Our science staff is trying to lead a more scientific life in 2018. Throughout the week, we’ll be questioning whether some of our favorite habits and hobbies are based on junk science or real evidence. Here’s the first entry, on personality quizzes.
Another big part of my personality: I really like online personality quizzes. Maybe you could tell.
But I’ve never really taken these tests seriously. Not even the Myers-Briggs — a test that is frequently used in professional development and hiring settings and costs $50 to take online. ($55.94 with tax. I’m an ENTP.) Call me cynical. Call me a skeptic. Call me a Ravenclaw with a dash of Slytherin. The point is, I always regarded personality quizzes as strangely addictive horse hockey, good for trading memes with friends, excellent at consuming your cash (or your employer’s — sorry, Nate), but not much more. “Astrology for nerds,” I called it. And as my colleagues and I compiled a list of the junk science we were resolved to let go of in the new year, I fully expected to be writing about how I was going to stop taking these damn things.
Instead, I get to spend 2018 immersed in a new series of personality tests — ones that are actually evidence-based and scientifically sound. That’s because, while most of the personality tests shared around the internet are, indeed, bogus procrastination devices, there is a science to personality, and it’s something that researchers really can put into a quantified, testable format, said Simine Vazire, a psychology professor at the University of California, Davis.
The most popular — used by the vast majority of scientists who study personality — is called the Big Five, a system that organizes personality around five broad clusters of traits: extroversion, agreeableness, conscientiousness, neuroticism and openness to experience.
Those clusters were not randomly chosen. Instead, the categories stem from research that began in the 1920s and ’30s, when researchers first theorized that you might be able to figure out the anatomy of a personality by studying the words we used to describe what people are like. But it wasn’t until the 1970s and ’80s that scientists finally had enough computing power to test their hunches. Researchers took thousands of surveys about the words people used to describe themselves and others, applied factor analysis, and came up with five big themes the traits clustered around, according to Christopher Soto, a psychology professor at Colby College. (Some researchers use a similarly derived model that adds a sixth trait: honesty-humility.)
The idea behind the Big Five is that everyone’s personality has a little of all five trait groups. What the test does, essentially, is tell you where you fall on the spectrum of each of the clusters. Your results are based on comparing you to all the other humans who have taken the test. So, for instance, when I took the Big Five through a website run by Soto, I ended up in the 99th percentile for extroversion, the 58th percentile for agreeableness, the 29th percentile for conscientiousness, the 43rd percentile for neuroticism and the 99th percentile for openness to experience.
That result is a bit different from the results you get with most online personality tests, which tend to group people by type — you’re a Hufflepuff, or a Charlotte, or an ISFJ. This is one of the big problems with pop culture ideas of personality, from a scientific standpoint. They try to fit us all into a set of immutable types. “That’s why we don’t like Myers-Briggs,” Vazire said. “We shouldn’t be talking about types of people.” That’s because, like most things with humans, personality traits fall on a bell curve and most of us will be near the middle of that distribution. When you try to categorize people by type, you end up with a lot of people who are placed in boxes that seem far apart, but whose distribution of personality is actually pretty close to each other. “Types create more artificial boundaries, where most people are really close to the boundary line,” Vazire said. “That’s the nature of human difference.”
The Big Five also differs in the way it asks questions. With the Big Five, you get direct statements — I am a person who is outgoing and sociable — and you agree with that, or you disagree. Sometimes you’re given a spectrum of agreement to choose from — agree strongly, disagree moderately. You aren’t asked about hypothetical situations. You aren’t asked about which words you like best. You aren’t given five images of different sunsets and asked to pick which one best reveals your inner soul.
“People feel like those are magic,” Vazire said. “I don’t want to take that feeling away from people because it’s not really harmful. But it harkens back to Freudian ideas of unconscious. The better and more valid way is to ask you pretty transparent questions.” The Big Five, she told me, has produced results that can be shown to remain largely consistent across a person’s lifespan and that can be used to predict at least some part of a person’s likely academic achievement, dating choices and even future parenting behavior. It has also been validated cross-culturally to some extent, Soto told me. Although, to do that, researchers re-create the model from scratch, using dictionaries of local languages, and the fifth cluster — openness to new experiences in the English-language version aimed at Americans — is often something different in other countries, influenced by different cultural values.
But none of that scientific evidence does much to make the Big Five popular online. In fact, when personality scientists think about their pet peeves with online quizzes, they take themselves to task. “I think we feel like we’ve done such a bad job of marketing the scientifically valid stuff,” Vazire said. Their science resolution, she said, isn’t so much to get people to stop taking personality quizzes, but to get those people who love quizzes to transfer some of that enjoyment to the Big Five. That’s something Soto and his team have been working on — creating a Harry Potter version. Of course, because it’s the Big Five, Soto’s test doesn’t tell you an absolute personality “type.” Instead, it tells you how compatible you’d be with each of the four Hogwarts Houses. I’m 69 percent compatible with Slytherin, 44 percent compatible with Gryffindor and 43 percent compatible with Hufflepuff. And Ravenclaw? I’m 99 percent compatible.1
Turns out, sometimes, the scientifically valid answer isn’t so different from Buzzfeed’s.