The Bias Is Strong With MBTI

raph @ 2014-10-29 12:51

Once upon a time, at work, I took a personality test.

It was a three day course, and the whole team participated. Early on, we took a twenty-minute test and added up the scores to end up with four numbers. Each number tells you where you sit one of four linear scales. Each scale is split through the middle and - depending on which half your number falls on - you get one of two letters per scale.

The resulting four letters combine to form your Myers-Briggs Type Indicator (MBTI). By knowing someone’s MBTI, you’re supposed to be able to better predict and understand another’s preferences and behaviour.

“Enterprise” loves MBTI because of how it puts people into neatly labelled boxes. Consultants love MBTI because enterprise loves MBTI. Which is why I was sitting in said course.

I got INTP.

I’ll be honest and admit that when I began writing this, I couldn’t even remember what the last three stood for. I had to refresh my memory on Wikipedia.

25% science.

The outcome on the first scale can either be introvert or extrovert. In our group, this scale seemed to yield the clearest results. Given that most attendees were software developers, it should come as no surprise that most of us ended up getting I.

Why? There’s a large body of research and a variety of literature on the introvert-extrovert spectrum. Software development is consistently described as a field that attracts introverts. So I’m going to let this one pass as it’s both scientificallly grounded and relevant in day-to-day life.

I won’t reiterate what is already well-described in aforementioned books. But just to give one example, it’s why many of us prefer the asynchronous nature of electronic means of communication (i.e. email, chat) over synchronous face-to-face “talk”.

75% percent bullshit.

My beef with the other three is not that they don’t apply to me. Quite the opposite actually. It’s because they are defined in such a way that they apply to everybody and are inherently context dependant. Testing for them only captures one’s preference at the time of the test, under the conditions of the test. But under different circumstances I might as well come out as the polar opposite.

At one point we were presented with a hypothetical scenario: you are the coach of a children’s football club. You manage to get just two tickets for a big game. How to do you decide who goes, and how do you communicate that to the children?

According to MBTI, a typical thinker would rank kids by some level of skill and reward the best with the tickets. A feeler would be more worried about the other kids being sad or dissapointed, and try to work out a more social solution with the group.

The problem here is, t and f are polar opposites of the same scale. I was scored to be a strong t based on the questions in the test earlier. And yes, the thought of ranking the kids by their dedication to the club did cross my mind. But I was just as concerned about the other kids being dissapointed as the next f.

I had suggested alternatives, one being to take all the other kids to another event rather than not rewarding them at all. However, they depended on variables not provided. The scenario - being part of a pre-defined course programme - was contrived in order to set the stage for the purposes of MBTI. It lacked much information which would have influenced my solution. Like the number of children, their age, the price of tickets vs family income…

But alternatives weren’t part of the programme. The trainer made the point she had come to make. And I learned that there is no room for alternatives in the littlle boxes which are MBTI (stereo)types.

Life is not linear.

A linear scale is simply not sufficient to accomodate this spectrum of decision making. Flattening it all like MBTI does disposes too much information to have relevance in the real world.

If I was to be generous (which apprarenltly I’m not, because I’m not a f) I could admit that MBTI gave a correct prediction that I had considered the ranking option, and as a result could have decided something similar. But that’s pretty thin, and practically impossible to disprove.

When people end up doing what MBTI predicted, it appears to confirm the correctness of MBTI. When people decide differently, it’s either ignored or handwaived away by the general cop-out:“MBTI only describes a preference”

See also: confirmation bias.

I did raise my concerns during the course. The defense was the same every time. Our “trainer” would argue that MBTI merely describes a preference, not a general rule (huge red flag). I might have agreed if people hadn’t instantly fell into a pattern of reducing others to their scores. “P’s do x while J’s do y”, the trainer would say. So even if we gave MBTI the benefit of the doubt, and went with the preference hypothesis, in practice it is not used as such. Human minds don’t work like that.

Most of us wore their score like a badge. Literally: many wrote their scores in big letters onto the name signs in front of them. Gradually, people with similar scores formed little teams. In what superficially appeared to be innocent fun, teams ganged up and mocked the others’ MBTI-stereotypes. “Haha, you X be so crazy.”

See also: othering.

Pseudo-science is pseudo.

The sixteen MBTI type profiles are strangely reminiscent of zodiac signs. They are written in such a way that they could apply to anybody. After we got the hard-earned score, we had a chance to read all profiles. At this point we had already been told what we “are”, and our minds were already set to read other profiles as if they didn’t match.

I suspect if we had been given the profiles without any prior labelling, it would have been very difficult to find a single one that matches. People would have identified with different profiles under different circumstances. People are not one-dimensional.

The fine print states MBTI is just a preference, but the scores quickly become comfortable boxes. Once you’re convinced that you are an XXXX, you identify with it so much that you become it. When a situation arises that seems to fit the profile, it’s “wow, that’s obviously because I’m X!”. When the profile is at odds with behaviour, it is brushed off as an outlier and ignored.

See also: astrology.

See also: pseudo-science.

Why is MBTI so popular then?

Most people enjoy taking personality tests and sharing the outcome. We are quick to identify with the results, especially when they confirm our expectations. Just look at how successful pseudo-scientific tests are in magazine’s or social media (i.e. Facebook). MBTI tests or derivatives thereof are very popular on the web.

It’s easy to do the test. And passing a test gives you a sense of accomplishment. It’s fun to label oneself and others. It’s fun to make fun people’s quirks. It makes sense on the surface and contains just enough science to make it get past our bullshit sensors.

I buy into the introvert-extrovert spectrum, but the rest is too dynamic for anyone to have a meaningful score. MBTI has very little practical relevance. It oversimplifies human psychology and the complexity of decision making. The resulting simplicity is what makes it so tempting to buy into MBTI. I think it’s dangerous how it sets wrong expectations when people use it as a predictor. Which they do, despite the disclaimers.

So, looking at all this I suppose I’m a typical t after all.

(Or am I?)