Myers-Briggs Type Indicator -- Pseudoscience?

[lpetrich]

Adam Ruins Everything - Why the Myers-Briggs Test is Total B.S. | truTV - YouTube a rather whimsical swipe against the creators of the Myers-Briggs Type Indicator. Doesn't address the MBTI itself, however.
The Bizarre Untold Origin Story of the Myers-Briggs Personality Indicator - VICE
The Myers-Briggs Personality Test Is Bullshit - VICE
 Myers–Briggs Type Indicator

The MBTI states that human personality has four dimensions:

• Extroversion - Introversion
• iNtuiting - Sensing
• Thinking - Feeling
• Judging - Perceiving

As usually applied, these dimensions are treated as binary variables and not continuous ones. Thus, one is either extroverted or introverted, and not anywhere in between, like being ambiverted. Since each of the four dimensions is treated as binary, this yields 2^4 = 16 personality types.

However, there is no evidence that we are binary in these dimensions, and no evidence that we fall into the 16 MBTI personality types. Furthermore, the canonical descriptions of these types are all rather general and rather flattering. This results in the Forer effect, much like horoscopes.

It is also curious that the MBTI is much better known outside of psychology research than the mainstream theory of personality, the "Big Five" five-factor model. In it, human personality has these five dimensions:

• Extroversion - Introversion
• Openness - Closedness
• Conscientiousness - Impulsiveness
• Agreeableness - Antagonism
• Neuroticism - Emotional Stability

The Big Five is based on rigorous hypothesis-testing, about as rigorous as one can reasonably get in studies of personality. However, the MBTI was constructed by observation and speculative theorizing. Parts of it were by Carl Jung, and these parts were expanded on by Katharine Cook Briggs and Isabel Briggs Myers.

Then there is the curious issue that the MBTI is much better known than the Big Five, even though the Big Five are more empirically based.

 

[phands]

Yes. It's pseudoscience. So is DISC and similar manage-o-babble.

 

[rousseau]

I'd call it less pseudoscience, more 'unrefined' science.

If you were to answer a set of 300 questions about your personality, in theory you should be able segment yourself into kind of a vague, ill-formed group, and you kind of, sort of know a little bit more about yourself. But the problem is that this group isn't necessarily predictive of your actual behavior.

Not unlike a lot of the psychology stuff that was going on in years past.