Letters and colours

What this is, and what it is not

You will see each letter and digit three times, in a shuffled order, and pick a colour for it each time. You are not being asked to remember what you picked before — you are being asked again, and the measurement is how close the three answers land.

That is why the test works on one person. There is no group to compare you to and none is needed: the thing being measured is the agreement between your own answers.

This result will never be worded as: “You have synaesthesia.” — and it is not a diagnosis, does not clear you for anything, and buys you nothing here or anywhere.

Your key

Your browser has made you a random key. It is the only thing linking this sitting to your next one, and we store a one-way hash of it, never the key itself. Write it down — with it you can come back in a fortnight and see the pair, and you can delete everything. Without it we cannot, because we will have no way to know which entries were yours.

…

Already have one? Use that key

The test

108 trials, about 12–18 minutes. You can stop at any point; a partial sitting is scored on what you finished and says so.

Start Finish and score what I did

What this page cannot do

• Your display is not calibrated and cannot be.sRGB is a standard, not a description of your panel. Its actual primaries and transfer curve are unknown and unknowable from JavaScript, and auto-brightness means the display can change itself mid-session. Every exam here is therefore built as a within-person, within-display measurement.
• Your screen cannot show every colour.An sRGB panel cannot display a saturated spectral green; where a stimulus falls outside the triangle the browser silently clips it to the edge. Clipping is not noise — it is a systematic compression that makes different people’s settings look more similar than they are.
• The room you are in is part of the stimulus.A screen in a bright room has a grey veil over it: contrast falls, the black point rises, and your adaptation is set by the room rather than by anything we sent you.
• Your three answers all pass through the same wrong screen.Which is exactly why this exam survives an uncalibrated display: the score is a distance within one sitting on one monitor, so whatever your panel does to a colour, it does to all three of them. What it does not license is comparing the actual hues you picked with somebody else's.

Where the method comes from

• Eagleman et al. (2007), J Neurosci Methods 159(1):139–145 — the standardised battery.
• Rothen, Seth, Witzel & Ward (2013), J Neurosci Methods 215(1):156–160 — colour spaces compared.
• Carmichael et al. (2015), Consciousness and Cognition 33:375–385 — the failure modes.
• Simner et al. (2019), Atten Percept Psychophys — pairings can change in adulthood.

Rothen, Seth, Witzel & Ward (2013), J Neurosci Methods 215(1):156–160 give refined, space-specific cut-offs with reported sensitivity and specificity. They are NOT reproduced here because they have not been read off the paper in this build, and an invented threshold is worse than none. The score below stands on its own; it does not need a cut-off to be meaningful.

The rules this page is bound by

• Report the measurement, not the category."Your VVIQ score was 19 out of 80. Scores at or below 24 are the range in which most published aphantasia research recruits participants." Never "you have aphantasia."
• Name the reference class, every time.A percentile is meaningless without saying among whom. Ours is: among people who found this page and chose to take the test. That group is systematically enriched for exactly the traits being measured, because people who suspect something about themselves go looking.
• Consistency, not confidence.For every instrument the second administration is part of the instrument. A one-shot result is a reading, not a finding.
• Never route a result to anything consequential.No credential tier, no karma weight, no grant eligibility, no payment. The moment a score buys something, people optimise it and the data dies.