Taking the MBTI with HFA

---

Back in the early 2000s I was working in America for Cable & Wireless. One of my colleagues was Craig, a Duke University professor of psychiatry; his wife, Dee Ann, was also trained in psychotherapy. I remember talking to her at one of our works dinners about the Myers-Briggs classification scheme and when it didn't work. She replied that it tended to give erratic results with people who exhibited abnormal mental conditions, such as schizophrenia and Asperger's syndrome.

And I left it there.

Today I'm wondering what happens if a person with High-Functioning Autism (formerly Asperger's syndrome) takes the Myers-Briggs Inventory. How does their MBTI come out?

ChatGPT tells me...

---

How HFA People Score on the Myers–Briggs Type Indicator

The Myers–Briggs Type Indicator rests on population averages of neurotypical psychology. It was never built to describe autistic cognition. When people with high-functioning autism (HFA) take the MBTI, the results often wobble or misclassify. The same is true of the Five Factor Model, though in a different way.

Take the S/N dimension. Some autistic people live in concrete domains: bus timetables, mechanical systems, the fine grain of sensory detail. They look like S. Others immerse themselves in mathematics, philosophy, speculative worlds. They look like N. Both are common.

The test assumes the two poles are exclusive; autism ignores that rule. A developer who can design high-performance real-time transaction systems may come out as a classic N for abstract thinking. Yet ask him about politics or engage in small talk and he seems incurious, even obtuse - answers that tilt back toward S.

Extraversion is another fault line. HFA people can be talkative, sometimes performatively gregarious, yet maintain no close friends. The MBTI interprets their verbal energy as E, while their lived experience is solitary. A literal answer to “Do you enjoy going to parties?” might be “Yes, I like the food and the noise,” even if the same person leaves early, interacts with no one, and has no social bonds. The test confuses stimulus-seeking with sociability.

On the T/F scale, many answer as T. They rely on explicit reasoning, exploiting rules and consistency. That does not mean they lack affect, only that their affect is not measured in the MBTI’s idiom. Similarly, the J/P axis is confounded: autism’s need for sameness, its preference for routine, shows up as J. But this is neurology, not “personality preference”.

If you shift to the Big Five, the pattern persists. “Openness” can score high if interests are abstract, or low if they are repetitive and concrete. “Conscientiousness” may appear strong on orderliness but weak on adaptive planning. “Agreeableness” is pulled down by low empathy-related items, though fairness and honesty are often high. “Neuroticism” is inflated by anxiety, a frequent comorbidity. Instability is the rule.

The real problem is structural. These instruments assume traits cluster as they do in the neurotypical bell curve. Autism reshuffles the deck. Literal answering, context-bound interpretations, and construct mismatch ensure that the outputs are noisy. What looks like inconsistency is the wrong measurement framework.

Better alternatives exist. The Empathising–Systemising Quotient, the RAADS-R, and sensory processing profiles capture dimensions orthogonal to MBTI and FFM. They measure the axes on which autistic cognition varies.

So when an HFA person emerges from the MBTI as ENTJ one month and ISFP the next, it is not a crisis of self-knowledge. It is the instrument failing to fit the mind it tries to describe.

---