Wellington vs 16Personalities

What does 16Personalities measure?

16Personalities is a free online quiz that reports one of sixteen types written as a four-letter code, in the style made familiar by the Myers-Briggs Type Indicator, with a fifth letter (A or T) for what it calls Identity. Its makers describe the model as their own, drawing on trait theory and Big Five terminology rather than on the official MBTI instrument. It is quick, well designed and taken by a great many people, which is why it is often the first personality test someone meets.

The result is a type. Whatever your position on each scale, you land on one side of it, and the page you read describes the side. If you sit near the middle of a scale, as most people do on most scales, the description you get is a coin toss between two quite different portraits.

The fifth letter is the interesting one. Identity, reported as Assertive or Turbulent, tracks roughly the ground the Big Five calls Emotional Stability, the dimension the official MBTI leaves out entirely. Adding it was a genuine improvement on the four-letter format. It is still reported as a letter rather than a position.

What does Wellington measure?

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 questions and scores 6 broad dimensions: Extraversion, Agreeableness, Conscientiousness, Emotional Stability, Openness to Experience and Honesty-Humility. Each is a continuous score placed against a reference sample, reported as a percentile and a band, with a written page for the band you fall in. Nothing is rounded to a type. The Wellington Membership adds 84 traits underneath, including every Big Five facet and all 24 VIA character strengths, and grows the report toward all 255 traits over 104 short weekly modules.

The first ten questions are answered before anything is asked of you, so you can see what the statements are like before deciding to go on. After question ten you sign in with an email address, so your answers are saved and your results can be shown to you. No card is needed for the Snapshot. If the vocabulary is new, the Big Five glossary defines all six dimensions in plain words, and the Big Five explained is the longer version.

Why does types versus traits matter?

Personality traits are distributed roughly like height: most people are near the middle, and few are at either end. A type test draws a line through the middle and assigns everyone on each side the same label. Two people one point apart can receive opposite types, and the same person can change type on a retest a few weeks later. Studies of the MBTI, the model 16Personalities' format follows, found that its scales behave as continuous traits lining up with four of the Big Five rather than as true dichotomies (McCrae & Costa, 1989), and a review for practitioners concluded that a four-letter type formula does not support the inferences commonly drawn from it (Pittenger, 2005).

A trait score keeps the information the type throws away. If you are at the 55th percentile on Extraversion, Wellington tells you that, describes what a balanced position looks like, and does not pretend you are an extravert or an introvert.

The practical consequence is what you can do with the result. A percentile can be compared with your own score a year later, or with the facets beneath it, or with what the research says the trait predicts. A letter can only be compared with other letters. Reading a score well is a skill in itself, and how to read personality test scores walks through it.

Why does a type description feel so accurate?

Part of the answer is that the scales are real, so a description built on them will catch something true. The rest of the answer is older than any type test. In 1949 Bertram Forer gave his students what he presented as an individual personality sketch drawn from a test they had just taken. Every student received the same sketch, which he had assembled from a newsstand astrology book, and the class rated it as an excellent account of themselves (Forer, 1949).

The effect has been reproduced many times since. Statements that could fit almost anyone read as uncannily personal when they are mildly flattering, when they name both a strength and a matching doubt, and when the reader believes a rigorous process produced them. This is not a trick anyone is playing on purpose. It is a feature of how people read descriptions of themselves.

• Why types are exposed to it. One page per label means the page has to fit roughly one sixteenth of everybody, which forces it toward statements that fit everybody.
• Why percentiles are less exposed. A band page is written for a stated position on a stated scale, so it can say something that does not apply to people in the other bands.
• How to test any result. Read the description you did not get. If it also sounds like you, the description is doing the work, not the measurement.

How many questions does a reliable result need?

Reliability is not a matter of taste. The precision of a scale rises as you add items that tap the same trait, and trimming a questionnaire to save a minute costs more than most people expect. When very short Big Five measures were examined in two samples of employees and students, they understated how much the traits mattered for behaviour and inflated the apparent importance of rival constructs, with single-item measures the worst offenders (Credé et al., 2012). The authors' point was that even slightly longer scales improve validity substantially, at little cost in the respondent's time.

A sixty-question quiz spread across five scales leaves about a dozen items per scale. That is a workable number for a broad domain and nowhere near enough for the facets beneath it, which is why no short quiz can tell you whether your Conscientiousness is orderliness or industriousness. Wellington measures each Snapshot dimension with ten statements, then adds the Portrait for anyone who wants the facets: 90 traits with the membership rather than 6 domains.

Which is more accurate?

Accuracy in a personality test means two things: whether it gives the same answer when nothing has changed (reliability), and whether its scores relate to real behaviour and outcomes (validity). The Big Five model that Wellington is built on came out of decades of lexical research and survived repeated attempts to replace it (Goldberg, 1993), and the six-factor HEXACO version, which adds Honesty-Humility, has replicated across languages (Ashton & Lee, 2007). We are not aware of independent, peer-reviewed validation of the 16Personalities instrument itself; its site reports its own figures.

Wellington measures each trait with several questions rather than one, uses a 5-point accuracy scale, and reports where you sit relative to other adults. Our reference norms are provisional while our norming sample is completed, and the methods page says so plainly. The bands and the Portrait names exist to make the result easier to talk about; the percentile is the measurement, and it is always shown alongside. Wellington is a wellness tool for reflection, not a medical or psychological diagnosis, and it does not replace care from a qualified professional.

Neither we nor anyone else can tell you that a result is true of you in the way a blood test is true. What a good instrument can offer is a score that holds still when you do, a model that survived being tested in other languages and other decades, and a plain account of how confident it is. How accurate are personality tests sets out what that standard looks like across the whole field.

Why does Wellington use scores and bands instead of a letter?

Take the shape of the data first. Every scale behind the five letters is continuous, and a large sample piles up in a bell, so the middle holds most of the population and the extremes very few people.

That shape has no seam in it, so the place where a letter changes is set by whoever wrote the scoring. Rounding to a side discards how far out a person was, which matters most for the people the line runs closest to, whose letters can turn over between one sitting and the next. Scales of this format behave as continuous dimensions rather than distinct types (McCrae & Costa, 1989), along a trait structure built up over decades of descriptive research (Goldberg, 1993).

What Wellington does with that is report the percentile against a reference sample and wrap three bands around it, with the cut-offs written down: Quiet up to the 30th percentile, Balanced up to the 70th, High above. Those norms are provisional until the norming sample is complete.

The last step is a name, because a name travels and a percentile does not. The combination of bands is summarised as a named Portrait, chosen from the scores and printed with them rather than in place of them. The name is for practicality. The numbers are the measurement.

Side by side

Neither test diagnoses anything, and neither should be used to select people for jobs. Wellington never sells answers or profiles and its analytics never contain an answer; see the privacy policy.

If you have taken 16Personalities and enjoyed it, nothing here says you were wrong to. The types are a decent conversation and a poor instrument. Take the free Snapshot as the measured version of the same curiosity, and compare the two readings yourself. A wider field is covered in the best personality tests, compared.

Questions people ask

Is 16Personalities the same as the MBTI?

No. The official Myers-Briggs Type Indicator is a paid instrument from The Myers-Briggs Company, usually delivered through a certified practitioner. 16Personalities is a free online quiz that uses the same four-letter naming style but describes its model as its own, closer to trait theory, with a fifth letter the MBTI does not have. Both report a type rather than a position on each scale, which is the part that matters for what the result can tell you.

Does Wellington give me a type?

Not in the sixteen-boxes sense. Wellington gives you a percentile and a band on each of 6 dimensions in the free Snapshot, and on every trait you add after that. It also names your combination of bands, one of 243 Portraits, as a shareable summary. The bands and percentiles underneath are what the report is actually built on, so the name is a headline rather than the finding.

Can I take both?

Yes, and many people do. If you already have a 16Personalities type, the free Wellington Snapshot takes about 7 minutes and shows you the continuous scores behind the letters: how far above or below the middle you sit on each dimension, rather than which side of it you fell on. Comparing the two is a good way to see what a type rounds off.

Is Wellington free?

The Snapshot is free. You answer the first ten questions before anything is asked of you, then sign in with an email address so your answers are saved and your results can be shown to you, and no card is involved. The Wellington Membership is $4.99 a month, and adds the facets and the character strengths. You can cancel any time.

Why did my 16Personalities type change?

Almost certainly because you were near the middle on at least one of the five scales. Most people are, on most scales, and a letter has to fall one way or the other. A review of the same four-letter format concluded that a type formula cannot carry the inferences people draw from it (Pittenger, 2005). A percentile moves a few points in the same situation, which is a more honest picture of what changed.

Why do the type descriptions sound so right?

Partly because they are built on real trait differences, and partly because people rate generic descriptions of themselves as accurate even when everyone has been handed the same text (Forer, 1949). A quick check is to read the description for the type next to yours. If it also fits, the writing is doing more work than the measurement is.