work — interviewing high performers, watching shifts, reviewing artifacts of the work product, documenting the actual decisions people make. It's labor-intensive. It's also the legal backbone of every defensible hiring test.
The mistake most HR teams make: assuming vendor validation transfers
This is the most expensive misunderstanding in the field, so it deserves a section of its own.
When you buy a standardized pre-hire battery from a major vendor (SHL, Criteria, HireVue, Wonderlic, etc.), the vendor will hand you a validation report. The report will reference studies — often thousands of them — showing the test's predictive validity across many jobs and industries. That report makes the test sound bulletproof.
Here's the catch: the vendor's validation evidence is for the test in general — not for your use of it. The legal question in any challenge is whether the test, as you applied it to your specific jobs, was job-related. The vendor's general validation might support that argument. It might not. The Uniform Guidelines actually address this directly — they require the user to establish that the validation evidence applies to their use of the test, not just to the test in the abstract.
In court, this often comes down to: did you do a job analysis showing your role matches the jobs in the vendor's validation studies? If not, the vendor's validation report doesn't automatically save you. It's evidence — but it's not the full answer.
The cleaner answer is to build the validation argument around your actual operation, not someone else's. That's what HireGauge does, and it's why "custom" can be more defensible than "off-the-shelf" — if it's done right.
A vendor's validation study is evidence. It is not a substitute for showing the test fits your jobs.
Two landmark cases worth knowing
Griggs v. Duke Power Co. (1971)
The case that started the modern framework. Duke Power required employees to have a high school diploma and pass two aptitude tests to qualify for higher-paying jobs. The Supreme Court ruled unanimously that even neutral selection criteria can be illegal if they have disparate impact on a protected group and aren't shown to be job-related.
The phrase the court used became the foundation of every adverse impact case since: "What is required by Congress is the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification."
Albemarle Paper Co. v. Moody (1975)
Four years after Griggs, the Court tightened the standards for what counts as adequate validation. Albemarle had hired an industrial psychologist to do a hasty validation study after they were sued. The Court rejected it, ruling that validation studies must follow accepted professional standards — they can't be reverse-engineered to defend a test that's already failed disparate impact analysis.
The takeaway: validation has to be done up front, properly, by qualified people — not patched together after a problem surfaces.
A practical checklist: is your hiring test actually defensible?
The 7 questions
- Have you measured adverse impact? If you've never run the four-fifths analysis on your test's outcomes by protected group, you have no idea whether you're exposed.
- Do you have a real job analysis — not a job description — that documents what the role actually requires?
- Can you tie every section / item / scoring rule of your test back to that job analysis? If not, the test is floating without a legal anchor.
- If you're relying on vendor validation, have you documented that your jobs match the jobs in the vendor's validation studies? Without that, the vendor's evidence doesn't automatically apply to you.
- Do you have an ongoing validation loop? Following up hires with performance data and updating cutoff scores accordingly. The Uniform Guidelines treat validation as continuing, not one-time.
- Are your scoring decisions, cutoff scores, and section floors documented and reviewed? Arbitrary cutoffs are one of the most common ways a test fails challenge.
- Have you reviewed all of this with an employment attorney? Not as the validator — as the person who'll defend you if it's challenged.
If you answered "no" or "not sure" to three or more of those, your current setup probably wouldn't survive serious scrutiny. That's not a niche risk — disparate impact suits are routine in operational hiring, and the cost of losing one (settlement, attorney fees, mandated changes, reputational damage) dwarfs the cost of doing the validation work up front.
When to get a real attorney involved
To be blunt: from the moment your hiring test affects more than a handful of people, you should have an employment attorney reviewing it. Not optionally. The cost is small relative to the exposure.
Specifically, get counsel involved if any of these are true:
- You're rolling out a new test to a high-volume role for the first time
- You're changing scoring thresholds or pass criteria on an existing test
- You hire 100+ people per year through any single selection procedure
- You're in an industry with elevated EEOC scrutiny (healthcare, financial services, federal contractors)
- Anyone has raised concerns — internally or externally — about adverse impact
What an attorney can't do for you is build the validation. That's a separate discipline. But they can review your validation work, identify your specific risk exposure, and tell you what to document and how. Pair them with a competent assessment designer (whether that's an internal I/O psychologist, a consultant, or HireGauge) — that's the team you actually need.
The bottom line
Off-the-shelf hiring tests aren't inherently more or less defensible than custom ones. What matters is whether the test you're using has been properly validated for your specific use, with a documented job analysis anchoring it. Most companies haven't done that work — for either flavor. That's the actual exposure.
The custom approach, done correctly, is usually more defensible than a generic battery — because the job analysis is real, the items are visibly tied to observed work, the criterion validation runs against your actual outcomes, and the whole chain of evidence is built for your specific roles. A generic battery can get you there too, but only if you do the work to show it applies to your jobs. Almost no one does.