The 16 personality types
Four preference pairs, sixteen combinations. Here is how the codes are built, what each one describes, and how much weight the model can actually carry.
A four-letter personality type is a summary of four choices. Each choice is between two ways of doing something ordinary — where attention goes, what kind of information registers first, what settles a close call, and whether the outside world is preferred settled or open. Two options in each of four pairs produce sixteen combinations, and each combination gets a code and a nickname.
What the code is good at is giving a name to a pattern you have probably already noticed in yourself. What it is not good at is precision. The pairs are continuums, most people sit somewhere along them rather than at one end, and a code has to round each one to a side — so two people with the same four letters can be strikingly unalike, and one answer changed on a near-even pair produces a different type entirely.
The pages below take each type seriously without pretending the sorting is exact: what the letters mean in practice, the function stack underneath them, where the pattern helps and where it costs, which types it gets confused with, and which popular claims about it are simply wrong.
The four preference pairs
Every code is one letter from each of these four pairs — 2 × 2 × 2 × 2 = 16. Any two types therefore differ by between one and four letters, and the ones that differ by a single letter are the ones people most often confuse.
Engagement is generative. Thinking happens partly out loud, and contact with people and events tops the tank back up.
Depth is generative. Thinking happens inwardly first, and unbroken quiet is what restores capacity.
Concrete information registers first — what is actually here, what was actually said, what has actually worked before.
Pattern registers first — what this resembles, what it implies, where it appears to be heading.
Close calls are settled by consistency — does the reasoning hold, and would it hold for anyone else in the same position?
Close calls are settled by values and effect — what matters here, and what will this do to the people inside it?
The outer world is preferred settled. Deciding brings relief; open questions carry a low background cost.
The outer world is preferred open. New information is welcome late, and early commitment feels like a door closing.
All sixteen, grouped by family
The middle two letters — how information is taken in, and how decisions are made — do most of the work of grouping the sixteen. Types that share both tend to recognise each other quickly, whatever else they differ on.
Grouped with the other Intuitive-Feeling types, this is a temperament organised around meaning. Work that is competent but hollow tends to be quietly unbearable, and the question of whether something matters usually gets asked before the question of whether it is efficient.
Shared preferences: Intuition and Feeling. What they have in common is the question they ask first — what does this mean, and who is it for.
Grouped with the other Intuitive-Thinking types, this is a temperament organised around competence and coherence. Understanding a system properly is treated as worth doing for its own sake, and an argument from authority alone lands as no argument at all.
Shared preferences: Intuition and Thinking. What they have in common is a demand that an explanation actually hold together before it is accepted.
Grouped with the other Sensing-Judging types, this is a temperament organised around responsibility and continuity. Things that other people assume simply work usually work because someone with this temperament is maintaining them, and that maintenance is rarely made visible.
Shared preferences: Sensing and Judging. What they have in common is that they finish things, and that the finishing is usually invisible.
Grouped with the other Sensing-Perceiving types, this is a temperament organised around immediacy and skill. Competence is proven by doing rather than by describing, and a long theoretical preamble before a task gets started is genuinely costly.
Shared preferences: Sensing and Perceiving. What they have in common is a bias toward the live situation over the description of it.
How to use a type — and how not to
Sorting people into sixteen boxes is the weakest part of this model, and it is worth knowing exactly where the weakness is before you lean on the result.
The four underlying dimensions hold up considerably better than the sixteen labels built on them. If one of your pairs came out near even, treat every conclusion resting on that letter as provisional — which is why every result on this site shows how firmly each axis leaned.
The useful sentence is 'I tend to think before I speak, and it costs me in fast meetings' — not 'I am an INTJ, so I cannot do that.' A type is a description of a preference, not a limit on behaviour, and using it as an excuse is the most common way it does harm.
Four-letter type has poor evidence as a predictor of job performance, and the publisher of the original instrument says it should not be used for hiring. It can support a conversation about working style. It cannot support a decision about someone's livelihood.
Type says nothing about ability, values, mental health, motivation or how someone was raised. Two people with identical codes can differ on every one of those, and every one of them will explain more about a working relationship than the letters do.
Where to go next
Questions about the 16 types
Where do the 16 personality types come from?
From Carl Jung's Psychological Types (1921), which proposed that people habitually favour certain modes of perceiving and judging. Katharine Cook Briggs and Isabel Briggs Myers turned that theory into a sortable questionnaire during the 1940s, adding the fourth pair (Judging/Perceiving) that Jung had not framed as a preference. Every modern four-letter system descends from that lineage, including this one.
How are the 16 types made?
Four preference pairs, two options each: 2 × 2 × 2 × 2 = 16. Every type is one choice from each pair, which is why any two of the sixteen differ by between one and four letters — and why the types one letter apart are the ones people most often confuse.
How accurate are personality types?
Sorting people into sixteen boxes is the weak part of the model. Reviews of four-letter typologies report that a substantial share of people receive a different code on retest weeks later, because a preference measured near the middle has to be rounded to one side. The four underlying dimensions themselves hold up considerably better than the labels built from them — which is why every result on this site shows how firmly each axis leaned rather than only the code.
Which personality type is the best or the rarest?
Neither question has a defensible answer here. No type is better: the letters describe preference, not competence or maturity, and every type contains excellent and mediocre practitioners of everything. As for rarity, published frequency estimates come from self-selected or dated samples and disagree with each other, so we do not print percentages.
Is this the MBTI?
No. MBTI is a trademarked commercial instrument from The Myers-Briggs Company, usually administered through a certified practitioner. This is an independent twelve-question indicator built on the same four preference pairs. It is free, it is not affiliated with them, and it does not reproduce their items.
Should personality type be used for hiring?
No, and the instrument's own publisher says so. Four-letter type has poor evidence as a predictor of job performance, and using it to select or screen candidates is both unreliable and unfair to them. It is a reflection tool — useful for a conversation about working style, useless as a filter.
Where this framework comes from
- [1]Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal, 57(3), 210–221. ↗
A standard review of the reliability and validity limits of four-letter type sorting. It is why this test is presented as reflection rather than measurement, and why the result page marks any axis that landed on a tie.
- [2]The Myers-Briggs Company — the official MBTI instrument ↗
For disambiguation: MBTI is a trademarked commercial instrument administered by certified practitioners. This test is not it, is not affiliated with it, and does not reproduce its items.
Which of the sixteen sounds like you?
Take the free five-minute test — no account, no email, and your answers stay on your device.
Start free test →