It cannot tell you why
Sleep, medication, anxiety, effort, language and simple unfamiliarity with testing all move scores, and none of them are the ability being measured. A low score is a question, not an answer.
Score interpretation
Your attention score, translated into something usable
A raw number means nothing until you know the scale it sits on and the group it was compared against. Convert between every common scale below, or get your own profile starting with a free two-minute task.
2 minFree 2-min test
6scales converted instantly
23skills scored separately
Score converter
The scales
Six scales, one underlying position
Every one of these expresses the same thing differently: how far from the average of a matched group a performance sits. They exist to avoid negative numbers, decimals or false precision when a result is written into a report or read out to someone.
| Scale | Mean | SD | Where you meet it |
|---|---|---|---|
| z-score | 0 | 1 | Research papers and statistical work. The raw statement of position: how many standard deviations from the mean. |
| T-score | 50 | 10 | Behaviour rating scales and personality inventories. Exists mainly to avoid negative numbers and decimals in reports. |
| Scaled score | 10 | 3 | Individual subtests of Wechsler batteries. |
| Standard score | 100 | 15 | Composite and index scores, including IQ-type indices. |
| Stanine | 5 | 2 | Educational testing. Nine coarse bands, deliberately imprecise. |
| Percentile | 50 | — | Feedback to non-specialists. Easy to grasp and easy to misread, because the spacing is not even. |
Free test
Put your abilities to the test, free, in two minutes
You do not have to commit to a full assessment to find out where you stand. This short task measures processing speed and visual attention with a symbol-digit paradigm, and gives you a result straight away.
Two minutes, no cost. The full assessment battery is there afterwards if you want the complete profile.
The most common misreading
Percentiles are not evenly spaced
This single fact causes more misinterpretation than anything else in cognitive testing. Because scores bunch up in the middle, ten percentile points near the centre are a small real difference, and ten points near the edge are a large one.
−3 / −2 SD
−2 / −1 SD
−1 / 0 SD
0 / +1 SD
+1 / +2 SD
+2 / +3 SD
2% of people
14% of people
34% of people
34% of people
14% of people
2% of people
Moving from the 50th to the 60th percentile is about a quarter of a standard deviation. Moving from the 88th to the 98th is more than a full one. When you need to compare two results, compare them in standard deviations and translate to percentiles only for the explanation.
Precision
A score is a band, never a point
A reported score is an estimate, not a reading off a ruler. How wide the band is depends on the reliability of the instrument. With a reliability of 0.90 on a standard-score scale, the ninety-five per cent interval runs roughly nine points either side, so a score of 104 and a score of 96 are not meaningfully different. Any report that gives a single number with no interval is telling you less than it appears to.
Two things a score cannot do, and one it must not
It cannot travel between instruments
Two tests can both report a standard score of 95 and mean genuinely different things if one was normed on a national sample and the other on undergraduates. Compare within a battery, not across publishers.
It cannot be read as a diagnosis
A score describes performance relative to a matched group. A diagnosis is a clinical judgement built from history, examination and multiple sources of evidence.
Where the numbers come from
The tasks that produce an attention score
Interpretation is easier once you have seen what is being counted. These are the paradigms behind the attention figures in a cognitive profile, and you can open any of them.
Sustained attention
Continuous performance task
Omissions, commissions, mean reaction time and reaction time variability all come out of a task like this one. Open it to see what is being counted.
Open the task →Inhibition
Stroop interference
Scored on the cost of the incongruent condition rather than on raw speed, which is why the interference index matters more than the total time.
Open the task →Divided attention
Dual-task tracking
Two demands at once. The score reflects the drop in each task when the other is running, which is the actual cost of dividing attention.
Open the task →The CogniFit scale
How a single session becomes a profile
CogniFit reports on a 0 to 800 scale where 800 is the maximum. Performance is grouped into five broad domains, attention, memory, reasoning, perception and coordination, and broken down further into twenty-three individual skills such as inhibition, shifting, updating, working memory, planning and response time. That structure is what lets one session produce both a headline figure and an explanation of which components carried it.
1
Raw performance
Each task records accuracy, speed, variability and error type, which is a great deal more information than a single total correct.
2
Normative comparison
Those variables are compared against a sample matched for age, drawn from more than three million completed assessments.
3
Profile and plan
The result is a score per domain and per skill, which then drives a training programme weighted towards the weaker components.
Training
Attention exercises behind the scores
8 of the exercises in the concentration programme. Each adapts its difficulty to your performance, which is what keeps a score moving for the right reason rather than through familiarity alone.
Credibility
What makes a normative comparison trustworthy
A score is only as good as the group it is compared against and the reliability of the task that produced it.
3,188,798
Completed assessments in the normative database, which is what allows a result to be read against age-matched peers rather than in the abstract.
Class II SaMD
The professional battery, CAB PRO, is an FDA-registered device for use by qualified health professionals. Registration number 3017544020.
Reliability reported
Internal consistency and test-retest figures are published per cognitive ability rather than only for the battery as a whole.
What this is and is not. The converter on this page is an educational tool that applies standard normal-distribution arithmetic. It does not know your reference group, your instrument or its reliability, so it cannot interpret a result on your behalf. CogniFit assessments measure performance on cognitive tasks and do not provide a medical diagnosis. A score is not a clinical finding. If a result worries you, discuss it with a qualified professional rather than acting on the number alone.
Questions
Before you start
How do I convert a z-score to a percentile?
Find the area under the normal curve to the left of the z-score. A z of 0 is the 50th percentile, +1 is roughly the 84th and -1 roughly the 16th. The converter on this page does it for any value on any of the common scales.
Is there a free test I can take to get my own score?
Yes. The free two-minute task on this page uses a symbol-digit paradigm to measure processing speed and visual attention and gives a result immediately. The full battery, 30 to 40 minutes, produces a complete profile across five domains and 23 abilities.
What is a T-score?
A rescaled z-score with a mean of 50 and a standard deviation of 10. It is used widely on behaviour rating scales because it avoids negative numbers and decimals in a report.
Why is the gap between the 50th and 60th percentile smaller than it looks?
Because scores bunch up in the middle of a normal distribution. Ten percentile points near the centre is about a quarter of a standard deviation; ten points near the top can be more than a full one. Compare in standard deviations, then translate to percentiles only for the explanation.
What does the confidence interval around a score mean?
It is the band within which the true score most likely sits, given the reliability of the instrument. With a reliability of 0.90 on a standard-score scale, the 95 per cent interval is roughly nine points either side, so 104 and 96 are not meaningfully different.
How does the CogniFit 0 to 800 scale work?
Performance is reported on a scale where 800 is the maximum, broken into five broad domains and twenty-three individual skills. That structure is what lets one session produce both a headline figure and a profile showing which components carried it.
Is a low attention score a sign of ADHD?
No. A score describes performance on a task, not a condition. Attention performance is affected by sleep, stress, medication and motivation, and diagnosis requires a clinical assessment across multiple sources of evidence. If you are concerned, speak to a health professional.
Can I compare my score with someone of a different age?
Only through the normed scales. Raw performance changes across the lifespan, so a raw score is not comparable between a fifteen-year-old and a sixty-year-old. Percentiles and standard scores are already adjusted, which is exactly what makes them comparable.
Stop converting other people's numbers
Two minutes free to get started, or one full session for an attention profile on a scale you can read, compare against your own age group and track from then on.