NeuroHaven

Where your mind comes home

How NeuroHaven Age™ works

Versión en español

What it measures, how results are calculated, and what to expect. NeuroHaven, November 2026.

Summary

NeuroHaven Age™ is a 15 minute online test of five thinking skills. It compares your performance with published research on adults your age and gives you your NeuroHaven Age™, the age your performance looks closest to, and a result for each area.

Your results are provisional. The research behind these comparisons used similar tests, not this exact one, so we converted the numbers to our format. We are now calibrating NeuroHaven Age on people taking this exact test. NeuroHaven Age is for general wellness and performance tracking. It is not a diagnostic test, and this page is not a peer reviewed study.

1. What it measures

We chose tasks that research links to age, that give similar results when repeated, and that work online.

Processing speed: symbol matching

Match symbols to numbers for 90 seconds. Symbol matching tasks have a strong link with age: r = 0.60 for the touchscreen version used in UK Biobank [1].

Reaction speed: four choice reaction time

Press the matching key or tap the matching box as a target appears. A similar four choice test repeated at r = 0.83 in a small sample of 20 people, tested twice in a row [2].

Visual memory: spatial span

Watch squares light up, then repeat the order. Averaging across sequence lengths repeated better than using the longest one correct: 0.84 versus 0.68, in 49 young adults [3].

Focus under conflict: color words

Answer the ink color of a word, not the word itself. A related online task in a different format showed a moderate link with age, r = 0.31, in adults aged 50 to 79, so our comparison here is only a rough estimate [4].

Mental flexibility: connect the circles

Connect numbers and letters in alternating order as fast as you can. The touchscreen version used in UK Biobank showed r = 0.57 with age [1]. A similar online task in adults aged 50 to 79 showed a much weaker link, r = 0.14, which is one reason we will check this against our own data [4].

You also answer a few questions about how you are today, such as sleep and stress, to help put your results in context. These answers do not change your results.

2. How your results are calculated

  1. Each task result is compared with what research suggests is expected at your age.
  2. Your NeuroHaven Age™ uses three tasks: symbol matching, spatial span and connect the circles. Device differences affect these speed tasks less than reaction time [5].
  3. Because one session includes some chance variation, we pull the result slightly toward the middle.
  4. We convert that into your NeuroHaven Age™. Reaction speed and color words appear in your five areas but do not count toward it.

The research gives us the age trends. The conversions to this exact format, such as adjusting a paper test to a screen, are our own estimates [6, 7, 8].

3. What your results will look like

Your NeuroHaven Age™. For example, 41 when you are 45. This means your performance today looks closest to adults around 41. Within 3 years of your real age, we say you are right at your age. It is a comparison of performance on these tasks. It does not measure the age or health of your brain.

Your likely range. It sits in the small note under your NeuroHaven Age™, because no single sitting is exact. It narrows with every sitting.

Your five areas. Each shows your result next to what is estimated for your age, labeled Ahead, Typical or Room to grow.

4. How accurate is it?

A short online test can only be so precise. These figures are our own estimates, not published results:

  • Chance variation means one session can shift your closest age match by about 10 years.
  • The true value is within about 12 years of our estimate about 7 times in 10. So your NeuroHaven Age™ is our closest match, not an exact answer.
  • Comparing against people taking this exact test should remove part of this error as our data grows. Repeating the test on different days should help most.

Variation is normal. In one study of healthy adults retested after a week, about a third showed a reliable drop on at least one test and about a third a reliable rise [9].

5. Retests

Retest in about a week, then monthly, on the same type of device. Once you have more than one session, we combine sessions from the same type of device, at least a day apart.

People usually do better on these tasks with practice, even after weeks [1, 6]. A higher number on a retest may be practice, so we do not call it improvement. We tell you whether a session moved within your normal range or beyond it, using a standard method for reliable change [10].

6. What it cannot tell you

  • It does not diagnose, screen for or rule out any condition.
  • It does not predict your future or tell you why a result is high or low.
  • Results can reflect sleep, stress, distraction, your device, your schooling and how familiar you are with computers [3, 5, 7].
  • If you have concerns about your thinking or memory, talk with a health care professional, whatever your result.

7. How to get your most accurate result

  • Use a computer with a keyboard and mouse if you can, and the same type of device every time.
  • Give it 15 minutes in a quiet place, with notifications off, and stay on the page.
  • Take it when you feel rested. Avoid testing when very tired, unwell or after alcohol.
  • Wear your usual glasses, and do it on your own without notes.

References

  1. Fawns-Ritchie and Deary, 2020. PLOS ONE. UK Biobank cognitive tests.
  2. Deary, Liewald and Nissan, 2011. Behavior Research Methods. The Deary-Liewald reaction time task.
  3. Woods et al., 2016. Frontiers in Human Neuroscience. Spatial span.
  4. Troyer et al., 2014. Frontiers in Aging Neuroscience. Online cognitive assessment.
  5. Passell et al., 2021. Behavior Research Methods. Cognitive test scores vary with choice of personal digital device.
  6. Pham et al., 2021. npj Digital Medicine. Smartphone symbol digit test.
  7. Tombaugh, 2004. Archives of Clinical Neuropsychology. Trail Making Test norms.
  8. Marrie et al., 2021. Frontiers in Neurology. Regression based cognitive norms.
  9. Karr et al., 2024. NIH Toolbox test retest reliability and reliable change.
  10. Jacobson and Truax reliable change method, summary.

For general wellness and performance tracking. Not a diagnostic test.