Four letters. Sixteen combinations. A vocabulary so widespread that people announce it on dating profiles and put it in email signatures, and a research literature that has spent decades explaining why it should not work. The Myers-Briggs Type Indicator is the most successful personality instrument ever made and among the least respected by the discipline it borrowed its ideas from. Both of those facts deserve an explanation, and neither is well served by mockery.
A mother, a daughter, and a book from Zurich
The story starts in Michigan with Katharine Cook Briggs, born in 1875, a college graduate at a time when few women were, and a woman with an enormous appetite for the systematic study of human character. She turned her home into what she cheerfully described as a laboratory for child-rearing, kept detailed records of her daughter's development, and by the early 1920s had sketched out a homemade typology of her own.
Then, in 1923, the English translation of Carl Jung's Psychological Types reached her. Jung proposed that people differ in the direction their attention naturally flows, outward or inward, and in which of four mental functions they rely on most: sensation, intuition, thinking or feeling. Briggs recognised something better built than her own scheme and adopted it wholesale. She wrote about it for general readers in the mid-1920s, translating Jung's dense prose into something a magazine subscriber could use.
Her daughter Isabel Briggs Myers inherited the project. Isabel was not an academic; she was a prize-winning mystery novelist whose 1929 book Murder Yet to Come had won a national competition. What she brought to her mother's enthusiasm was narrative economy and an unusual gift for writing questions people were willing to answer honestly.
A test born of wartime reshuffling
The Indicator's practical impulse was the Second World War. Women were entering factories and offices in vast numbers, often into work chosen by nothing more considered than whatever vacancy existed. Myers believed that if people understood their own inclinations they would land in roles that suited them, would do those roles better, and would be less miserable. She began building the questionnaire in the early 1940s.
She had no training in psychometrics, so she acquired some. Myers apprenticed herself to Edward Hay, a Philadelphia personnel consultant, and learned item construction, scoring and validation from the inside of a human resources department rather than from a university. Along the way she made a decision Jung had not authorised: she added a fourth pair, judging versus perceiving, describing how a person prefers to deal with the outside world. It is arguably the most useful of the four in everyday life and the least defensible as Jungian scholarship.
How four letters became a business
The Educational Testing Service picked up the Indicator in 1962 and published a research edition. Internally, the reception was cool; ETS staff psychologists were unconvinced by the evidence and the relationship eventually ended. Myers found other allies, notably the psychologist Mary McCaulley, with whom she founded a centre in Florida devoted to the instrument, and in 1975 Consulting Psychologists Press took over publication and began selling it in earnest to organisations.
That was the turn that mattered. Corporate training departments in the 1980s and 1990s wanted a language for talking about difference that did not sound like criticism, and the Indicator supplied one. Every type is framed as valuable. No result implies illness, weakness or deficit. A manager could run a workshop on team dynamics without anyone leaving the room feeling insulted, which is a commercial property of enormous value and has very little to do with predictive validity. Millions of people take some version of it every year, most of them at work.
What psychometricians object to
The complaints are specific rather than sneering, and worth stating plainly.
The first is the dichotomy problem. The Indicator reports you as either an extravert or an introvert, and to do so it must draw a line. But scores on the underlying questions are not bimodal, with two clumps on either side of a gap; they are roughly normally distributed, with most people bunched near the middle. The cut point therefore falls exactly where the greatest number of people sit, so a modest change in a handful of answers flips a letter and, with it, your entire type.
The second follows directly: retest instability. David Pittenger's much-cited 1993 review noted that when people retake the Indicator after only a few weeks, a large proportion, in some studies approaching half, come out as a different type. A measure that reassigns your identity between one month and the next is describing something, but it is not describing a fixed type.
The third is what it leaves out. When researchers correlate the four Myers-Briggs dimensions with the five-factor model, four of them map on reasonably well: extraversion to extraversion, sensing-intuition to openness, thinking-feeling to agreeableness, judging-perceiving to conscientiousness. The missing one is neuroticism, the trait most strongly associated with mental health, stress and relationship satisfaction. That omission is not an accident but a design principle. Myers wanted a system in which no result was bad news, and the price of that kindness was excluding the dimension that carries much of the predictive weight.
Formal reviews reached similar conclusions. In 1991 a National Research Council committee, asked by the US Army to evaluate a range of popular performance-enhancement techniques, examined the Indicator and concluded that its popularity was not matched by the research evidence supporting its use for that purpose.
Why it is still loved
None of which explains why a lapsed INFJ will still tell you, with feeling, that the description was the first time anyone had put words to how they experience a crowded room. It is worth taking that seriously.
Part of the answer is the writing. Isabel Myers was a novelist, and the type descriptions read like character sketches rather than score reports. They describe patterns of attention and inner experience, not behaviour, which is precisely the territory people find hardest to articulate about themselves and most relieved to see named. Part of it is the flat structure: sixteen equally legitimate ways of being, none ranked. And part of it is social. Types create a shared shorthand, a way of saying "I need an hour alone after that meeting" without it sounding like a complaint.
Merve Emre's history of the instrument, The Personality Brokers, is unsparing about its evidential thinness and simultaneously alive to what it gave people: a permission structure for being different, invented by two women who were themselves shut out of the profession that would later dismiss their work.
Holding it lightly
The reasonable position is neither reverence nor contempt. As a formal measure of stable individual differences, the Indicator underperforms alternatives that cost less and predict more. As a conversation starter, a vocabulary and a prompt for self-reflection, it does something real, and the fact that the letters are unstable matters less when nobody is making a hiring decision with them.
The trouble begins only when four letters are treated as a fact about a person rather than a description they happened to endorse on a particular afternoon. Used as a mirror, it is a good one. Used as a filter, for jobs, promotions or partners, it is asking a novelist's character sketch to do an actuary's work.
