Does The Myers-Briggs Test Work? | QI

This clip is from QI Series Q, Episode 5, 'Questions & Qualifications' with Sandi Toksvig, Alan Davies, Ade Adepitan, Nish Kumar and Holly Walsh.

[tricesimo]

"I took a personality test... it came back negative."

[MerricMaker]

My dad took the MMPI while in college, was told that it was confidential of course. Took the test with the goal of coming across as the biggest psycho possible, and his memory was good enough to just stay consistent on the baseline questions.

Test done, a day or so passes, guess who got called into the Dean's office due to the results of a "confidential" test.

[MZZenyl]

Tested it myself some years ago - took the same online test twice with roughly three months in between, with the same questions, and two letters had changed.
It doesn't test your personality, so much as it tests your mood and state of mind at the time of taking the test, with sufficiently vague categories that anyone can see themselves in all categories to some extend.
A horoscope for people who think the word "business" is hot.

[Nodnarb]

I took the Meyers-Briggs test before, and it said that I had the personality of a serial killer. Bullocks. ONE person. ONE PERSON!

[dunbarfinger]

The problem I have with the introvert/extrovert designation is in some cases the preference for introverted pursuits wasn't because of internal traits but was a learned behavior caused by external social pressure. A kid who was unpopular in school may grow up to think "Social activities means teasing, shunning, and bullying so of course I don't like them.". In reality that kid who is now an adult may subconsciously want more extroverted activities but have too much fear about them to consciously realize it.

Calling that person an introvert almost feels like letting his childhood bullies off the hook for the damage they did. It's like saying no harm was done because he *likes* being alone.

There are people who genuinely do like being alone. I'm not saying they don't exist. I'm just saying that they get lumped in together with the ones who learned it as a defense mechanism and it hides a problem to do so.

Because these kinds of defense mechanisms often get done subconsciously, the subject doesn't often realize it when answering Meyers-Brigs questions.

[seadawg]

I got a book years ago and could never get past the first set in questions because almost every single question I thought “well, it depends on the situation and my mood”
I later found out the Jung found the idea “finding your type” to be exactly the opposite of what he thought was good. It’s ideal to be most fully in the middle and be able to react appropriately to circumstances.

[hjt]

The Myers Briggs test is a classic example of the Barnum effect, and Sandi comparing it to horoscopes is pretty accurate

[DemstarAus]

I'm a Ravenclaw! Has just as much relevance to an employer as INTP or whatever.

[tvdan]

I've spent most of my life being beat over the head by that damn test. Had to take it before college so they could "better match" roommates. My roommates and I couldn't have been more poorly matched. It was a disaster.

Professionally, I have found that companies that use it tend to want their employees to be robots. I say that as an escapee of Big Box Retail.

[alastairmciver]

I think some people are comforted by finity. They like the idea that there are a limited number of personality types, or a limited number of story-plots. The numbers vary, but people will always come out with these ludicrous, limiting classifications because they like a world where everything fits neatly into boxes. People who are a bit too smart for horoscopes will turn to Myers Briggs, but it is the same impulse, to limit and classify, that drives both. Personally, I hate the idea of a limited, tickybox world, so I'm delighted to know that Myers Briggs is bollocks.

[puirYorick]

A lot of the methods used by many HR people amount to woo hidden by many over-priced brand names all in the name of selling consultancy and annoying self-help reading materials that you're meant to order from their web site. I still have a couple of certificates I was awarded for playing along during training sessions where I used to work.

[BuildinWings]

Thank you for bringing back scolding-Sandy in the bumper. She reminds me of my high school German teacher.

[samuelrobinson]

'I am a special agent of God'

*Man with long beard puts hand on head*

[ankyfire]

Honestly, I feel like D&D characters are better descriptors than MBTI.
I'm chaotic neutral and everyone who knows me (and is nerdy enough to know d&D) agrees.

[MsumireStory]

I would have immediately screamed, "Define unusual!". I always thought there was actual research behind those tests...

[akirakolappalainen]

A friend of mine had to take it when applying for a job as a bus driver. He didn't get the job literally because of the test, and was told he's too Machiavellian to drive a bus????

[SylviusTheMad]

I always wondered why Myers-Briggs was so accurate in some ways and then so wildly wrong in others.

Turns out it's all just made up, which I didn't realize until I had an HR department implement the even weirder Enneagram, which drove me to do some research.

[johaquila]

The Briggs-Myers test tests 4 dimensions, with a binary result (one of two letters) in each. This results in one of 2x2x2x2=16 possible 4-letter outcomes.

The problem isn't that the dimensions aren't valid. They mostly are. They correspond roughly to 4 of the 5 well-established dimensions known as Big Five personality traits in academic psychology. (The one dimension it lacks is the one called neuroticism.)

Self-reporting is probably not the problem, since the Big Five personality traits are also determined through self-reporting. The real problem is fitting the test result into a binary outcome. This is completely unscientific. It's OK to fit continuous data into discrete categories, but the categories need to be appropriate for the data distribution. In this case, for each of the 4 dimensions of the Briggs-Myers test, the outcome statistics form roughly a Bell curve. A sensible categorization would therefore be to use 3 categories for each dimension, such as extraverted, ambiverted and introverted, with the vast majority of people coming out ambiverted, i.e. 'normal'.

The problem with meaningful personality tests is exactly that they are meaningful. Knowing whether a person is VERY introverted, VERY extraverted, VERY curious, VERY cautious, VERY organized, VERY careless, VERY critical or VERY agreeable (most people are 0 to 2 of these things, but certainly not 4) is actually important when considering to hire them. But for each of these dimensions, most people are close to the center -- sometimes so close that they may come out on different sides when re-taking the test. And yet they are lumped together with those in one of the extremes. This takes the edge out of the personality tests, which is precisely what is needed for acceptability.

The Briggs-Myers test is like categorizing people into tall or short based on their heights. When seeing a crowd, we don't see 50% tall people and 50% short people. What we actually see is lots of people of average height, a few tall people, and a few short people. The same principles apply to personality dimensions.

[Acceleronics]

My company sponsored a "team building" 2-day off site that included the M-B test. It put me in the INTP box. Every description I've ever read about INTP fits me perfectly. My take is that these tests can be accurate, but it is easy to subvert them (see what I did there?).

[DexM]

Let's clarify: in Jung's topology, we're using both introversion and extraversion, just as we're all using sensing/intuition, we're all using thinking/feeling, and we're all using judgement/perception. If you look at the distribution across the population, you a have Gaussian (bell curve) on each of these axes (although the sensing/intuition one is skewed towards sensing). So most people (those who are not exactly at the center of the curve) have a PREFERENCE for introversion or extraversion, a PREFERENCE for sensing or intuition, etc. What Myers and Briggs did is to binarize this preference, which is completely stupid, since you can easily be classified as introvert one day and as extrovert another day if you're close to the center of the curve.