A test measures
Fixed rules, fixed length, no adaptive help, run every two weeks. Its only job is to give a comparable number over time.
Method
What we measure, how we train, and what we don't claim.
This page exists so you can judge the product on its method rather than its marketing. It is written for a curious non-specialist.
A game trains
Adaptive difficulty, hints, a named technique and an error review. Game scores never enter your ability profile.
The six baseline tests, coming in an update
The tests are not in the app yet. Each will be an implementation of a published task. We link one or two review papers per method rather than a single flattering study.
Digit span
Forward and backward recall of digit lists. Reported as items held.
n-back
Updating and monitoring in working memory. Reported as level and accuracy.
Stroop
Interference control. Reported as the effect in milliseconds.
Simple reaction time
Median and variability across 30 trials, with display lag compensated.
Matrix reasoning
Pattern completion of increasing difficulty. Reported as items solved.
Mental rotation
Same-or-mirror judgements at several angles. Reported as accuracy and time.
How the score is built
Raw units come first — items, milliseconds, errors. A percentile against an age band is shown only once we have enough data for that band, and it is labelled as an estimate until then.
Children are compared with children of the same age band, never with adults. Tasks have age-appropriate versions with the same structure.
Games are validated against the tests: if a game does not move its target test over a cohort, it gets reworked or dropped.
What we don't promise
- A higher IQ
- Better school grades or exam results
- Prevention or treatment of any condition, including dementia
- A diagnosis of ADHD or any other disorder
If you are worried about your memory or attention, talk to a doctor. This app is not a substitute for that conversation.