Science and transparency
What this test can tell you, and what it cannot
PersonalityVIEW is a self-report questionnaire about preferences. It reports which way you lean on four axes and how strongly you lean, and that is the whole of what it reports. This page sets out how the scoring works, where the limits are, what the AI coach does with what you tell it, and which numbers we have not published yet.
What the test measures
Four preference axes. Every question is a statement about you, and your answer is how far it fits, from strongly disagree to strongly agree. Each answer counts toward one side of one axis. At the end each axis carries two things: the side you lean toward, and how far from the middle you landed. Nothing else is inferred, and no question feeds an axis it does not belong to.
Orientation
Extraversion or Introversion
Where you tend to send your attention: outward, toward people and activity, or inward, toward reflection before acting.
Insights
Intuition or Sensing
How you tend to take information in: through patterns and what could follow, or through concrete detail and what is present.
Decisions
Thinking or Feeling
How you tend to reach a conclusion: by weighing the logic of it, or by weighing the values in it and the effect on people.
Style
Judging or Perceiving
How you tend to organise your life: with things settled in advance, or with the options deliberately kept open.
What it is not
These are not disclaimers added by a lawyer. They are the boundary of what the questionnaire is capable of.
It is not a clinical instrument. It does not screen for, diagnose or rule out any condition, and no result here should be read as a statement about your mental health.
It does not measure ability. A preference is not a skill. A leaning toward Introversion says nothing about how well you handle people, and a leaning toward Thinking says nothing about how clever you are.
It is not a reason to give anybody a job, or to refuse them one. It has not been validated for selection, it does not predict performance, and using it that way is a misuse of it.
No type is better than another. There is no ranking, no score to raise and nothing to pass or fail.
It is not permanent. Preferences move with age, role and circumstance, and a result is a description of a moment rather than a fixed fact about you.
Reliability figures
Not published yetWe have not published reliability statistics yet, and we are not going to publish figures we have not measured. They will appear here once 100 people have completed the 32-question Standard test, whatever the numbers say, together with the exact version of the question bank they were measured on.
What will appear here
- How many people's answers were measured.
- Which version of the question bank they answered.
- Internal consistency for each of the four axes, one figure per axis.
- The date the figures were produced, so you can judge how old they are.
Until then this section stays empty on purpose, rather than being filled with numbers borrowed from another instrument.
The questions people actually ask
Short answers, written from the code that scores the test rather than from a brochure.
How is my type decided?
Each answer adds to one side of one axis. At the end, each axis is a single total compared with the midpoint, and the side with more behind it gives the letter. The report also shows how far from the midpoint you landed, as a percentage, so a leaning of 52 percent is never dressed up as a leaning of 90 percent. The arithmetic is a sum, not a model: the same answers always produce the same result.
What happens if an axis comes out exactly even?
It is not rounded away in silence. When the two sides of an axis land on exactly the same total, a small reserved set of tie-break statements is consulted first, then a single anchor statement, then the pole your previous result gave for that axis. If none of those settle it, a fixed default is used, and that default is written down rather than improvised. However the tie was broken, the axis is reported as borderline and the split you see stays an honest fifty fifty.
What does borderline mean on my result?
It means the axis landed too close to the middle for the letter to be firm: the answers on that axis net out to less than about two thirds of the spread you would expect from someone with no preference at all, which is roughly one strongly answered statement on the twelve-question test and three on the seventy-two. The cutoff grows with the number of statements, so a short test is flagged more readily than a long one, as it should be. We label it rather than hide it. A borderline axis usually means you draw on both sides depending on the situation, which is more worth knowing than the letter that happened to win. It is also the axis most likely to come out the other way if you take the test again, and that is the questionnaire being honest rather than being unreliable.
What does "your answers look rushed" mean?
Half the statements on every axis are worded one way and half the other, so somebody who agrees with nearly all of them, or disagrees with nearly all of them, has agreed with statements that contradict each other. When that happens on most of a sitting, or on two or more axes, the result is shown with a note saying so. It still counts as your result. It is left out of the reliability figures on this page, because one rushed sitting lowers the consistency of every axis at once and says nothing about the questions.
Can an employer use this to hire me, or to turn me down?
No, and we ask that nobody try. The questionnaire is not validated for selection, it does not measure competence and it does not predict how anyone will do a job. It is built for reflection and for conversation. Using it to decide who gets hired, promoted or placed on a team is a misuse of it, whatever the four letters say.
Is the AI coach a therapist?
No. The coach is a language model. It is labelled as AI wherever it appears, and it is available to adults only. It reads two things: your own profile and the PersonalityVIEW content we have published. It does not diagnose, it does not treat, and it is not a crisis service. If you are struggling, please speak to a qualified professional or contact a crisis service where you live.
What happens to what I tell the coach?
Coach conversations are kept for 90 days and then removed. You can clear a conversation yourself at any time, and clearing it removes it. To produce a reply, your messages and a summary of your result are sent to the model provider. Please do not put anything into a coach conversation that you would not want stored or processed.
What do you hold about me, and how do I get it back or delete it?
Your account, your answers, your results and whatever you have chosen to make public. Your answers are never shown to another person: only the four-letter result and the strengths you pick can appear on a public profile, and only if you turn that on. The app exports everything held about you as a single file, and deleting your account deletes that data rather than hiding it. The privacy policy names every collection and every processor, one by one.
Are the famous people in the type profiles evidence?
No. Nobody on those lists took the test. They are editorial illustrations, picked to make a type easier to hold in mind, and they are not a claim about any real person's personality.
If something on this page is wrong, we want to know
This page is written from the code that scores the test and from the document that governs what we do with data. If you find a sentence here that does not match what the app actually does, that is a bug rather than marketing, and we would rather hear about it than have it sit here.
Corrections and privacy questions: privacy@personalityview.com