Is the MBTI scientifically valid?

Partly. The traits it measures are real and correlate with the Big Five, but the binary sorting, the weak test-retest reliability, and the lack of predictive power for outcomes like job performance mean it should not be used for high-stakes decisions.

Is MBTI useful at all?

Yes, as a shared vocabulary for self-reflection and for talking about differences with partners, family, and teams. It becomes a problem when a type is treated as a verdict or a hiring tool.

The Insight Game › Guides › Is MBTI Pseudoscience? An Honest Answer From a Therapist

How Type Works

Is MBTI Pseudoscience? An Honest Answer From a Therapist

By Michael A. RoBards, LCSW

Ask a psychologist about the MBTI and you may get an eye roll. Ask someone who found the description of their type and felt seen for the first time, and you will get a completely different answer. Both of them are responding to something real. Here is the honest middle.

What the critics are right about

The reliability is weak. In reviews of the research, roughly half of people receive a different four-letter type when retested a few weeks later (Pittenger, 1993; 2005). A measure of stable personality should not do that.
The binary sorting is the problem. Studies of the underlying data find the four dimensions are continuous, not two humps (Stein & Swan, 2019). Most people sit near the middle on at least one. Cutting a bell curve in half and naming the halves creates sharp categories where nature put a slope.
It does not predict much. Type is a poor predictor of job performance or life outcomes, which is why using it to hire or promote is a genuine misuse.

What the critics get wrong

“Pseudoscience” implies it measures nothing. That is not what the data say. The MBTI dimensions correlate meaningfully with the Big Five — the Extraversion scale, for example, tracks Big Five Extraversion at roughly .74 (McCrae & Costa, 1989). It is measuring real traits, less precisely and in a less useful shape than the best instruments do. That is a design criticism, not evidence of nothing.

It also matters what you use it for. A tool that is a bad fit for hiring can be a good fit for a couple trying to understand why one of them needs the weekend planned and the other needs it open.

How I actually use type

I am a therapist, and I use type in my practice. Not as a diagnosis, never as a label to hide behind, and not to predict anything. I use it as a vocabulary. When a husband can say “I'm not being cold, I'm a Thinker and I process out loud through logic,” and his wife can say “I'm not being dramatic, I feel the impact first,” the conversation changes. They stop fighting about whose way is correct and start translating. That is the useful, honest, non-pseudo version of type.

The rules I keep

Treat a type as a description, not a verdict. If it stops fitting, it is allowed to change.
Read the near-the-middle letters as nuance, not error.
Never use it to decide who to hire, date, or trust. Use it to understand someone you already know.
Be suspicious of anyone, including a test, that claims to have you figured out in twelve minutes.

Where the Insight Game stands

We do not claim to have escaped these problems — no typing tool has. What we try to do is be honest about them: tell you when your result is close, point you to both neighboring profiles, and publish our live agreement data on the research page so you can judge for yourself. Type held lightly is a good tool. Type held tightly is where the pseudoscience accusation starts to stick.

References
McCrae, R. R., & Costa, P. T. (1989). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality, 57(1), 17–40.
Pittenger, D. J. (1993). The utility of the Myers-Briggs Type Indicator. Review of Educational Research, 63(4), 467–488.
Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal, 57(3), 210–221.
Stein, R., & Swan, A. B. (2019). Evaluating the validity of Myers-Briggs Type Indicator theory. Social and Personality Psychology Compass, 13(2).

Play the free 5-minute Insight Game →

The Insight Game is built on the clinical Type and Temperament work of Dr. Martine J. RoBards, Ph.D. (1946–2017). Live community accuracy data is on the research page.