What the Tests Are Actually Measuring

The Myers-Briggs Type Indicator is taken roughly 2 million times per year. Corporations use it for team-building and hiring. Coaching practices are built around it. People list their type in dating profiles and social media bios. The four-letter shorthand — INFJ, ENTP, ISFP — functions as a personality language that has become widespread enough to feel authoritative. Personality psychologists are largely unanimous in their assessment of it: the MBTI is not a valid measure of personality. The consensus among researchers who study personality measurement is unusually clear on this point, and has been for decades. This doesn't stop the instrument from being used everywhere, which itself tells you something interesting about how people select psychological frameworks.

The Psychometric Problems

Valid personality assessments need to satisfy several criteria, the most fundamental being reliability (the test produces consistent results across time and testing conditions) and construct validity (the test measures what it claims to measure in a way that predicts real-world outcomes). The MBTI fails reliability testing in basic ways. Test-retest studies — where the same person takes the test twice, weeks or months apart — find that up to 50 percent of respondents receive a different four-letter type on the second administration. A measure of stable personality traits should not produce different results for the same person half the time. The construct validity problems are structural. The MBTI organizes personality into four dichotomies — introversion/extraversion, sensing/intuiting, thinking/feeling, judging/perceiving — and treats each as a categorical distinction. You are one type or the other. But the underlying traits these dimensions are attempting to measure exist on continuous distributions in the population. Most people score near the middle of each scale, not at the poles. Forcing continuous variation into binary categories discards information and produces the unstable typing that the test-retest data reveals. A review from Yale's psychology faculty examining the MBTI alongside validated instruments found that the four dimensions of the MBTI mapped imperfectly onto established personality dimensions and had significantly weaker predictive validity for occupational and relationship outcomes than the Big Five model, which has robust support across cultures and contexts.

Why It Persists

The persistence of the MBTI despite its psychometric problems is itself a phenomenon worth understanding. Several factors contribute. The test is proprietary, and the organization that owns it — the Myers-Briggs Company — has significant commercial interest in its continued use and actively markets it to corporations. The types are described in entirely positive language — there are no bad types — which makes the test feel affirming rather than evaluative. Barnum effects do significant work: the type descriptions are specific enough to feel accurate while vague enough to fit most people who receive them. There's also the community function. Having a shared typing system creates common vocabulary and the satisfying experience of self-categorization. People who share a four-letter type form communities around it, reinforcing the sense that the type captures something real. The social reality of the community makes the underlying type feel validated even when the measurement instrument isn't. Research from Cornell's human resources management department examining why corporations continue using personality assessments with poor validity found that decision-makers rated assessments as more valid when they produced results that felt intuitive — that matched pre-existing impressions of the person being assessed. The MBTI's descriptions are written to feel intuitively correct regardless of their predictive accuracy.

The Tangent Worth Taking

Astrology is worth naming here directly, because the MBTI comparison is genuinely apt and worth taking seriously rather than using as a dismissal. Both systems offer personality typologies based on a small number of categories. Both are described in affirming language that fits most people who receive the description. Both produce strong community identification among adherents. Both persist despite lack of empirical support for their predictive validity. The primary difference is cultural: astrology is coded as spiritual or folkloric, while the MBTI is coded as psychological and corporate, which makes it feel more rigorous than it is.

What Valid Personality Assessment Looks Like

The Big Five framework — measuring openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism as continuous traits — has been replicated across cultures, languages, and methods. It shows consistent predictive validity for occupational performance, relationship satisfaction, health behaviors, and life outcomes. It's less fun to talk about at parties because it doesn't produce memorable four-letter types. That's roughly the extent of its disadvantage. If you're interested in understanding your own personality in ways that are actually predictive, the Big Five is the place to look.