The personality test's origin isn't a psychology lab. It's the U.S. Army in 1917, trying to screen recruits for what was then called "shell shock" risk before sending them into a war that had no real psychiatric triage system in place. Robert Woodworth's Personal Data Sheet — a list of yes/no questions about anxiety, nervous symptoms, and emotional stability — was built fast, for a specific wartime need, and its guiding idea (asking people direct questions and scoring the answers) turned out to outlast the war by about a century.
The mid-century split
Two very different traditions grew out of that early instinct. One, associated with Raymond Cattell's factor-analytic work in the 1940s and 50s, tried to let statistics reveal how many genuinely separate personality dimensions actually existed in people's self-descriptions, rather than assuming the number in advance. The other, arriving around the same period, took a different path entirely: Katharine Briggs and Isabel Myers built the Myers-Briggs Type Indicator directly on Carl Jung's theory of psychological types — a framework built from clinical intuition and philosophical argument, not from factor-analyzing large datasets of how people actually describe themselves.
That difference in origin matters more than it might seem. The factor-analytic tradition kept getting re-tested against new data, revised, and converged over decades toward the five-factor structure — Neuroticism, Extraversion, Openness, Agreeableness, Conscientiousness — that most trait researchers now treat as reasonably well-established. The MBTI, built on a fixed theoretical framework rather than derived from data and re-tested against it, didn't have the same built-in mechanism for revision — which is a meaningful part of why it remains popular despite persistently weak evidence for the four-letter types it produces.
Where the public-domain instrument comes in
By the 1990s, the five-factor model had a well-validated commercial instrument — the NEO-PI-R — but no public-domain equivalent, meaning any product or researcher who wanted to use it had to license it. Lewis Goldberg's International Personality Item Pool project, started in the mid-1990s, set out to fix that: a public-domain bank of personality items, free for anyone to use, that could measure the same well-established structure without a license. The IPIP-NEO-120 that Bearing runs on is a direct descendant of that project — four items per facet, thirty facets, built specifically so a product could use a validated five-factor instrument without either paying for the proprietary one or inventing a new, unvalidated one from scratch.
The throughline
A century of personality testing runs from an urgent wartime screening tool, through a genuine and consequential fork between statistically-derived and theory-derived models, to a public-domain item bank built explicitly to make good measurement available without a paywall. Most people's actual experience with personality tests — a workplace MBTI, a magazine quiz — sits on the less-validated branch of that fork. The five-factor model, and the public item pool that makes it accessible, is the other branch: less culturally famous, and considerably better supported by the evidence.