Instrument register
Every test on this site, what it is allowed to claim, and how many people we have measured. The count is the table below, and the answer to the last part is nobody.
2 published instruments. Reference sample: 0 people, for both.
No test appears on this site without a row on this page, and the row is written before the test ships, not after. If a test is live and missing here, that is a fault — tell us and it becomes a correction.
| Instrument | Domain | Norming rung | Reference sample | Version | Published | Last reviewed |
|---|---|---|---|---|---|---|
| Digit span |
Working Memory | Rung 0 | 0 | Methodology 1.2 | 3 Sep 2026 | 5 Sep 2026 |
| Odd one out |
Pattern Recognition | Rung 0 | 0 | Methodology 1.2 | 5 Sep 2026 | 5 Sep 2026 |
| Symbol match |
Processing Speed | Rung 0 | 0 | Methodology 1.2 | 6 Sep 2026 | 6 Sep 2026 |
| Paper folding |
Spatial Reasoning | Rung 0 | 0 | Methodology 1.2 | 6 Sep 2026 | 6 Sep 2026 |
| Word meaning |
Verbal Comprehension | Rung 0 | 0 | Methodology 1.2 | 7 Sep 2026 | 7 Sep 2026 |
| Arithmetic |
Numerical Facility | Rung 0 | 0 | Methodology 1.2 | 7 Sep 2026 | 7 Sep 2026 |
| Ignore the Word |
Attention | Rung 0 | 0 | Methodology 1.2 | 7 Sep 2026 | 7 Sep 2026 |
Four of these five make a shape. Digit span, odd one out, symbol match and paper folding each add an arm; the fourth completes it and gives you a number for how far you got through what those tasks can measure. Word meaning is not an arm yet — arms will be added in one deliberate version change when the new domains are complete, rather than one at a time, because every added arm changes the number that already-shared links are holding.
Build your Cognitive Fingerprint →Nothing stored. Your scores travel in the web address.
What each column means, and what would change it
Norming rung
The rung says what a result from this instrument is permitted to claim. It is a property of the instrument, and it is versioned and dated like everything else. It is the single most important column on this page, because it is what stops a number from being dressed up as more than it is.
| Rung | What it is based on | What a result may say |
|---|---|---|
| 0 — Descriptive | Nothing. Bands come from how the task is built, not from data about people | What you did on the task, what the task tested, and what it cannot tell you. No comparison to other people, in any form |
| 1 — Provisional | Real observed data, too few people to publish as a norm, and an unrepresentative sample | Rung 0, plus "compared with the first N people who took this test" — with N stated on the page. Still no "average" |
| 2 — Reference | Enough real data, with the sample described honestly | Comparison to our test-takers, characterised as such. Never a claim about the general population |
| 3 — Normative | Enough data to break out by group, with methodology open to outside review | Comparisons by age or segment, with their uncertainty stated. Never anything diagnostic. At any rung, ever |
We are on Rung 0 and we will not use the word "average" about our own results until Rung 2. That is a commitment written down now, while it costs us nothing, so that it constrains us later when it does.
How many results Rung 2 requires has not been decided, and we are not going to invent a number for it. A threshold has to come from how much the instrument's own scores vary, which we cannot know until we have some. Publishing a round figure today would be a guess wearing the clothes of a standard.
Reference sample
How many people's results we hold for that instrument. It is 0 for both, and that is not a rounding. Neither test stores a result — your answers and your score stay in your browser and are never sent to us, so there is nothing to accumulate. The privacy notice describes the one thing that is sent, which is a counter with nothing attached to it.
This number changes only by a deliberate decision, and that decision means asking you first. It will not creep up quietly while this page says zero.
Version
A result carries the methodology version it was produced under, so a past result does not silently change meaning when we change a rule. A version goes up when a scoring rule, an exclusion rule, the domain model or the norming rung changes — not when we fix a sentence. Fixing a sentence is a correction, and those are listed separately.
Last reviewed
When we last read the instrument's pages against the system as it actually is. These two drift apart on their own, and every correction on this site so far has come from that drift rather than from the measurement being wrong.
Why a register this short is worth publishing
This page is not impressive. It lists a handful of tests and tells you we have measured nobody.
That is the reason it exists. The columns on this page — rung, sample size, version, review date — are the ones that decide whether a score means anything, and they are the ones most often absent from cognitive tests on the internet. A site with seven tests and no register is not further along than this one; it is the same information withheld.
The commitment is the point, not the table. Every future instrument gets a row here, filled in before it is published, including the parts that are unflattering. When this page has more rows, you will be able to check whether the zero moved and what we said when it did.