HR Consulting

blogheader

Personality Assessments for Hiring: When They Earn Their Keep (and When They’re Snake Oil)

I am not an attorney. This article is general information based on my experience as an HR professional, and it is not legal advice. For legal advice about your situation, consult a qualified attorney.

The short version

  • A selection assessment has to be job-related and backed by validation evidence. Most marketed personality tests are neither.
  • A test designed to reveal a mental impairment is a medical examination under the ADA and cannot be given before an offer.
  • California’s automated-decision regulations, effective October 1, 2025, reach personality tests, games and reaction-time challenges.

These rules change every year, and California adds its own. A subscription gives you a California-certified HR professional to check your answer before you act. Call (562) 444-8076, email service@accomplicenter.com, or come by our store at 3217 Carson St in Lakewood to sign up.

A bad hire costs a small business between 30 and 50 percent of the employee’s annual salary, by most credible estimates. For a $60,000 hire, that’s $18,000 to $30,000 lost. For a $100,000 manager, it’s a number that hurts.

The assessment industry knows this and has built itself around the promise of de-risking the decision. Personality tests, cultural fit assessments, “behavioral profiles,” and similar products are sold to employers as ways to predict who will perform and who won’t. Most of these products are junk. A few are validated science. The difference matters, and most operators don’t know how to tell.

What makes an assessment defensible

The Equal Employment Opportunity Commission and federal regulations under the Uniform Guidelines on Employee Selection Procedures don’t prohibit personality assessments in hiring. They require that any selection procedure used for employment decisions be job-related and consistent with business necessity, and that it not have an unjustified disparate impact on protected groups.

Translated for an operator: if you’re going to use an assessment to make a hiring decision, the assessment has to actually predict performance on the job, the prediction has to be backed by validation evidence, and the assessment has to be administered and interpreted by someone qualified to do so.

Most hiring personality tests fail the first part of that test. They were never built to predict job performance. They were built to be entertaining, or to generate a marketing report for the candidate, or to fit a vendor’s pricing model. They produce results, but the results don’t correlate with how the person will actually perform.

The minority of assessments that do predict job performance share a few characteristics:

  • They were built specifically for occupational use. Not adapted from a clinical instrument, not based on a personality theory designed for therapy, not built for self-help. Built from the ground up for hiring decisions or workplace performance prediction.
  • They have published validation studies. The vendor can show you peer-reviewed research demonstrating that the instrument correlates with relevant job-performance criteria. The studies should be in mainstream industrial-organizational psychology journals, not vendor-published white papers.
  • They have published reliability data. Reliability is a measure of how consistent the test is. Bad tests give different results to the same person on different days. Good tests are stable.

They have demonstrated low adverse impact across protected groups, or the validation evidence justifies any impact that exists. The Uniform Guidelines requirements aren’t optional, and the better instruments have built their validation around demonstrating defensibility.

The assessments that meet these criteria are a small set. The CPI 260 is one of them. The Hogan suite is another. The 16PF is another. There are a few others. Most of what gets marketed as “personality assessment for hiring” doesn’t make this list.

When to use one

Assessments aren’t appropriate for every hire. The cost of administration and interpretation, the time it adds to the process, and the candidate experience implications all matter. The math works for some hires and not others.

The hires where assessments earn their keep:

  • Leadership roles. A manager who sets the tone for a team has outsized impact. The CPI 260 in particular was developed and validated against managerial performance, and the data on selection assessments at this level is robust. A miscast manager costs you the team they’re managing, not just their own seat.
  • Key technical hires. A senior engineer, a head of operations, a finance lead at a small company. Roles where the wrong person sets you back twelve months and the right person sets you ahead twelve months.
  • Partner-level decisions. If you’re hiring someone who will own equity, who will speak for the company externally, who will represent the brand, the cost of a wrong hire goes well beyond their salary. An assessment is appropriate.
  • Difficult-to-evaluate roles. Some roles are hard to interview for because the work itself is hard to simulate or describe. Sales is a common example. Customer-facing roles in regulated industries. Specialized professional roles. When the interview alone isn’t giving you a reliable read, an assessment can fill the gap.

The hires where assessments aren’t worth it:

  • Hourly entry-level roles. A dishwasher hire, a cashier hire, a receiving clerk hire. The cost of the assessment exceeds the cost of a wrong hire at this level, and the validation data is thinner for these roles.
  • Short-tenure roles. Seasonal hires, temporary roles, project-specific hires. The assessment cost is amortized over too short a tenure to make sense.
  • Roles where the existing process works. If you have a hiring process that’s giving you reliable results, don’t add steps for the sake of adding steps. Save the assessment for the hires where you’re not getting reliable signal.

How to use one well

If you decide an assessment is appropriate for a hire, a few things matter.

  • The job description has to be done first. Assessments are interpreted against the requirements of the position. Without a clear definition of essential functions and the competencies the role requires, the assessment results don’t connect to anything actionable. The job description grounds the interpretation.
  • The assessment can’t be the whole decision. EEOC guidance is clear that selection decisions should rest on multiple sources of evidence. The assessment is one input. The interview is another. The reference checks are another. The candidate’s track record is another. Treating the assessment as the deciding factor is both legally risky and bad practice.
  • The administration has to be qualified. This is where most operators get into trouble. They buy the assessment, give it to the candidate, get the results, and try to interpret them. The interpretation requires training. The patterns in the data, the relationships between the scales, the boundary cases, the way the results inform interview questions, all of it requires someone who has been trained on the specific instrument and has experience reading the output.
  • The decision has to be consistent. If you use an assessment for the leadership role, you should use it for every leadership role at that level. Inconsistent application is one of the surest ways to invite a disparate-impact challenge. Set a clear policy for which roles get assessed and apply it.

The two legal lines

Two legal lines matter here. The first is the ADA’s bar on pre-offer medical examinations. A personality instrument crosses that line when it was designed to reveal a mental impairment, regardless of how you score it or what you use it for: the Seventh Circuit held exactly that about the MMPI in Karraker v. Rent-A-Center. The CPI 260 sits on the safe side because it is a normal-range occupational instrument rather than a clinical one, but the safe practice is the same for any tool. Use normal-range, job-related instruments, avoid items that probe mental health symptoms, and keep anything clinically derived until after the offer. The second line is new. California’s Civil Rights Department regulations, effective October 1, 2025, treat an automated tool that analyzes physical or mental traits, including personality tests, games, puzzles and reaction-time challenges, as a potential unlawful pre-offer inquiry under FEHA, and require four years of records on the system and its results. Illinois barred AI-driven employment discrimination effective January 1, 2026. The EEOC withdrew its own AI technical assistance in January 2025, so there is no federal guidance to point to. If your assessment is scored or interpreted by an algorithm rather than a trained human, the state rules are what you answer to.

Where I come in

The CPI 260 is the assessment I administer most often for clients. It was developed at the University of California Berkeley, has more than fifty years of validation research behind it, and was built specifically for occupational and managerial assessment. It’s not a quick screen. It’s a 260-item instrument that takes the candidate about 30 minutes and produces an interpretation report covering 26 personality dimensions, organized around four broader themes that map directly to managerial performance.

I’m John Hamilton. I have an MS in HR Management. My undergraduate degree is in psychology with a concentration in neurobiology, from California Polytechnic State University. I hold the PHRca and PHR certifications from HR Certification Institute. I administer the CPI 260 for clients who are about to make a hire that matters more than the average hire, and I write the interpretation in language that’s actionable for the hiring decision rather than clinical.

If you’re hiring someone whose miscast costs you real money, call us at (562) 444-8076 or email service@accomplicenter.com. A manager, a key technical hire, a partner, a leadership-track employee. We’ll talk about the role, the candidate pool, and whether the CPI 260 is the right call. Most of the time it is. When it isn’t, I’ll tell you that and we’ll talk about a different approach.

The cost of the assessment, including administration and interpretation, is a fraction of the cost of a wrong hire. The math has favored doing this for decades.

To subscribe, call us at (562) 444-8076, stop by our store at 3217 Carson St in Lakewood, or email service@accomplicenter.com.

See AccompliCenter HR plans →

Store Hours

Today
  • Tue
    Nov 17

    8:30AM
    —
    6:00PM
  • Wed
    Nov 18

    8:30AM
    —
    6:00PM
  • Thu
    Nov 19

    8:30AM
    —
    6:00PM
  • Fri
    Nov 20

    8:30AM
    —
    6:00PM
  • Sat
    Nov 21

    9:00AM
    —
    4:00PM
  • Sun
    Nov 22

    10:00AM
    —
    3:00PM
  • Mon
    Nov 23

    8:30AM
    —
    6:00PM