The Hiring Test Even Its Own Maker Says to Skip

Ask a psychology teacher what students most want to know about the working world, and one question comes up again and again: why do so many employers make job candidates take a personality test before they've even sat for an interview? Mrs. Julia, who teaches AP Psychology and Sociology at the International Community School of Bangkok, fielded exactly that question from her students on a recent episode of the school's podcast, Students Incorporated. Her answer landed on one of hiring's stranger open secrets: the personality test used at the vast majority of Fortune 500 companies is one that the organization which owns it says was never built for hiring at all.

A Test Built for Something Else

The test in question is the Myers-Briggs Type Indicator, the four-letter framework (INTJ, ESFP, and so on) that has become shorthand for personality in offices, dating profiles, and group chats alike. Its reach in the corporate world is hard to overstate: the assessment has been used at roughly 80 percent of Fortune 500 companies and 89 percent of Fortune 100 companies, according to a review by employment-assessment firm Criteria Corp, and more than 3.5 million people take it every year across 115 countries. Consulting giants like McKinsey have used it in their hiring pipeline for years.

Mrs. Julia's point to her students wasn't that personality testing is worthless. It was narrower and more pointed than that. "If you look up the studies on the company that actually made the Myers-Briggs, they themselves would say that test isn't so great for actual hiring," she told her class, "but it'd be great for like team building and being able to assess which team members would work well together." That distinction tracks with what the Myers & Briggs Foundation itself has said publicly: using the MBTI to make hiring decisions isn't an appropriate application of the tool. The U.S. Equal Employment Opportunity Commission adds a legal edge to that caution, requiring that any assessment used in hiring be job-related and properly validated for the role, a bar critics say the MBTI was never designed to clear.

The company that actually made the Myers-Briggs, they themselves would say that test isn't so great for actual hiring.

What the Research Actually Says

None of this means personality has nothing to do with job performance. It means the wrong test is being asked to answer the question. Mrs. Julia pointed her students toward alternatives psychologists trust more: the Big Five trait model, often remembered by the acronym OCEAN, and the older California Psychological Inventory, both of which she described as more statistically reliable than the Myers-Briggs or the similarly popular Enneagram. Research backs that ranking up. A widely cited meta-analysis by industrial psychologists Frank Schmidt and John Hunter, which compared roughly 19 different selection methods using decades of hiring-outcomes data, found conscientiousness, the Big Five trait tied to organization, diligence, and follow-through, to be among the more consistent personality-based predictors of job performance the field has measured. A 2022 Society for Human Resource Management report found that 57 percent of employers already use some form of personality assessment in hiring, meaning the practice isn't going away. The more useful conversation, Mrs. Julia suggested, is which instrument is doing the measuring and whether the people administering it are trained to interpret it responsibly.

The Same Problem, New Software

That tension between the appearance of scientific objectivity and what an assessment can actually justify hasn't gone away with paper-and-pencil tests. It has followed hiring into its newest tools. Algorithmic and AI-driven hiring assessments, now used by a growing share of large employers to screen resumes and score video interviews, are facing the same scrutiny the MBTI has weathered for decades, except with regulators now paying closer attention. In March 2025, the ACLU of Colorado filed a complaint with the EEOC and the state's civil rights division alleging that an AI interview tool used by Intuit and built by vendor HireVue was inaccessible to deaf applicants and likely to score non-white applicants and speakers of dialects like Native American English less favorably. A separate case, Mobley v. Workday, cleared a major hurdle in early 2026 when a federal judge allowed age-discrimination claims against the software company's AI-based applicant-screening tools to proceed, reinforcing that anti-discrimination law applies to algorithms just as it applies to a human recruiter. California's Automated Decision Systems regulations, which took effect in October 2025, went further, formally bringing AI hiring tools under the state's employment discrimination law. The EEOC has been explicit that intent doesn't matter here: an employer can be liable if an AI tool produces a selection rate for a protected group substantially lower than for others, regardless of whether anyone meant for that to happen.

The throughline connecting a 1940s-era personality inventory to a 2026 hiring algorithm is the same: a tool marketed as removing human bias from a decision can just as easily encode it, and neither the maker's marketing nor the appearance of data can substitute for actually checking whether the tool measures what it claims to.

Awareness Isn't the Same as Fairness

That throughline is exactly where Mrs. Julia's classroom conversation went next, moving from hiring instruments to the underlying psychology of bias itself. She broke the concept into three working parts for her students: bias as a tendency to believe some people or ideas are better than others, prejudice as the emotional version of that bias, and discrimination as the behavioral version, someone acting on the bias they hold, consciously or not. Stereotyping, she added, is the cognitive shortcut that often feeds all three. Her sociology class had taken an implicit-association test on skin tone the same day, an exercise designed to surface unconscious bias that, she told her students, "really all of us have," regardless of the culture someone grew up in.

She connected that directly to representation, describing a student-led "skin tone matters" campaign at ICS Bangkok that pushed back against narrow beauty standards common in Asian marketing, and tied it to the psychological concept of the mere-exposure effect: seeing a wider range of skin tones, body types, and ethnicities represented as beautiful in media gradually shifts what audiences perceive as normal. That is not a fringe observation. Colorism, the preference for lighter skin within the same racial or ethnic group, remains a well-documented feature of beauty marketing and hiring perceptions across much of Asia. The global skin-lightening products market was valued at roughly $17.9 billion in 2025, with the Asia-Pacific region accounting for more than half of that demand, according to market-research firm Straits Research. That scale is part of why a school campaign about skin tone isn't a small or purely personal topic; it's pushing against an industry, and a set of assumptions about beauty, that Mrs. Julia's students will keep running into long after they graduate.

Asked what advice she'd give graduating seniors heading into workplaces and campuses far more diverse than the one they were leaving, Mrs. Julia didn't promise the problem would disappear. She told her students that ignorance doesn't always come from malice, and that the more useful move, when they encounter it, is to engage rather than react: to be, as she put it, "a point of contact and point of reference and a point of relationship" for someone who may never have had one before. It's advice that doubles as the missing piece in the hiring-test story: a personality inventory or an algorithm can flag a candidate, but it can't decide, on its own, whether the people reading its output are actually paying attention to what it does and doesn't know.

None of that argues for abandoning assessments altogether. It argues for the same discipline Mrs. Julia was modeling in her own classroom: ask what a tool was actually built to measure before trusting what it tells you.

← Back to all posts