Two kinds of test are everywhere online and both are routinely oversold: colour vision screening, and quizzes that assign you a number.
Neither is worthless. Both become useful the moment you know what they can and cannot establish.
Why online Ishihara plates do not work
The Ishihara test is the dotted circles with a number hidden inside. It works because the dots use specific printed pigments whose relationship to each other is precise, viewed under controlled lighting.
Reproduce that on a screen and several things go wrong at once. Your display has its own colour gamut. Your brightness setting changes the relationship between the colours. Your operating system may be applying a colour profile, a night-shift filter, or both. The image itself has been through JPEG compression, which is specifically designed to discard colour detail.
The result is a test where a person with normal vision can fail and a person with a deficiency can pass. It is measuring the monitor as much as the eye.
What does work on a screen
Hue arrangement — putting colours in order so each sits between its neighbours. This is the principle behind the Farnsworth–Munsell tests, and it degrades far more gracefully.
The reason is that it asks about relationships between colours shown together rather than about absolute colour values. A screen that shifts every hue slightly shifts them all together, and the ordering task survives it.
The score is a total error: for each cap, how far its position is from where it should sit relative to its neighbours. Zero is perfect, and because the caps form a ring, starting at a different point or going round the other way is not an error.
What a good score does and does not mean
It suggests your colour discrimination is in the usual range. It does not tell you which type of deficiency you might have — separating protanopia from deuteranopia requires specific confusion axes tested under known conditions, which in practice means an anomaloscope in a clinic.
A poor score has an obvious first explanation worth eliminating: screen brightness. Turn it up, try again, and only then consider seeing an optometrist.
Quizzes that assign you an age
These work by attaching a number to each answer and averaging them. That is the entire mechanism, and stating it plainly is the difference between a bit of fun and an implied claim.
"Mental age" was a real concept, introduced by Alfred Binet in 1905 to identify French schoolchildren who needed extra help. William Stern turned it into a ratio — mental age divided by actual age, times 100 — which is where IQ came from.
It was abandoned for adults because it stops distinguishing anything. Cognitive development plateaus, so two adults twenty years apart answering identically have no meaningful difference in "mental age" to report. Modern assessments score against an age-matched group instead, which is a different thing.
Binet, for what it is worth, objected to the use his test was quickly put to. He considered it a snapshot of a moment rather than a measure of fixed capacity.
Questions people ask
Should I trust an online colour test at all? As a prompt to get a proper one, yes. As a diagnosis, no.
Why does my result change between attempts? Screen brightness, ambient light, and practice. All three move the number.
Are personality quizzes measuring anything? They measure which options you chose. Whether that maps to anything real is a separate question, and for most online quizzes the answer is no.
What is a real cognitive assessment like? Several hours with a psychologist, standardised instruments, and a report. Not eight questions on a web page.
The colour blindness test uses hue arrangement rather than reproduced plates, and the mental age test says on the page exactly what it is doing. For screen questions specifically, the monitor test and colour contrast checker are the practical ones.