Psychology 101
Is the Myers-Briggs (MBTI) Accurate? What the Research Says
The CQ Team · 6 min read · Oct 16, 2026
Partly. Its scales give consistent scores, but about half of people get a different four-letter type on retest, and there is no evidence it predicts job performance. What the research says.
Its four scales give consistent scores: retaken weeks later, they correlate .81 to .86, the Myers & Briggs Foundation reports. The four-letter type is less steady, because most people score near the middle of at least one scale, and about half get the same whole type on retest. There is no evidence it predicts job performance, and its publisher says so itself.
Six things people say about the Myers-Briggs, and what the research says.
| Claim | Verdict | What the research says |
|---|---|---|
| Half of people get a different type when they retake it. | Roughly true | The publisher says about 50% get the same whole type on retest, and 90% keep at least three of the four letters. Older forms did worse: a median of 40% in its 1985 manual. |
| Its four scales give consistent scores. | True | Retaken 6 to 15 weeks later by 1,721 adults, the four scales correlated .81 to .86, according to the Myers & Briggs Foundation. |
| Most people fall clearly on one side of each letter. | False | In about 12,000 people, the scores were “strongly center-weighted”: most sat near the middle of each scale, not in two groups (Bess & Harvey, 2002). |
| It predicts how well you'll do in a job. | No evidence | Its publisher says there have been “no meaningful studies” of this, and calls using it for hiring unethical. |
| It measures some of the same things as the Big Five. | True | In 1989 psychologists Robert McCrae and Paul Costa found its four scales measure parts of four of the five Big Five traits. Emotional stability is the one it leaves out. |
| If the description sounds like you, the test is accurate. | Not on its own | In a classic demonstration, 39 students rated one identical, generic description 4.3 out of 5 for accuracy. Feeling described is easy. |
The two retest numbers measure different things. The scale scores move very little between tries, so they correlate highly. But because most people sit near the middle of at least one scale, a small move is enough to cross the line, and the letter flips.
| Same type by pure chance | 6% |
| Same four letters on retest | about 50% |
| Same on at least three letters | about 90% |
Retest figures from the publisher's MBTI Facts; chance is 1 in 16.
Less than trait tests do. In a 2024 study of 559 US adults, a Big Five test predicted about 40 life outcomes about twice as well as an MBTI-style type. Back in 1991, a National Research Council committee found “not sufficient, well-designed research” to justify using the MBTI in career counselling (In the Mind's Eye).
| Big Five | 0.23 |
| Big Five without emotional stability | 0.18 |
| MBTI-style, as scores | 0.15 |
| MBTI-style, four letters | 0.11 |
Average accuracy (R) predicting about 40 life outcomes, 559 US adults.25. The MBTI-style test was Clearer Thinking's own, and the Myers-Briggs Company called the comparison misleading. Clearer Thinking, 2024
Partly because it is pleasant to read. Type reports describe preferences rather than weaknesses. In the same study, 10% disagreed that the type report made them feel good about their personality, against 19% for the Big Five report. In a November 2024 follow-up, 241 people rated the type report as the most accurate description of themselves. The 1991 committee noticed the same thing: the MBTI gives “generally positive feedback”.
The Foundation describes it this way: it sorts for preferences and “does not measure traits, ability, skill, or character” (myersbriggs.org). Used that way, as a shared vocabulary for how people prefer to work and talk, it does what it claims. For decisions about people, such as hiring, its own publisher says not to use it.
Its scales are reliable and overlap with four of the Big Five traits. The four-letter types are less well supported: scores do not split into two groups, and about half of people get a different type on retest.
Critics use the word because of the type claims. The scales do measure real differences, so a fairer summary is weak as a set of types and useful as a description of preferences.
Its publisher says it is unethical to use it for hiring or selection, and that it was not designed to predict job performance.
For predicting real-life outcomes, the Big Five. In one 2024 study it predicted life outcomes about twice as well as an MBTI-style type test.
No. 16Personalities uses the same letters but says its model reworks the Big Five, and it adds a fifth letter, -A or -T.
Myers-Briggs Type Indicator, Myers-Briggs and MBTI are registered trademarks of The Myers & Briggs Foundation. Cerebral Quotient is not affiliated with The Myers & Briggs Foundation or The Myers-Briggs Company.
Curious where your needle sits? Take any assessment free and find out in minutes.
Take a test →