Self-Assessments: The Myth of Personality Types

Or: Mind the Bell-Curve

First off, here’s a fun article about the Myers Briggs (MBTI) that I wish I had written myself. It speaks from my heart.

But even apart from the Myers Briggs, any theory claiming that people come in distinct personality “types” (e.g., the “Eneagram”, “True Colors”, “Are you a dog or a cat person?”, etc.) has a very fundamental problem: none of those types make sense, for two simple reasons. (Geoffrey Miller explains them in more detail and eloquence in his book “Spent”, which I had reviewed earlier on this blog.)

1. Personality traits have been documented in a huge body of research. After decades of studies by a multitude of independent groups, and after many data-driven revisions of initial theories, one dominating model suggest that there are five distinct factors, also known as the “Big Five”. They have been labeled Openness to Experience, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. There’s also a newer model that finds six dimensions (adding the factor of Honesty/Humility to the other slightly modified five factors).

In those models, each of the dimensions is normally distributed, along a bell-curve (as is the case for  almost any other conceivable human trait). This means that most people find themselves somewhere in the middle of each of those dimensions, being moderately conscientious, agreeable, etc, with fewer people having extreme traits. The dichotomies of typologies (such as feeling vs thinking; or judging vs perceiving in the Myers Briggs) simply don’t make sense if the underlying traits have a bell-curve distribution.
The bell curve also explains why these tests are notoriously unreliable, meaning that most people fluctuate between different types if they take the tests repeatedly.

2. The five or six personality factors are statistically independent of each other. That is to say, knowing a person’s score on some of those factors gives you no information whatsoever about all the other aspects of their personality.

Together, points 1. and 2. are what statisticians call a multivariate normal distribution: each dimension shows a normal distribution with most people near the middle, and each dimension is independent of the others.

Together, they also tell us that distinct personality types are an illusion.

Why, then, are we so fascinated by them, and why do we find it so intriguing to be assigned to a specific type? I’m assuming it has to do with our talent for story-telling and pattern-seeking, but I would welcome other people’s thoughts on that topic.

With regard to the Myers Briggs, I also take issue with the idea that different people should be “thinking” vs “feeling”, or “sensing” vs “intuitive” types. I believe we all need to do all of the above, not either/or. But that would be a topic for a whole new post. In fact, it’s the topic of a whole book we already wrote.

by Ursina Teuscher (PhD), at Teuscher Decision Coaching, Portland OR