MBTI’s Shocking Lack of Reliability Revealed by Decades of Research
I have a complicated relationship with personality typing. Not because the frameworks are useless — some of them are genuinely illuminating — but because of what happens to them once people get hold of them. The type becomes a verdict. The instrument designed to open up self-understanding becomes a box with a label on it, and people use the label to explain themselves to themselves in ways that close more doors than they open. The MBTI, the Enneagram, the Big Five, the DISC — each of these captures something real. None of them captures everything real. And the ways in which they get misused tell us something worth understanding about how difficult genuine self-knowledge actually is.
What Personality Research Actually Says
The scientific status of personality models varies considerably. The Big Five — Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism — has the most robust empirical support. Decades of research from multiple countries and independent labs confirm that these five dimensions capture meaningful and relatively stable variation in human behavior. They predict outcomes across a wide range of domains: job performance, relationship quality, health behaviors, political orientation. The MBTI, by contrast, has a significantly weaker empirical record. Research from multiple universities has found poor test-retest reliability — a large percentage of people who take the MBTI and retake it weeks later get a different type result. The Myers-Briggs Company has invested in improvements over the years, but the basic structure was built on Jungian theory rather than empirical personality science, and that origin shows in the psychometric properties. What research from Vrije Universiteit Amsterdam and other institutions consistently finds is that personality traits are real, moderately heritable, and relatively stable over adulthood — but also responsive to experience, role demands, and deliberate effort in ways that fixed-type frameworks understate.
The Type Becomes the Self
Here is the thing I see happen repeatedly. Someone takes a personality assessment and receives a result that resonates. They feel recognized — the description captures something they have always sensed about themselves but could not articulate. That recognition is real and sometimes genuinely valuable. But then the type starts doing something more than describing. It starts explaining, and then excusing, and then predicting. "I'm an introvert, so I need to leave early." "I'm a Seven, so commitment has always been hard." "I'm a high D, so that's why I come across that way." The type, which was meant to illuminate, becomes a script. Behavior that might be examined and changed instead gets attributed and accepted.
A Tangent Worth Following
The philosopher Ian Hacking introduced the concept of "looping effects" — the way that human categories, once named and circulated, change the people they describe. When a category of personality or behavior becomes widely known and applied, people begin to identify with it, perform it, and organize their self-understanding around it. This is different from how categories work in the natural sciences: electrons do not know they are electrons and do not behave differently as a result. People do. The proliferation of personality typing in popular culture is a genuine looping effect in action. The categories shape the people they were designed to merely describe.
What Types Get Right
This is not an argument for abandoning personality frameworks. At their best, they do useful things. They normalize diversity of temperament. They help people understand that their needs — for solitude, for stimulation, for structure, for variety — are legitimate rather than defective. They provide a vocabulary for conversations about difference that can improve team dynamics, relationships, and self-compassion. A type is a useful approximation. The trouble starts when the approximation gets mistaken for the territory.
The Starting Point Frame
The most accurate way to hold a personality type is as a starting point. Here is a reliable pattern in how you tend to behave, weighted by your experience and disposition. Now: where is it serving you well, and where is it limiting you? Where have you grown beyond it? Where have you chosen deliberately against the pattern? Research on personality change — a field that has expanded significantly in recent years — consistently finds that personality traits do shift over the lifespan and can shift in response to deliberate effort. Work from the University of Illinois at Urbana-Champaign found that people who set goals specifically aimed at changing trait-related behavior showed measurable trait change over a sixteen-week period. The trait is real. The verdict is not. You are more than your type. The type is an entry point, not an endpoint. Use it to learn something and then keep going.
Want to discuss this with Sophie Laurent?
No signup needed · Start chatting instantly
Ask Sophie Laurent About This →