Why Psychologists Use the Big Five Instead of Personality Types

Take a popular four-letter type quiz today and you might come out as an introvert who plans everything. Take it again in a month and there is a fair chance one of the letters has flipped. Nothing about you changed. What changed is which side of an invisible line you landed on.

That small, irritating experience is a good way into one of the clearest disagreements between popular psychology and research psychology. Outside universities, personality is mostly talked about in types. Inside them, it is almost always measured with five continuous dimensions known as the Big Five: openness, conscientiousness, extraversion, agreeableness and neuroticism (sometimes called emotional stability, scored the other way round). Here is why.

A theory built from a dictionary

The Big Five did not start with a theorist deciding what personality ought to be. It started with a bet about language, usually called the lexical hypothesis: if a difference between people matters, speakers will have invented words for it. So instead of asking an expert, you can ask the dictionary.

In 1990 the psychologist Lewis Goldberg published one of the most thorough versions of this exercise. In the first of three studies he took 1,431 English trait adjectives, grouped them into 75 clusters and analysed how people's ratings on them hung together. The same five-factor structure appeared in ten replications, each using a different statistical procedure. A second study, with 479 common terms, found the same five factors in self-ratings and in ratings of people by their peers. Crucially, none of the factors beyond the fifth held up across samples.

This matters because nobody designed the five in advance. They are what was left standing after the data had been sorted many different ways. That is a very different origin from a type system built from a theory and then turned into a questionnaire.

Boxes versus sliders

The deepest problem with types is not which boxes are used but the idea of boxes at all. A type system assumes people fall into distinct groups: you are either an introvert or an extravert, with a meaningful gap between the two.

When researchers look at the actual scores, that gap is usually missing. In 1989 Robert McCrae and Paul Costa gave the best-known four-letter type questionnaire to 468 adults aged 19 to 93, alongside a Big Five inventory. They found no support for the idea that it measures truly dichotomous preferences or distinct types. What it did measure, reasonably well, were four independent dimensions, and those dimensions lined up with four of the Big Five.

A much bigger picture came in 2012, when Nick Haslam, Elise Holland and Peter Kuppens reviewed the whole body of taxometric research, a family of methods designed specifically to tell whether a trait comes in categories or in degrees. The review covered 177 articles and more than 530,000 participants. Once weaker methods were accounted for, the estimated rate of genuinely categorical findings was about 14%, and the domain of normal personality produced little persuasive evidence of categories at all.

If most traits are dimensions, the retest problem follows almost automatically. Scores on something like extraversion bunch up in the middle, with fewer people at the extremes. A type test draws a line through the busiest part of that distribution. Anyone sitting near the line can cross it because of a bad night's sleep or a slightly different reading of one question. Sorting people into two boxes also throws information away: someone just past the cut-off gets the same letter as someone at the far end of the scale.

Does it predict anything?

A description of personality is only useful if it tells you something beyond itself. Here the Big Five has a substantial record.

In 2007 Brent Roberts and colleagues reviewed prospective longitudinal studies, which follow people over years, and compared personality traits with socioeconomic status and cognitive ability as predictors of three hard outcomes: mortality, divorce and occupational attainment. The effects of personality traits were indistinguishable in size from the effects of the other two.

Psychology has had a difficult decade with replication, so it is fair to ask whether such findings hold up. In 2019 Christopher Soto ran preregistered, large-sample replications of 78 published links between the Big Five and life outcomes. 87% of them were statistically significant in the expected direction. The effects were also typically smaller than first reported, about 77% as strong, which is an honest reminder that original studies tend to overestimate.

None of this makes the effects large for any single person. Traits shift probabilities across many people. They do not write your biography.

Are types completely dead?

Not quite, and the honest version of this story includes the exceptions. In 2018 a team led by Martin Gerlach analysed Big Five data from more than 1.5 million people and reported evidence for at least four clusters where people were more densely packed than expected. A critique in the same journal the following year argued that those clusters were neither as robust nor as exhaustive as claimed. Either way, these are clusters found inside dimensional data, not the fixed boxes of popular type systems.

The Big Five has limits of its own. It is mostly measured through self-report, so it captures how people see themselves. The original lexical work was done largely in English, and when researchers repeated it in other languages, some found six factors rather than five. In 2004 Michael Ashton and colleagues reported a similar six-factor solution in seven languages, adding a dimension called honesty-humility. That is why some researchers now prefer a six-factor model, and why you will find a separate honesty-humility test on this site.

What this means for you

If you enjoy your four letters, there is no need to throw them away. Just read them as rough positions on sliders rather than as a verdict. The useful question is not which box you are in, but how far along each scale you sit and whether that position holds up in different situations.

• Look at the strength, not just the letter. A result near the middle means you are flexible or the test is unsure, not that you belong firmly to one side.
• Retake it after a few weeks. If a letter flips, that is information about how close you are to the line.
• Compare it with a dimensional measure. Our Big Five test scores each trait as a continuous percentage, and our sixteen types test deliberately shows the four axes as sliders so you can see how firmly you hold each letter.
• Ask someone who knows you. Self-report is only one view. How others see you is a second, independent source.

Types are a tidy story. Dimensions are a less satisfying story that happens to fit the data better, and that trade-off is exactly why researchers chose them.