A Short History of Personality Testing: From Galton to Big Five

The history of personality testing is shorter than people expect and weirder than it sounds. It starts in a Victorian sitting room with a man counting adjectives in a dictionary. It runs through wartime conscription, a mother-daughter team with no psychology training, and a slow grind of statistical research that eventually produced something that actually works. This post is the short version.

Galton and the lexical hypothesis (1884)

Francis Galton was a Victorian polymath who liked to count things. In 1884 he published a short paper noting that the English language has thousands of words for describing personality, and proposing that those words could be a starting point for studying it ¹.

The idea — later called the "lexical hypothesis" — is simple: if a personality difference is important enough to matter in everyday life, the language will have evolved words for it. Counting the words and seeing how they cluster would reveal the structure of personality itself.

Galton did not get far with the project himself. The idea sat for fifty years before anyone took it seriously.

Allport and Odbert (1936)

Gordon Allport and Henry Odbert finally ran Galton's experiment. They pulled every personality-relevant English word out of an unabridged dictionary — about 18,000 of them — and started trying to organize them into groups ².

This was a slog. They had no computers and no statistical methods sophisticated enough to handle 18,000 variables. They produced lists, hand-grouped them, and left the next step to later researchers. The list itself became one of the most-cited documents in personality history.

Cattell and the sixteen factors (1940s-50s)

Raymond Cattell took the Allport-Odbert list, used early factor analysis to find the patterns, and concluded there were sixteen primary personality factors. He built an assessment around them — the 16PF — which is still in use ³.

Sixteen turned out to be too many. Later researchers, using cleaner factor-analytic methods on Cattell's data, found that the sixteen factors collapsed into a smaller number when the math was done right. But Cattell's work created the template every later model would follow: take a large set of trait descriptors, factor-analyze them, see what falls out.

Eysenck and the three giants (1950s-60s)

Hans Eysenck argued that the data really showed only three factors: Extraversion, Neuroticism, and Psychoticism ⁴. He built a different test — the Eysenck Personality Questionnaire — and ran a long, productive feud with Cattell over how many factors were correct.

Eysenck's three were closer to the modern Big Five than Cattell's sixteen. But the question of "how many factors" stayed open for another two decades.

The Myers-Briggs detour (1940s-80s)

Meanwhile, outside academic psychology, something different was happening.

Katharine Cook Briggs and her daughter Isabel Briggs Myers were not psychologists. They were a mother-daughter pair who became fascinated with Carl Jung's writing on psychological types. During World War II, they decided to build an assessment that would help women entering the workforce identify jobs that matched their type ⁵.

The MBTI was published in 1962. It used four either/or categories — Extraversion vs Introversion, Sensing vs iNtuition, Thinking vs Feeling, Judging vs Perceiving — to produce sixteen "types" like INTJ or ENFP.

Academic psychology never embraced it. The dichotomous structure forced continuous traits into either/or categories, the test-retest reliability was weak, and the predictive validity was modest at best ⁶. Pittenger's 1993 review in Review of Educational Research is the standard academic critique.

The popular world embraced it anyway. The MBTI is now one of the most widely used personality instruments in corporate training and self-help. The gap between its academic standing and its commercial success is one of the strangest features of the field.

The Big Five emerges (1980s-90s)

While the MBTI was conquering corporate America, academic researchers were converging on a different answer.

Through the 1980s, multiple research groups — working independently with different samples, different languages, and different methods — kept finding the same five factors when they factor-analyzed personality descriptors. Lewis Goldberg's 1990 paper consolidated the consensus and gave the model its standard form ⁷:

• Openness to Experience
• Conscientiousness
• Extraversion
• Agreeableness
• Neuroticism

The names varied (the OCEAN acronym came later) but the structure replicated across English, German, Dutch, Czech, Japanese, and dozens of other languages. That cross-cultural replication was the key. The Big Five was not a feature of English-language psychology. It looked like a feature of human personality itself.

By the late 1990s, the Big Five was the dominant model in academic personality research. It still is.

HEXACO and the sixth factor (2000s)

The Big Five story has one more chapter. In the early 2000s, Kibeom Lee and Michael Ashton ran the lexical analysis across more languages and on cleaner data. They consistently found a sixth factor the Big Five had folded partly into Agreeableness: Honesty-Humility ⁸.

The HEXACO model — six factors — is now the second major contender in academic personality research. It has not replaced the Big Five, but it is the standard option when researchers specifically want to measure ethical-domain traits.

Where we are now

The modern state of the field, in a paragraph:

The Big Five is the academic standard for measuring personality. HEXACO is the standard alternative for ethical-domain research. The MBTI is the popular standard with weak academic support. Every other assessment you have probably heard of — DiSC, Enneagram, StrengthsFinder — sits somewhere between MBTI and Big Five in terms of evidence base.

The history shows a slow, century-long process of throwing out what does not work and keeping what does. Galton had the right instinct. It just took until the 1990s to build the math and run the studies that turned the instinct into a model that actually predicts things.

Take the modern Big Five assessment (12 min, free) →

References