How to Read a Creyos Report: Scores, Percentiles and Domains

A Creyos report turns a set of short computer puzzles into standard scores that compare you with people your age; for what Creyos is and why clinicians use it, start with the Creyos guide. If you are holding your report now, the short version is this: 100 is the middle, scores from 88 to 112 are labelled average, and the labels at either end are statistical, not clinical. Below we explain each part of the report, walk through an illustrative example, and cover what “below average” does and does not tell you.

What a report can and can’t tell you

Creyos reports are designed to be shareable with the person assessed, so you may receive a copy. They are still clinical documents: Creyos’s own guide says the platform is not a diagnostic tool and should be used with other information and clinical judgment.

The two kinds of pages on a report

The summary page

The first page lists who was assessed (an ID, birth date and gender), how many tasks were completed and when, and the comparison group, such as “Females, 35–44.” Below that, every task appears on one line with the skill it measures and your standard score, plotted against a band marking below average, average and above average. Tasks are sorted from your highest score to your lowest, so the list doubles as a quick strengths-to-weaknesses view.

The task pages

Each task then gets its own page with a picture of the puzzle, a short description of the skill, your standard score on a bell curve, a shaded percentile range, the classification label, and a few everyday activities that draw on the same ability (adults and under-18s see different examples). If you have taken the task more than once, a trend section shows your past results with dates, a line of best fit through up to seven recent scores, and whether the change counts as meaningful compared with your first and your most recent result.

Notice what is not there: the standard report is built task by task. The four domains (memory, reasoning, concentration, verbal) are useful for organizing your thinking, but each task stands on its own score. Our task-by-task guide explains what every puzzle asks you to do.

Standard scores and percentiles in plain numbers

Creyos converts each raw task score into a standard score with an average of 100 and a standard deviation of 15, the same scale many IQ-style tests use. Because every task uses that scale, a 95 on Rotations and a 95 on Digit Span mean the same relative position. The comparison is with your own age group (and, when entered, gender) in a normative database of more than 85,000 people.

The 87 and 113 cut points and their approximate percentiles come from Creyos’s interpretation guide; the 130, 100 and 70 rows are ordinary bell-curve arithmetic for a mean of 100 and a standard deviation of 15. The percentile on your report is shown as a shaded range with a midpoint rather than a single exact number, because any one sitting has measurement error. Creyos also adjusts scores using each task’s reliability, which pulls extreme results slightly toward the middle to account for regression to the mean.

An annotated mock report

Imagine a 29-year-old who completed all 12 tasks; the comparison group is people aged 25–34.

How a clinician might read this example:

1. Start with the spread. Most scores sit in the average band, one is above it, and two are just below. A mixed profile like this is common in healthy people.
2. Look for a pattern, not a single number. The two lowest scores are both concentration tasks. That pattern is worth a conversation; it is not proof of an attention disorder.
3. Check how close the low scores are to the line. An 86 and an 84 sit only a few points under 87, well within the range that could shift with sleep or a distracting room.
4. Treat the flag as missing data. Token Search has no score, so nothing can be concluded about it until it is redone.
5. Bring in everything else. Sleep that week, mood, anxiety, how the session went, history since childhood, and questionnaire answers all shape what the numbers mean.

What “below average” does and doesn’t mean

“Below average” is a statistical label: by definition about one in five people your age scores there on any given task. It means “lower than most of the comparison group on this puzzle on this day.” It does not mean impaired, damaged or disordered. Creyos’s guide notes that it is normal for a profile to include both high and low scores, because tasks in different domains measure largely separate abilities.

What makes a low score more meaningful is consistency: several low scores clustered in one domain, a low score that repeats on retesting, or a result that fits other evidence such as lifelong symptoms. Even then, the next step might be closer questioning, labs, a sleep review or a referral. Creyos’s guide itself notes that unexpectedly low results can prompt a referral to a neuropsychologist or neurologist. The difference between a screen and a diagnosis is covered in mental health screening versus diagnosis.

The validity flag

When a task’s pattern of responses falls outside the bounds that 99% of normative scores fall within, the report replaces the score with a “potentially invalid” message. Common causes are a distraction mid-task, misreading the rules, a technical hiccup or (rarely) writing answers down. Creyos notes that clinicians often reassess a flagged task after making sure the rules were clear.

A flag is not an accusation and not a bad result; it just means that number shouldn’t be used. The reverse matters too: an unflagged low score can still reflect a distracted or exhausted day, because the flag catches only highly unusual patterns.

Things that move scores besides ability

• Sleep. In a large online study using these tasks, people reporting about four hours of sleep a night scored as if they were roughly eight years older on overall cognition, though short-term memory was not affected (Wild et al. 2018).
• Mood and anxiety. Depression and anxiety can slow thinking and pull attention inward; scores taken during a hard stretch may not reflect your usual ability.
• Effort and engagement. Rushing, giving up early on hard trials, or test anxiety all lower scores.
• Setup. A small screen, a trackpad instead of a mouse, a noisy room or interruptions can cost points on timed tasks.
• The comparison group. The normative sample is internet-recruited volunteers who were not screened for health conditions, so “average” means average for that broad population, not for a screened healthy group.
• Medications. Some medicines affect alertness. Don’t change any medication around a test on your own; your clinician will tell you what to do on test day.

Results over time

If you repeat tasks, each task page gains a trend section and a Meaningful Change Indicator, which compares the new result with both your baseline and your previous attempt. Scores are anchored to first-time performance, so modest gains on a repeat are expected from practice alone. Retesting with Creyos explains practice effects, the indicator’s fairly lenient statistical threshold, and how to make repeat sessions comparable. Repeating measures over time is one tool within measurement-based psychiatric care.

Your results visit

At Willow & Stone, Dr. Stacey Forbes, DNP, APRN, PMHNP-BC, reviews Creyos results with you as part of a clinical evaluation; the report supports her assessment and never replaces it. When results are reviewed, and the fee, are confirmed when you book. Useful questions to bring:

• Which scores matter most for the question we are trying to answer?
• Do any results look affected by how I slept or felt that day?
• Should any task be redone, and when?
• How do these results fit with my history and questionnaire answers?
• What is the next step: monitoring, labs, a treatment change, or a referral?

Creyos is registered and listed with the FDA as a Class II, 510(k)-exempt computerized cognitive assessment aid; it is not FDA-cleared or approved and does not diagnose any condition. For the broader picture of what follows an evaluation, see after a psychiatric evaluation. Fees for visits are on our pricing page, and you can request a consultation here.

Common Questions

Sources & Further Reading

• Creyos — Interpreting Creyos Health Cognitive Assessment Results (PDF)
• Creyos — Understanding the Creyos Health Normative Database (PDF)
• Creyos — The Creyos normative database (blog, Feb 2026)
• Wild et al. 2018, Sleep — daily sleep duration and cognition (PMID 30212878)
• Creyos — Online cognitive tasks (overview of the 12 tasks)
• eCFR — 21 CFR 882.1470, computerized cognitive assessment aid

This article is general education, not a diagnosis or advice for your situation. Creyos results support, and never replace, a clinical evaluation; only a clinician who has evaluated you can diagnose a condition or recommend treatment. Do not start, stop, or change a medication without talking to the clinician who prescribes it.