MBTI: Debunking the Debunkers (again)

I have been having a "discussion" in The Science Enthusiasts group in Facebook. Someone posted another "debunking the MBTI" article (from 2014). Surprisingly, this one doesn't cite "Wharton Scholar, Adam Grant".

The article includes this statement: "Several analyses have shown the test is totally ineffective at predicting people's success in various jobs, and that about half of the people who take it twice get different results each time."

Well, yes. Both of those things are, at least partially, true. Neither, however, is an argument that "debunks" the MBTI.

... the test is totally ineffective at predicting people's success in various jobs

This statement is true (surprise!) However, it doesn't "debunk" the MBTI. The MBTI is not meant to be "effective at predicting people's success in various jobs". The test is also ineffective at predicting people's IQ, how they'll vote, or where they live.

In fact,

It is unethical and in many cases illegal to require job applicants to take the Indicator if the results will be used to screen out applicants. The administrator should not counsel a person to, or away from, a particular career, personal relationship or activity based solely upon type information.

(From Ethical Guidelines at myersbriggs.org)

Type may give a person some clues, but the only thing that can come close to predicting success in a job is past history in a similar job.

The statement is true but, so what? Next?

... about half of the people who take it twice get different results each time

This one is only partially true (it's actually less than half) as well as both misrepresentative and irrelevant.

On retest, people come out with three to four type preferences the same 75% to 90% of the time. When a person changes type on retest, it is usually on one of the dichotomous pairs (e.g., E-I or S-N), and in a dichotomy where the preference clarity was low.

(From Reliability and Validity at myersbriggs.org)

The MBTI is a set of 93 forced-choice preference questions. Each asks, of the two options, do you prefer A or B? That's it. There's no option for "sometimes". There's no option for "both" or "neither". A or B?

Do you prefer to write with your right hand or your left? Do you prefer studying STEM or the Humanities? Do you prefer to play in the band or on the football field? Choose one.

The questions do not change from one assessment to the next. Therefore, the only way for a person to get a different result is to answer the questions differently the next time.

If you answer differently, you should have a conversation with yourself to determine: why did you answer differently? Also, I challenge anyone to tell me about a "test" that they have taken in which they can answer the same questions differently each time and still get exactly the same results.

To misquote Shakespeare, "The fault, dear friend, is not in the instrument / But in ourselves..."

I could write an article saying that "several analyses have shown that whether someone is left-handed or right-handed is totally ineffective at predicting people's success in various jobs, and that about half of the people studied actually use the other hand at different times or for some activities" and readers would laugh at me. Nevertheless, statements like the ones in the article sound good to people who don't understand the basic theory.

All articles like this one prove is that neither the person who wrote it, nor the person being quoted in it, understand Psychological Type theory or the MBTI.

Back to the Facebook Discussion

I've been trying to explain what Type and the MBTI are (and what they are not) to a bunch of people who don't understand the underlying concepts. That is painfully difficult when they think they already know.

I'm being accused of

• only supporting the MBTI because I'm certified and I get paid to do so (I wish!)
• only supporting the MBTI because "nearly everyone has some pet pseudoscience or woo that they defend"

If all psychological theories are "pseudoscience or woo", I suppose that latter claim could be true. There is, however, brain study data that supports, at least, introversion and extraversion, as well as data that may support the other dichotomies of Type. Also, we should not completely discount anecdotal data. Millions of people have not only verified their own Type, they have similarly verified other Types as being definitely not like them.

When one person on Facebook asked me, "If [the MBTI] serves a legitimate purpose, maybe you can explain what that is", I attempted to do so, replying

The MBTI is a self-development tool. Learning about Type can help you better understand and communicate with family, friends, and coworkers.

He then replied "That's more like a sales pitch than an explanation." (Um...?)

I have no idea what he wanted me to say. I guess he wanted to hear that there is some sort of predictive result. But then we're back to what the MBTI isn't.

Another person said: "The thing is that people can have legit complex neuro dx that requires medical intervention and they walk around thinking they're whatever letters the test told them they are. It's pretty dangerous in some regards."

Saywhat? The statement in the first half of the first sentence here is entirely unrelated to the part after "and". People can have legitimate disorders and walk around thinking they're good singers too. There's no correlation or contradiction or any other conclusions that can be drawn between "complex neuro dx" and the MBTI. It's a "yeah, so?" argument.

Person 3: "That's exactly it though, the idea that personality is fixed and will be constant in a persons life ... People change day to day. I may learn something today that changes my viewpoint on X later. etc."

Yes, but, Type never says someone's personality is fixed. The personality is a combination of many things: core Type, developed self, contextual self, learned skills and behaviours... Type is innate; personality flexes. Nothing said here contradicts the theory (although the first half of the statement shows a misunderstanding of the theory).

Person 4: "MBTI has very low validity and has been superseded by the far more robust Big 5/OCEAN (Costa and McRae). MBTI has terrible test retest validity, whereas OCEAN far better and much more thoroughly researched. You can pretty much discount anyone who talks in MBTI terms."

No. Just, no.

First, because the MBTI does not have "very low validity". Validity is the degree to which an instrument measures what it intends to measure, and the degree to which the “thing” that the instrument measures has meaning.

If personality type is real (or rather, if it reflects the real world with accuracy), then we should be able to use MBTI type to understand and predict people's behavior to some degree. Type should help us differentiate the values, attitudes, and behaviors of different people.

Many studies over the years have proven the validity of the MBTI instrument in three categories: (1) the validity of the four separate preference scales; (2) the validity of the four preference pairs as dichotomies; and (3) the validity of whole types or particular combinations of preferences. Many of these studies are discussed in the MBTI® Manual (published by CPP).

Reliability and Validity at myersbriggs.org.

Second, comparing the MBTI to the Big 5 simply makes no sense. One is Type theory; the other is Trait theory. It's like saying that geometry has been superseded by the far more robust trigonometry. It's just not a useful comparison.

Sigh.

I've backed away from the "discussion" and am taking long slow breaths.