Please hit the ❤️ “Like” button at the top or bottom of this article if you enjoy it. It helps others find it.
The Myers-Briggs Type Indicator (MBTI) is pseudoscience.
For the uninitiated, the Myers-Briggs (as I'll refer to it since it's the way it's colloquially known) is a personality questionnaire/typing tool that categorizes people based on four dimensions: extraversion versus introversion, sensing versus intuition, thinking versus feeling, and judging versus perceiving.
For each dimension, people are put on one side or the other, giving 16 possible "personality types". People love to declare their type, especially "INTJ", the supposedly rare personality type that every person on the internet has.
The thing is, Myers-Briggs types are about as meaningful as taking a personality quiz to see what Schitt's Creek character you are.
(Actually, let me amend that. Apparently I'm Moira, which is so obviously incorrect it calls this quiz into question. This particular Schitt's Creek personality quiz is in fact worse than Myers-Briggs)
Myers and Briggs were a daughter-mother duo who came up with the "Myers-Briggs Type Indicator" based loosely on Jung's book, Psychological Types. The first articles they wrote about it were published in the 1920s.
They didn't use any systematic method to come up with their test and types. Instead they were based on the idiosyncratic observations of people (from Jung and their own experience). Which would be fine as an initial stab at a theory of personality.
The trouble is, it wasn't tested at all. Modern personality research is far beyond that, and there's no reason to be using a shot-in-the-dark model of personality when we have a much better modern model of personality: The Big Five.
The Big Five (or OCEAN) is the standard used in personality research, and has been for a long time. Actual personality researchers don't use Myers-Briggs because it just isn't a good measure of personality (more on that below).
Unlike Myers-Briggs, the Big Five personality traits were arrived at through a more quantitative approach, using lexical analysis combined with factor analysis on collected survey data (see Regan’s excellent post that details how researchers came up with The Big Five).
The analysis left them with five major personality dimensions: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism.
More importantly than the methodology to come up with these personality constructs is how they perform.
One way to test how they perform is by looking at how well they predict an outcome. If a measure is capturing a person's personality, it should predict something about their behavior.
Studies have repeatedly shown The Big Five is much better at predicting various behaviors of people than Myers-Briggs. For example, see this chart of their ability to predict various life events:
The amusing conclusion reached from this finding:
On average, the Big Five test was about twice as accurate as the MBTI-style test for predicting these life outcomes, placing the usefulness of the MBTI-style test halfway between science and astrology—literally. (Source)
Note that the figure has two values for Myers-Briggs: scores and categories. Scores means looking at where people fall in a continuous way on the four dimensions of a Myers-Briggs test, making it more similar to the Big 5 personality tests, which also give a continuous score on each of its dimensions.
Used as a score, Myers-Briggs doesn't do terribly, but it's clearly worse than The Big 5. Note that there's some overlap in what they measure. Both have a measure of "introversion versus extroversion", and studies have shown decent correlation between the Myers-Briggs measures and four of the five Big Five traits (neuroticism isn't well captured by Myers-Briggs).
Looked at this way, Myers-Briggs would just be a kind of outdated theory of personality, not a pseudoscience. The dimensions are wonky and don't stand up under modern research (like putting thinking/feeling on opposing sides of the spectrum), it lacks construct validity, and we have a better model. But hey, lots of scientific models are bad, that doesn't make them pseudoscience.
But a core part of Myers-Briggs is explicitly categorizing people instead of giving them a continuous score. Instead of being a 45 out of 100 on the introvert-extrovert scale, you're classified as an introvert.
When you use the categories, the validity of Myers-Briggs predicting anything useful drops further. The categories themselves have very poor reliability—if you test someone again after a few weeks, there is a good chance they will change categories. This isn't a great feature if you're claiming this is someone's personality type. The very definition of personality is something that remains relatively stable over time.
The reason people change types so frequently is because so many people fall in the middle of the distribution for personality traits (whether the Myers-Briggs traits or the Big 5).
If you're close to the mid-point on any trait, as most people are, a small change in score will tip you over the threshold and into another category.
Given all these problems with Myers-Briggs, why is it still so popular in the public consciousness?
A huge part of the answer is that Myers-Briggs is a business with a marketing department that sells certifications, courses, and books. They have a ton of connections to industry and education, and use them to pump money out of businesses and schools for their crappy personality tests that don't actually provide anything helpful. But since it feels psychology-ish, people mistakenly think it's based on real psychology research—a management professor I once had thought it was what psychology researchers use, and taught it to all her students.
But more importantly, I think it taps into our need for certainty and dichotomies in a way that better tools (like The Big Five) don't, due to its willingness to categorize people.
Black and White Thinking
It's much easier to write guidance for how to manage different personality types when you can put people into a few boxes. It's easier and more exciting to share your personality type online when it's a simple type (complete with a description) rather than a series of continuous scores.
I've written before about how messy data often gets boiled down to a simpler, definitive statement. Messy or continuous data is too cognitively expensive. Having a nice category to stick someone into makes them simpler to think about.
It also makes it easier for a company to write guides on how to treat their INTJ employees.
It's something we deal with in a lot of areas of life where we're given simplified stories. Complex data is boiled down to simple statements.
This is often necessary—we need to communicate things, and getting bogged down in every nuance means you'll lose the forest for the trees. Sometimes it's useful to just say "this person is introverted", rather than try to quantify how many standard deviations from the mean they are on the extraversion-introversion scale.
But it's good to remind ourselves to use some epistemic humility, that behind most nicely packaged simple statements there is a world of complexity that's been boiled down. As such, we should hold these sorts of statements—and our supposed Schitt's Creek character assignments—at arms-length.
Please hit the ❤️ “Like” button below if you enjoyed this post, it helps others find this article.
If you enjoy the newsletter, consider upgrading to a paid subscriber to support my writing.
If you’re a Substack writer and have been enjoying Cognitive Wonderland, consider adding it to your recommendations. I really appreciate the support.
I’m starting to rethink my “INTJ” neck tattoo now.
The addition of the validity chart showing correlation with life outcomes is key here. Yes Big 5 is more valid, but it’s still only a small correlation. In general practice, the cultural coupling of personality descriptors and future outcomes is over applied beyond this small correlation. It amazes me how even at Xmas my well educated family got into a whole conversation over star signs.