DEV Community

Lucian (LKB)
Lucian (LKB)

Posted on Originally published at lkforge.com

What the Average Big Five Result Actually Looks Like (n = 19,718)

Personality tests love to hand you a tidy "type". So we took 19,718 real Big Five responses — the IPIP-50 instrument, from the Open-Source Psychometrics Project — scored them into the five OCEAN traits, and looked at what the distribution actually says.

Nobody is average on everything

The first result is the bluntest one: 0 of 19,718 people scored in the top band on all five traits at once. Not a single person. "Average across the board" is a statistical fiction — real people are high on some traits, low on others, and middling on the rest. In fact 46% land in the middle band on three or more traits, which is why the confident one-word "type" a quiz gives you throws away most of the picture.

The traits barely relate to each other

The Big Five is designed so the five traits are close to independent, and the data agrees. The strongest correlation between any two traits is r = 0.33 (Extraversion and Agreeableness); across all ten pairs the average is |r| = 0.17. That's weak. Knowing someone is extraverted tells you almost nothing about whether they're conscientious or open. The common intuition — "she's outgoing, so she's probably organized too" — simply isn't in the numbers.

Which traits skew, and which spread

Not every trait is shaped the same. Openness and Agreeableness skew high (most people rate themselves above the midpoint), while Extraversion and Neuroticism have the widest spread — people are all over the range. So an "average" score on Openness is genuinely common, but an "average" score on Extraversion is almost a coin flip between reserved and outgoing.

The honest caveat

This is a self-selected online sample — people who chose to take a personality test. That biases the absolute means (the where of each peak), so don't read the exact averages as the human baseline. What it doesn't undermine are the shapes and the correlations: the near-independence of the traits and the near-total absence of "top on everything" people are robust to who showed up.

Try it / reproduce it

The free Big Five test uses the same IPIP-50 instrument, runs entirely in your browser, and asks for no signup. The full distribution study — every number above, plus the correlation matrix and per-trait histograms — is reproducible from a public script.

Originally published on LK Forge.

Top comments (0)