Methodology
Every challenge uses a fixed, versioned set of items and a simple scoring rule. Here is exactly how each one works.
Scoring rules (version 1)
- Reasoning and spatial: one point per correct item. Wrong or skipped items score zero. There is no negative marking and no conversion to an IQ number.
- Reaction time: the median of ten valid responses. Pressing before the cue is a false start. Times below 100 ms or above 2000 ms are flagged and excluded. After 20 attempts without ten valid ones, the session ends.
- Memory: one point for each sequence repeated in exactly the right order.
- Attention: hits, misses, false alarms and correct rejections, reported as balanced accuracy, with median response time on correct targets.
- Spot the difference: each changed region counts once. Wrong taps are recorded but not penalised.
- Colour, hearing and visual challenges: raw correct counts only. No diagnostic labels, hearing levels or acuity fractions.
- Multitasking: symbol task accuracy and memory task accuracy reported separately, plus joint accuracy.
Premium breakdowns
Premium shows the same session in more detail, such as accuracy by item type, response times trial by trial and which items you missed. It never changes how the basic score is calculated.
Score guide
The Stats tab places your latest result on a general score guide for that kind of test. For reaction time the bands follow typical browser-measured simple reaction times. For the other tests they are set by the share of items you got right. The guide helps you read your number. It is not a percentile, a population norm or a diagnosis.
Content status
The current items are original starter material for piloting. Difficulty labels are provisional until real data supports them.
Limits
Browser timing depends on your screen, input device and system. Colours and sound depend on your display and headphones. Nonverbal items reduce reading demands but do not remove cultural, educational, motor or device effects.