What gets measured

Five short tasks, run back to back, each mapped to a well-established laboratory paradigm. The whole thing takes about five minutes.

🧠 Memory — a Corsi-style span task. Cells light up in sequence and you reproduce the order. The span grows from 3 to 8. Score is (span reached − 2) × 20, so a span of 5 lands at 50.
🎯 Focus — a Stroop task. You report the ink colour of a colour word while the word itself says something else. 30 seconds; roughly 25 correct is a strong result.
⚡ Speed — simple reaction time, three trials. 200ms maps to 100 points, 500ms to 0.
🧩 Logic — number sequences. Find the rule, give the next term. 30 seconds.
🧭 Spatial — mental rotation. Decide whether two shapes are rotations of each other or mirror images. 30 seconds.

The five sub-scores are averaged into a total and drawn as a radar chart, so the shape matters more than the number.

Why the shape matters more than the total

A total of 65 tells you very little. Whether it came from five scores near 65 or from a 90 and a 38 tells you a great deal — the second profile has an obvious place to start practising.

That said, be careful about reading too much into small gaps. A difference of ten points between two areas is well within what a bad night's sleep produces. Your report page deliberately refuses to label a strongest and weakest area when the spread is under 15 points, and says so.

⚠️ This is not an IQ test

Three things this cannot do, and it is worth being blunt about them.

It is not standardised. A real cognitive battery is normed on a large sample under controlled conditions. This runs on your phone, in whatever room you are in.
It measures a narrow slice. Five short tasks touch working memory, inhibition, processing speed and spatial reasoning. That leaves out language, long-term memory, reasoning under uncertainty, and most of what people mean by "intelligence".
It cannot diagnose anything. If you are worried about attention or memory, a clinician with a proper battery is the right answer, not a web page.

What it is good for: seeing your own relative strengths, and tracking how sleep and fatigue move your own numbers over time.

Getting a clean measurement

Do it once, properly. The first run is the honest one — after a few attempts you are partly measuring familiarity with the tasks.
Same time of day when you repeat it. Morning and 2am are different people.
No interruptions. A single notification during the Stroop block costs several points.
Use a device you can tap accurately on. Reaction time on a laggy touchscreen measures the touchscreen.

What to do with a low area

Cognitive ability does not move much in a few weeks — that part is well documented and the "brain training makes you smarter" claims are mostly weaker than advertised. What does improve quickly is familiarity with the task format, which is worth real points whenever you meet these tasks in a testing situation.

Memory — flash memory, N-back
Focus — Stroop, task switching
Speed — reaction time, quantity comparison
Logic — ball weights, tower puzzle
Spatial — shape rotation, path building

Frequently asked

Q. My score jumped 20 points between attempts.
A. Entirely normal. Sleep, caffeine, time of day and how distracted you are all move these tasks substantially. That variance is the reason single scores mean little.

Q. Is the memory task the same as a digit span?
A. Same idea, spatial version. Digit span uses numbers you rehearse verbally; this uses positions, which is closer to the Corsi block-tapping task.

Q. Where is my result saved?
A. In this browser only (localStorage). Nothing is sent to a server, so it will not appear on another device, and clearing browser data removes it.