If the Shoe Overfits… Part One
August 22nd, 2008(This is part one of a two-part series on criticism of MBTI personality theory. In part one, I set the stage and cover some basic criticisms. In part two, I’ll cover some in-depth criticisms and wrap things up.)
I’ve written about the Myers-Briggs (MBTI) personality test in this blog in the past here, and in two parts here and here. As such, this post assumes you have a basic knowledge of MBTI theory. If you’re not familiar with MBTI tests, that post is a decent place to start.)
I’ve read a large number of books about MBTI personality theory, mostly on the Keirsey variant championed by psychologist David Keirsey. (I’m starting to think I should start my own book club or something.) I’ve read hundreds of emails on the INFJ (a particular MBTI type) mailing list. And I even went as far as to take the MBTI Step II test at my own expense. I’m both interested in and a fan of MBTI testing in general.
But something’s always bothered me about it. In the first paragraph of my first post on MBTI I said, “…most of the [personality] tests seemed like pseudoscience gobbledygook to me…so for years I never paid them any mind.” Ever since then I’ve always had my doubts on how accurate and reliable personality tests were. After much reading and deliberation, I eventually settled on MBTI as being the best, but even then, I only accepted the Keirsey temperament version of the test.
And almost from the get-go, I mostly only spoke about Keirsey’s four main temperaments - “Artisans”, “Guardians”, “Idealists”, and “Rationals” - as opposed to the full 16-type treatment. Four dimensions of 16 personality types, to me, has always seemed way too specific for something as malleable and vague as someone’s personality.
Fortunately, if you read Keirsey’s website (and his books), he seems to feel much the same way I do. On his site, the first thing you’ll probably see is a succinct summary of his four temperaments. And here’s a quote from Keirsey himself in “Please Understand Me II”, the updated version of his seminal book on temperament theory: “I soon found it convenient and useful to partition Myers’s sixteen types into four groups…” I’m more inclined to agree with him and the four types he saw (”Artisans”, “Guardians”, “Idealists”, and “Rationals”) rather than with Myers’s four types (which are somewhat different).
For those who don’t already know, Keirsey’s four types are easily explained by example. Here’s what I wrote on this topic earlier (”People Patterns” being an introductory book on temperament theory):
“The best example in ‘People Patterns’ is from The Wizard of Oz. The Cowardly Lion is the Artisan looking for courage, Dorothy is the Guardian looking to go home, the Tin Man is an Idealist looking for a heart, and the Scarecrow is a Rational looking for a brain. Whichever character you identify with most (if any) is probably your Keirsey temperament.
If you need a more contemporary example, you can use the houses in Harry Potter. Slytherin are Artisans, Hufflepuff are Guardians, Gryffindor are Idealists, and Ravenclaw are Rationals.”
And thus those are the four basic personality temperaments. (Unless you think “The Wizard of Oz” is secretly about William Jennings Bryan or something.) You’ll notice that Keirsey (as well as Myers and Briggs) studiously avoids any perjorative judgments about personality types in most of his writings. They are all usually described in neutral or even optimistic terms. This is because MBTI theory is more of a tool for personal exploration than evaluation. Also, I think having negative claims about certain personality types is both presumptuous and counter-productive. (I think Keirsey would agree.)
However, are personality tests like horoscopes, then? Are MBTI personality types so vague and optimistic that they suffer from the “Forer effect”? (The Forer effect is the semi-well-known phenomenon of people tending to identify with vague generalities if for some reason they believe these generalities apply specifically to them, especially if these things are positive.)
In my experience with MBTI tests this hasn’t been true. I feel pretty strongly about being a “Rational” and reading about the other types just reinforces this in my mind. Here’s what Wikipedia has to say:
“Skeptics criticize the terminology of the MBTI as being so vague as to allow any kind of behavior to fit any personality type, resulting in the Forer effect, where an individual gives a high rating to a positive description that supposedly applies specifically to them[22][23]. However, the descriptions offered for the 16 psychological types are often quite detailed and specific, unlike the ‘vague and general personality descriptions’[28] that characterize the Forer effect. For example, David Keirsey[29] examined how the four temperaments differ in terms of language use, intellectual orientation, educational and vocational interests, social orientation, self image, personal values, social roles and even characteristic hand gestures. Keirsey went on to describe the hierarchy of intellectual roles played by each of the four types within each temperament, resulting in sixteen unique descriptions which, unlike the Forer effect, rely not on the universal traits that make human beings the same, but on the specific traits that make human beings different from one another.”
What do people who take the test (other than me) have to say about it, then? From the same Wikipedia article:
“When people are asked to compare their preferred type to that assigned by the MBTI, only half of people pick the same profile.[26] Critics also argue that the MBTI lacks falsifiability, which can cause confirmation bias in the interpretation of results.”
That’s not a great statistic. However, that’s for people testing for the same 4-letter type as before (i.e. one of the 16 personality types). My argument, though, is that 16 types is too many. More on that later. As for confirmation bias, that’s a common criticism, and one not easily dismissed.
Confirmation bias is the simple idea that we are much more likely to observe and pick out evidence supporting our views rather than evidence that suggests otherwise. And it’s pervasive. One particularly compelling book about confirmation bias is Nassim Taleb’s “The Black Swan” (and, to a lesser extent, his other book “Fooled by Randomness”). One of the biggest themes of “The Black Swan” is that confirmation bias is harmful and widespread.
It’s obvious I take Taleb seriously since I wrote a post on his book and I include him in my sidebar. My response to him, in the case of personality tests, would be that I, like many others, am often a victim of confirmation bias. But I’m aware of that fact and I try to stay vigilant. The fact that MBTI theory usually never strays into pejorative territory will always make its personality profiles prone to confirmation bias (since you’re always much more likely to identify with neutral and positive statements about yourself). However, as I’ve said, I think the benefits of this approach outweigh the negatives, especially if you’re on watch for confirmation bias. (Is this statement itself a product of confirmation bias? Wait, that’s too skeptical even for me…)
I should also mention that the MBTI test is boolean, or as they call it in the jargon, “modal”. That means each question on MBTI tests has only 2 answers, each of which tests one of the four dimensions of personality. While this approach often feels constricting and forced (it doesn’t allow for shades of gray on particular questions), modal tests should suffer less from the Forer effect, since they force you into clear yes-and-no answers.
There’s something about the MBTI test that concerns me even more than confirmation bias, though. Since this post is getting a bit lengthy, I’ll cover that next time. Stay tuned!
| | | del.icio.us |