Every Toy in Its Box: The beautiful lie of Myers Briggs and modern personality testing
The first time I remember playing with a cootie catcher, I was perhaps in the fifth grade.
If you’ve never seen a cootie catcher (pictured here), it is simply a sheet of paper complexly folded, with a treasure of secret messages awaiting just below the folds. You would pick a number or a word and the person operating the cootie catcher would rapidly open and shut the origami contraption both vertically and horizontally in a manner that corresponded to the number you chose (or, if it was a word, the number of letters in said word). Eventually you would work your way to a single random fold, which would contain that most treasured of secret messages, the one that told you about yourself: You were strong and brave, or perhaps the smartest kid in the neighborhood. Or more likely, since it was fifth grade and you were just dipping your toe into the pool of sexual attraction, it would tell you who liked you, or who you were going to marry, or –most daring of all – who would let you kiss them were you only to ask.
I was thinking about cootie catchers this weekend, when Jaybird invited contributors and readers alike to take two separate personality tests, the Myers Briggs and the Enneagram. A lot of people took them, myself included, and it was a fun way to break the recent tedium of election bickering. Those of us that took them scanned the others’ scores, wondering who did or didn’t score exactly as might have been predicted, and whose score revealed an as-of-yet undiscovered intellectual soul mate. To me, these kinds of personality tests are the modern-day adult version of cootie catchers: they’re fun, they can break the ice and provide topics of conversation, and they allow us to focus on that most precious of topics, ourselves. The other and most important thing that personality tests have in common with cootie catchers is that they are terrible predictors for people’s behaviors, and provide no real insight save what we might glean from observing ourselves observe the results.
My first experience with the Myers Briggs was in the mid 1990s, when I needed to take (and, for lack of a better word, pass) the test as a condition of employment for a job. I was also required to undergo a handwriting analysis, which was a little worrying since I have pretty atrocious, doctor-prescription-like penmanship. But I passed each and was hired, and within a short time was heading up my department and reviewing other perspective employees’ Myers Brigg test results. (The absurd handwriting test was discarded shortly after my hire.) If I recall correctly, we paid about $350 for each Myers Briggs test administered. When I look back at the people we hired based on Myers Briggs scoring, I go back and forth on which part I find more amusing: the fact that the tests were such consistently terrible predictors of success, or the fact that my company clung so strongly to the belief that they were a necessary tool to building a stable, competent sales force despite all evidence to the contrary.
For a long while I wondered if reason for its failings was that the Myers Briggs was too easy to cheat. We were hiring salespeople. Most of the people applying really wanted to job. So perhaps, I theorized, they weren’t answering honestly and just telling the test what they thought it wanted to hear (in, ironically, pretty stereo-typical salesperson fashion). Now, however, I have come to believe that the entire concept is deeply flawed. Even if those taking the test are being honest, I believe its predictive capabilities rely on nothing more than random luck.
There are a number of reasons I think this, but the biggest is this: People are simply more complex than the prefabricated toy boxes the Myers Briggs-like tests wish to put us away in. As an example, let me take two questions from the tests Jaybird linked to, one from the Myers Briggs and one from the Enneagram:
I’ve been: (a) a bit cynical and skeptical, or (b) mushy and sentimental
YES or NO: You prefer meeting in small groups over interaction?with lots of people
Both of these question suffers from the same problem: each mistakenly assumes that people are divided along very neat dichotomies of black and white. These assumed dichotomies do allow us to process perceived patterns more easily. But they also force us into error, because those dichotomies are a little more than a beautiful lie. For myself, neither of these choices makes any sense, and I know whatever choice I make will lead to the test making an incorrect determination about my skills, comfort levels, and very essense.
The first question asks me, essentially, do I use reason or do I feel emotions; this is absurd because everyone I know does both. I might well choose either. In fact, whichever one I do choose probably says more about what meeting I just got out of or what book I just put down than it does who I am as a person.
The second question is even worse for someone like me. Because of the sentence structure, I am forced to answer “NO,” because I don’t prefer small groups to large ones – I like them both just fine. I could sit down and have a martini with the good Doctor Saunders or have pitchers of beer with the entire League staff, and both of those scenarios would be pretty space awesome. Neither makes me uncomfortable. But I am aware that if I answer “NO” the test is going to assume a lack of comfort with intimate inter-personal communication, and if I answer “YES” the test is going to assume a lack of comfort being in front of people. Neither of those things is remotely true. The truth is that I’m just more complex than that. Everybody is more complex than that.
Another problem with the tests is made obvious with questions such as these:
YES or NO: You readily help people while asking nothing in return
Much of my success has been: (a) due to my talent for making a favorable impression, or (b) achieved despite my lack of interest in developing “interpersonal skills”
Each of these questions relies on a high degree of self-awareness that far too many people lack. I suspect that almost everyone answers the first question “YES,” despite the obvious fact that under some circumstances almost everyone helps just for the sake of helping, and under most circumstances almost no one does. Similarly, I have done a lot of management and employee intervention work for clients over the years. And based on that experience, I can tell you that the more likely someone is to not question how awesome their “interpersonal skills” are, the more likely they are to be really terrible at playing with others. If there is a person out there that attributes his or her complete and constant inability to get along with most other people in a group for reasons other than “it’s the groups fault, I’m doing everything perfectly,” then I have yet to work with them.
Lastly, a huge issue I see with personality testing is the same one that I see in pop-psychology dream analysis books: It makes the error of assuming that words show up the same to all people under all circumstances. Above, I framed the question about small groups vs. large groups as to two competing social scenarios with my colleagues here at the League. It is important to note that not only is it possible that not everyone will frame it in a similar fashion, it’s quite possible that I won’t always frame it so. It depends entirely upon my mindset when taking the test. If I was taking the test having come out of a staff meeting, might I answer that question differently? Might I answer it differently if I was about to have a disciplinary talk with one of my children that had just been caught taking money out of my wallet without permission?
I thought about this yesterday while sitting in an airport in Denver, and it made me curious enough that I went back and quickly retook the Enneagram test an additional three times. The first was imagining that I was answering the questions as they pertained to managing my team at work, the second thinking of myself as a parent, and the third thinking about how I discuss difficult issues with my wife. My results for all four tests skewed like this:
Initial test: Dominant score in #3, The Achiever
Second test (thinking about managing people at work): High scores in both #3 The Achiever and #8 The Challenger
Third test (thinking about parenting): High scores both #1 The Reformer and #8 The Challenger
Fourth test (thinking about being a husband discussing difficult issues): High score both #9 The Peacemaker and #5 The Investigator
That’s a lot of different results. So who is the real Tod Kelly? Is he really the Achiever, the Challenger, the Reformer, the Peacemaker or the Investigator? Which of this really best describes him?
The answer, clearly, is all of them. We humans are, as I said above, far more complex and complicated than we like to imagine ourselves. We all have most or all of these various masks inside of us (and many more as well), and we slip back and forth between them so effortlessly that we’re not even aware that we’re doing so. I may well be an extrovert – (I am) – but I still spend huge chunks of time wanting to be happily alone with my thoughts. My sister is an extrovert as well, but I suspect for years she thought of herself as an introvert: If when we were growing up I was the universally identified extrovert, she was by default the introvert – except that she wasn’t, really. Not ever.
The company I work for now – the one I am retiring from next month – has a far better track record for hiring sales people than the one I worked for back in the 90s. In the past 15 years, they have hired just under 20, which is less than most of its competitors hire every two to three years. In that time, they have never had a salesperson or partner quit or retire (though that last bit is about to change), and in that time they have only ever felt compelled to terminate one salesperson. They are also the only firm in the area that does not use any form of Myers Brigg-like testing. I do not believe this is coincidental.
When my firm hires someone for a sales or executive position (which are the positions Myers Briggs are most often used for), the process is long – usually a couple of months or longer. There are many preliminary talks, and many of those are over lunches and dinners. By the time someone is made an offer, they have spent a lot of time with most of the partners and everyone who will be part of their team. By that time, everyone on both sides of the equations have a pretty good idea of how well all the pieces might fit together. When the other firms in our area decide they want to hire someone, they decide that they need to hire someone as soon as possible; they use Myers Briggs as a tool that allows them the luxury of not having to pay attention to whoever they are looking to hire. They want a spreadsheet to quickly, ridiculously and entirely quantify the very essence of someone they spent half an hour with a couple of days ago. But as I said earlier, believing you can do that with another human being is a beautiful lie we like to tell ourselves.
It tells us, in other words, far less about the potential people to be hired than it does about the companies looking to hire them.