MBTI vs Big Five: what each test measures and where they overlap

Take two personality questionnaires in one evening. The first hands you four letters and a page that reads like a character sketch. The second hands you five numbers, and one of them says you worry more than seventy percent of people.

Guess which one ended up on coffee mugs.

The MBTI and the Big Five are the two frameworks most people run into, and they are usually presented as rivals: the popular one and the scientific one. That framing hides the more useful fact, which is that they are mostly describing the same things. Where they part company is in what they do with the measurement afterward, and in one dimension that the four letters leave out on purpose.

One started from a book, the other from a dictionary

The MBTI began with a theory. Carl Jung described a set of psychological types in 1921, and Katharine Briggs and Isabel Briggs Myers built a questionnaire to sort people into them. The categories came first and the questions were written to fit. Who created the MBTI tells that story.

The Big Five ran the other way. Nobody proposed five traits and went looking for them. In 1936 Gordon Allport and Henry Odbert went through an unabridged dictionary and pulled out the words English uses to describe a person, on the reasoning that any difference between people that matters will have earned a word by now. Later researchers had people rate themselves and each other on those words and asked a statistical question: which descriptions travel together?

Five clusters kept turning up. Ernest Tupes and Raymond Christal reported them in a 1961 Air Force technical report, Lewis Goldberg gave them the name Big Five in 1981, and in 1985 Paul Costa and Robert McCrae published the NEO Personality Inventory, a questionnaire built to measure them. The five are openness to experience, conscientiousness, extraversion, agreeableness and neuroticism.

So one framework is a theory that was later given a test. The other is a pattern in data that was later given names. That difference in origin explains most of what follows.

Where the two overlap

In 1989 McCrae and Costa put both questionnaires in front of the same people, 267 men and 201 women drawn from the Baltimore Longitudinal Study of Aging and their spouses, and published the result in the Journal of Personality. Each MBTI scale turned out to track one of the five.

• E and I track extraversion.
• S and N track openness to experience.
• T and F track agreeableness.
• J and P track conscientiousness.

The matches are not equally tight. The first two are close: the correlations they reported were around .7, which is high for two questionnaires written decades apart by people with different aims. Somebody who comes out as an introvert on one will usually come out low on extraversion on the other.

The last two are looser, between .44 and .49. A preference for feeling leans toward agreeableness and a preference for judging leans toward conscientiousness, but with plenty of exceptions in both directions. The study also found the J and P scale picking up a little openness, and its authors called the usual interpretation of that scale into question.

Their overall reading was double-edged. The four scales measure real aspects of four of the five major dimensions. They do not sort people into distinct types, and the authors argued the five-factor model was a sounder basis for interpreting MBTI results than Jung's theory.

The dimension the four letters leave out

Four scales, five traits. The one with no partner is neuroticism: how readily a person feels anxious, discouraged or rattled, and how long it takes to pass.

McCrae and Costa called its absence the most conspicuous gap and offered two reasons for it. Emotional stability was not part of how Jung defined his types. And the authors of the test were committed to the position that every type is equally valuable, which is hard to maintain about a scale where one end is plainly less comfortable to live at than the other.

That is a design decision, and it has a cost and a benefit. The benefit is a result nobody is embarrassed to share. All sixteen descriptions are flattering, which is a large part of why the letters spread and the five numbers did not. The cost is that two people with identical letters can differ enormously in how steady they feel, and nothing on the page says so.

Some popular tests patch the gap with a fifth letter. What -A and -T mean after a type covers where that suffix comes from.

A category versus a position

The second difference is in the reporting. The Big Five gives you a position on each trait, usually as a percentile. The MBTI takes a position and reports which half it fell in.

Nothing in the data justifies the cut. McCrae and Costa looked for the two-humped pattern that true opposites would produce and found it on none of the four scales. People spread along each one continuously. A letter is a position with most of the information removed, and that is the mechanical cause of the retest complaint covered in why your result changes.

A category has one advantage, though. It is a name. "I am an INTP" is something a person can say across a table. "I am at the thirty-first percentile on agreeableness" is not.

An INTP in Big Five terms

Run the translation on a real code. The INTP letters say: lower on extraversion, higher on openness, and, more loosely, lower on agreeableness and lower on conscientiousness. An ENTP differs in the first trait only.

Two cautions about reading it that way. "Lower on agreeableness" sounds like an insult and is not one; it describes somebody who will tell you the plan has a hole in it. And the last two translations rest on the looser correlations, so they are tendencies across a crowd. Plenty of individual INTPs are warm and never miss a deadline.

The translation says nothing about the fifth trait. An INTP can sit anywhere on it.

Which one to take

It depends on what you want the result for.

If you want a position on each trait, with the uncomfortable one included, take a Big Five inventory. For measurement it is the stronger of the two. It is also harder to do anything with socially.

If you want a vocabulary you can use with other people, the four letters are the one that everybody already speaks, and they are not empty: four of the five traits are in there. Hold the letters loosely where your score was close to the middle. Is the MBTI accurate goes through what the research supports and what it does not.

Neither one is a clinical or diagnostic instrument, and a high neuroticism score is not a diagnosis any more than a P is. MBTI is a trademark of The Myers-Briggs Company, which has no connection to this site. I run it by myself and I am not a psychologist. The test here reports four letters, and the sixteen type pages describe each combination, which is the half of the picture that is easy to talk about.