A personality test asks you questions and composes an answer from what you ticked. That is its strength — fast, repeatable, cheap — and its fundamental limit: it can only ever measure self-report.
What the common instruments actually measure
MBTI
Sixteen types across four axes. Popular in organisations, widely criticised in psychometrics: test–retest reliability is weak — a substantial share of respondents change type within five weeks — and the axes are dichotomous while the traits they measure are continuous. Nobody is introverted or extraverted; everyone sits somewhere on a gradient.
Big Five
Five continuous dimensions: openness, conscientiousness, extraversion, agreeableness, neuroticism. It is the most robust model personality psychology has, validated across dozens of languages and cultures. It remains self-report.
Free online quizzes
Most have no validation whatsoever. They are built to produce a flattering, shareable result. The mechanism is the Barnum effect: a sufficiently general statement — "you have a need for recognition that you rarely show" — is recognised as accurate by almost everyone.
Three biases no questionnaire escapes
- Social desirability. People answer what is acceptable, usually without noticing.
- Limited introspection. A significant share of our preferences is not consciously accessible. We produce explanations after the fact, and we believe them.
- Question framing. Rewording a question shifts the answers. The test then measures its own drafting.
These are not design flaws. They are inherent to asking. Better questions do not remove them — they move them.
Measuring behaviour instead of declaration
Experimental psychology has long relied on indirect measures: reaction times, gaze paths, interpersonal distance, voluntary exposure time. These have a decisive advantage — they are hard to fake, because the subject does not know what is being observed.
Occupation of space is among the most legible. The distance you keep from an object, the time you give it for no reason, what you return to after seeing everything else: none of this passes through language, and therefore none of it passes through self-image.
What RÉSONANCE does
RÉSONANCE is a chamber. Twelve steles, a floor of black water, light falling from above. No question is asked, no instruction given.
What is measured: where you go first, how long you stay, how close you stand, what you avoid, what you come back to. Fifty-seven variables follow — not labels but degrees: need for withdrawal, sensitivity to light, tolerance for noise, relationship to material.
The result is not a type. It is a continuous profile, and a universe of presences derived from it — some of which contradict you deliberately, because a profile that only shows you yourself teaches you nothing.