What is a percentile in a personality test?

What does a percentile actually mean?

A percentile tells you where one score sits inside a distribution of other people's scores. If your Conscientiousness score puts you at the 62nd percentile, 62 percent of the reference sample scored lower than you did on the same measure, and about 38 percent scored higher. It is a statement about rank, not size: a 62nd percentile does not mean you are 62 percent conscientious, because Conscientiousness has no ceiling for 62 percent to be a share of.

Percentiles run from close to 0 to close to 100, and the midpoint sits around the 50th percentile, where half the sample scored higher and half scored lower. Most people cluster near that midpoint, because most personality traits are roughly bell-shaped, and a percentile far from 50 becomes steadily rarer the further it climbs or falls.

Is a percentile the same as percent correct?

No, and confusing the two is the most common misreading of a personality result. Percent correct describes a test with right and wrong answers, such as a maths exam, where 62 percent correct means you answered 62 out of 100 items the way the answer key wanted. A personality statement such as "I keep my belongings tidy" has no correct answer. Agreeing or disagreeing is simply a report of how you see yourself, and the test turns that agreement into a number rather than marking it right or wrong.

A percentile describes where that number lands next to other people's numbers, a different kind of comparison entirely. A high percentile is not a high mark and a low percentile is not a low mark. Both are positions in a line, and the line has no answer key.

Why does a percentile need a reference sample, and why does the sample matter?

A percentile only exists once you have chosen who to compare a score against, because the number is entirely relative to that group. The same raw score can sit at the 40th percentile against one sample and the 60th against another, and neither figure is wrong. They answer two different questions: where you stand among this group, and where you stand among that one.

Real differences exist between the people who might make up a reference sample, so the choice matters. Average levels of Conscientiousness and Agreeableness rise across early and middle adulthood, and in one large sample Neuroticism fell with age for women but stayed roughly flat for men (Srivastava et al., 2003). A reference sample built mostly from one age group, or skewed toward one gender, will describe someone outside that group less well.

Averages also differ by where people live. A survey of 17,837 people across 56 nations found that respondents in South America and East Asia differed from other regions on Openness, though the authors caution against turning that into a national league table (Schmitt et al., 2007). A percentile built from a sample in one country says less about someone answering from a different one, which is why what a norm group is and how it was assembled is worth checking before you put weight on a single percentile.

Why can a percentile move if you retake the test?

Every score is an estimate, not a perfectly exact reading. Classical test theory treats an observed score as a true score plus measurement error from wording, mood, attention and which particular statements happened to be asked (Nunnally & Bernstein, 1994). That error is what makes a percentile wobble a little from one sitting to the next, even when nothing about the person answering has changed.

The error shrinks as a scale asks more statements about the same trait. Very short measures, especially single-item ones, understate how much a trait matters and overstate the importance of whatever is measured alongside it (Credé et al., 2012). A percentile of 68 in March and 74 in June is more likely the same person on two ordinary days than two different people.

This matters most right at a band line. If a percentile reads 29 in one sitting and 32 in the next, the sensible reading is measurement noise crossing a boundary, not a change in the trait itself. A move worth taking seriously is a large one, seen more than once, on a trait built from several statements. How to read your personality test scores goes through what a genuine change looks like.

What do percentile ranges and bands look like?

Reports usually group percentiles into bands, because talking about "the 74th percentile" is more precision than most conversations need. The table below gives one common way of reading percentile ranges, alongside Wellington's three bands, which use the same cut-offs under simpler names.

The percentile is the exact measurement. A band is a rounder way of saying roughly the same thing, and a percentile just inside one band can read almost identically to one just inside its neighbour. The methodology page sets out how a raw score becomes a percentile and how the band cut-offs were chosen.

How does Wellington report a percentile?

Wellington is a science-based personality assessment from Therabot Labs LLC. It measures personality as continuous traits, built on the Big Five and HEXACO models, and writes the results back as a personal report rather than a type.

The free Snapshot asks 60 statements across six dimensions. Each dimension is the average of its own statements, placed against a reference sample of adults and reported as a percentile. Every trait is reported as a percentile against the reference sample. The three bands (Quiet below the 30th percentile, Balanced to the 70th, High above) and the Portrait names exist to make the result easier to talk about; the percentile is the measurement, and it is always shown alongside. Our norms are provisional while the norming sample is completed.

The Wellington Membership extends the same scoring to 90 traits, including every Big Five facet, so a percentile on a broad dimension can be read alongside the percentiles on the facets beneath it. Wellington is a wellness tool for reflection, not a medical or psychological diagnosis, and it does not replace care from a qualified professional. It is not designed for hiring, and you can export or delete your data at any time.

Questions people ask

What is a good percentile on a personality test?

There is no good percentile, because personality statements have no correct answer for a percentile to score you against. A high percentile means you sit further from the average on that trait, not that you performed well. Some traits relate to outcomes many people want, but the same trait carries real costs at its other end, so read a percentile as a description of your position rather than a grade.

Why do I get a different percentile from two personality tests?

Different tests almost always compare you with different reference samples, gathered from different groups of people, so the same underlying trait can produce two different percentiles. A shorter test also carries more measurement error than a longer one, which adds further wobble. Check which sample each test used before treating either number as the final word.

Does a percentile mean I answered a certain share of questions correctly?

No. Personality statements have no correct answer, so a percentile is not a score out of the number of questions asked. It compares the average of your answers with the same average taken from a reference sample, and reports what share of that sample scored lower than you, nothing about accuracy.

Can my percentile change if I retake the test later?

Yes, a little, even with no real change in you. Measurement error moves a score by a few points from one sitting to the next, and a percentile sitting near a band line is the one most likely to shift across it. A change worth trusting is a larger one that shows up again on a later retake, not a few points either way.