This is not a medical or screening tool
Before anything else: nothing on ReactionBench can diagnose ADHD, concussion, or any other condition. If you are worried about attention, memory or focus — for yourself or a child — talk to a doctor or a qualified clinician. A validated clinical assessment looks at far more than how fast you click a button, and this page exists to explain that gap honestly, not to soften it.
Reaction tasks in attention science
Psychologists have used reaction time for over a century as a rough index of how quickly the brain takes in a signal and acts on it. In attention research specifically, the interesting number is rarely a single fast click — it is how speed and accuracy hold up over many trials, and how well someone can stop a response already in motion. Two families of task do most of the work in this area, and this site's go/no-go and 5-choice tests are the closest available cousins of each.
Continuous performance tasks present a long, repetitive stream of stimuli where you respond to most of them and withhold on a rare few, tracking whether speed and accuracy drift as the minutes pass. The paradigm dates to 1956, when Rosvold and colleagues designed it to study lapses of sustained attention after brain injury; it has since become one of the most widely used tools in attention research generally, including — but far from limited to — ADHD studies. Closest here: the go/no-go test. Choice and inhibition tasks instead offer several possible signals, each needing a different response or none at all, tapping decision speed and the ability to hold back the wrong, faster answer; the go/no-go paradigm traces to Donders' 19th-century mental-chronometry work, and the closely related stop-signal task was formalised by Logan and Cowan in 1984 as a race between a "go" process and a "stop" process. Closest here: the 5-choice test.
The tests here are built for curiosity and self-benchmarking, not assessment. Validated clinical batteries use calibrated hardware, controlled rooms, age-matched norms and trained scoring — none of which a web page can reproduce.
What a reaction test can — and can't — hint at
Used honestly, a few metrics across repeated runs can be interesting to you personally, as a baseline to track. None of them are evidence of anything clinical.
| Signal | What it is | How much it tells you |
|---|---|---|
| Median reaction time | Your typical speed on a given day | Only meaningful against your own past runs on the same device — never an absolute verdict |
| Consistency (spread) | How much your rounds vary | More telling than raw speed, but wide variation can simply mean fatigue, boredom or distraction |
| False starts | How often you respond before the signal | Interesting as a personal habit to notice; a single session tells you almost nothing on its own |
| Drift over a session | Whether you slow down toward the end of a long run | Everyone drifts a little — a laggy tab or a noisy room produces the same pattern |
The gap between a home game and a clinical diagnosis
Attention conditions like ADHD are diagnosed by trained clinicians using a structured history, observations across multiple settings, standardised rating scales, and a picture built over months — not minutes. Reaction tasks are, at most, one supporting data point among many, and even purpose-built continuous performance tests used in clinical research are known to be sensitive but not specific: poor performance can come from tiredness, low motivation, or an unrelated condition just as easily as from an attention disorder, which is why clinical guidance treats them as a supplement to — never a replacement for — a full assessment.
Four gaps matter most between a browser game and a real evaluation. One setting, one moment: a clinician looks at behaviour across home, school and work over time; a session here captures one moment on one device. No calibration: clinical tools use controlled hardware and age-matched norms, while an ordinary monitor and mouse add their own uncontrolled latency. Confounds everywhere: sleep, caffeine, mood, practice and a slow connection all move a score far more than any underlying trait could. No trained scoring: real assessments are interpreted by professionals alongside history and rating scales, not by an automatic grade badge.
- 01Rosvold, H. E., Mirsky, A. F., Sarason, I., Bransome, E. D., & Beck, L. H. (1956). A continuous performance test of brain damage. Journal of Consulting Psychology, 20(5), 343-350. Link
- 02Logan, G. D., & Cowan, W. B. (1984). On the ability to inhibit thought and action: A theory of an act of control. Psychological Review, 91(3), 295-327. Link
- 03Riccio, C. A., & Reynolds, C. R. (2001). Continuous performance tests are sensitive to ADHD in adults but lack specificity: A review and critique for differential diagnosis. Annals of the New York Academy of Sciences, 931, 113-139. Link