What Is the Go/No-Go Test?
The Go/No-Go Test is a simple-looking challenge with two different rules. When the stage turns green and shows GO, you should act quickly. When it turns red and shows NO-GO, you must resist the urge to respond. Both actions matter: pressing at the right time and deliberately not pressing at the wrong time.
This type of task is often used to explore response inhibition, or the ability to withhold an action when the situation calls for it. In ReflexBenchmark's online version, GO trials are more common than NO-GO trials. That makes pressing a frequent part of the experience, so the less common hold trials require particular attention to the signal on screen.
This is an informal browser response control and attention practice test, not a standardized clinical assessment. Its main purpose is to show what happened across one session: how often you responded correctly, how often you pressed when you should have held back, and how quickly you responded on successful GO trials.
How to Play This Online Go/No-Go Test
- Read the two instructions before starting: press Space or tap for green GO, and do nothing for red NO-GO.
- Choose Start Test. The game presents a brief ready interval before each signal.
- For a green GO signal, press Space or tap/click the active test area within the one-second response window.
- For a red NO-GO signal, avoid pressing or tapping until the trial ends, even if you feel prepared to respond.
- Finish all 40 trials. The result panel shows your percentage, response categories, recorded GO times, five-block chart and saving status.
Each trial is followed by brief feedback that tells you whether it was a hit, miss, correct hold or false alarm. A fresh ready interval appears before the next signal. There are exactly 28 GO and 12 NO-GO trials, distributed across five eight-trial blocks, although their order is shuffled within the session.
Do not press repeatedly while waiting between signals. The task reacts only during an active stimulus. If you hide the browser tab during an attempt, the game interrupts the session and does not save the unfinished score. Choose Restart to begin a complete new attempt.
The Four Outcomes: Hit, Miss, Hold and False Alarm
A GO hit means you responded to a green signal before its one-second deadline. The result records a reaction time for that successful response. A GO miss, sometimes called an omission, means no valid response was received before the deadline. The test does not treat that as a correct trial.
A correct NO-GO hold means you did not respond to the red signal. Holding back is a successful action here, even though no click is made. A NO-GO false alarm means you pressed or tapped while the red signal was active. The result records it separately from a missed GO.
These categories are worth inspecting together. For example, responding to nearly every signal may produce many GO hits but also several false alarms. Being extremely cautious might reduce false alarms while increasing missed GO signals. Neither pattern is fully described by reaction speed alone.
How Your Go/No-Go Score Is Calculated
The saved score is overall correct responses as a percentage of 40 trials. Correct GO hits and correct NO-GO holds both count toward the total. GO misses and NO-GO false alarms do not. The formula is (GO hits + correct NO-GO holds) ÷ 40 × 100, rounded to the nearest whole percent. The possible score range is 0% to 100%.
The result also separates GO accuracy from NO-GO hold performance. There are always 28 opportunities to make a GO hit and 12 opportunities to hold correctly. Its NO-GO hold-rate display uses correct holds ÷ 12 × 100. The number of false alarms can also be expressed as a share of those same 12 NO-GO trials, but the page shows the false-alarm count directly.
Worked Scoring Example
Imagine you finish with 25 GO hits, 3 missed GO signals, 10 correct NO-GO holds and 2 false alarms. That accounts for all 40 trials: 25 + 3 + 10 + 2 = 40.
Your correct total is 25 + 10 = 35. Your overall score is 35 ÷ 40 × 100 = 87.5%, displayed as 88%. Your NO-GO hold rate is 10 ÷ 12 × 100 = 83% after rounding. These figures are fictional examples explaining the math, not average or expected user results.
Unlike some reaction games, this score does not award extra points for reacting a few milliseconds faster. A faster GO response matters to the reaction-time metric, but a valid GO hit always contributes one correct trial to the main accuracy calculation.
Understanding GO Reaction Time and NO-GO Hold Rate
The median GO reaction time is calculated only from successful GO responses. The test sorts their recorded times and selects the middle value, or averages the two middle values when there is an even number. Missed GO trials and NO-GO holds do not receive a reaction-time measurement for this median.
If you have no successful GO responses, there is no meaningful GO reaction-time median to display; the result shows a dash. A lower median indicates that your successful GO responses were quicker in that session, but it does not automatically mean the attempt was better overall. Look at misses and false alarms as well.
The NO-GO hold rate answers a different question: how often did you correctly withhold a response when the red signal appeared? A single false alarm changes that percentage because there are only 12 NO-GO trials. For personal comparisons, read the percentage together with the actual number of holds and false alarms, rather than treating a small percentage change as a meaningful long-term trend.
What the Five-Block Chart and Trial History Show
The result chart contains five bars covering trials 1–8, 9–16, 17–24, 25–32 and 33–40. Each bar counts successful GO hits and correct NO-GO holds in its eight recorded trials. A bar showing 6/8 means six of those eight responses were correct. These are actual attempt totals, not a generated progress curve or an estimate.
The expandable 40-trial history identifies each signal as GO or NO-GO and labels its observed outcome. For successful GO hits and false alarms, it also shows the browser-recorded response time. Seeing the sequence can help you notice whether mistakes occurred in a cluster or were spread across the session.
When enough comparable website attempts are available, the score-saving system may also display an approximate attempt percentile. It requires at least 30 comparable previous attempts for this test mode. A pending rank does not mean your score failed to save. Site attempt percentiles are not population norms, medical scores or predictions of a person's real-world self-control.
Practical Tips for More Consistent Results
- Learn the signal rule first. Identify green GO as “respond” and red NO-GO as “hold” before starting the timed sequence.
- Wait for the actual signal. Tapping during the ready interval will not create a useful response and can disrupt your own rhythm.
- Avoid automatic repeated tapping. Each NO-GO trial asks you to withhold the response even when the previous trials encouraged pressing.
- Find a stable position. Keep your hand near the Space key or your finger ready for touch input without resting on a control that triggers accidental presses.
- Use the same device for comparisons. Switching between a mouse, touchscreen and keyboard changes the physical input experience.
- Reduce distractions. Turn attention to the test stage rather than notifications, nearby conversations or other browser tabs.
- Compare multiple metrics. Review accuracy, false alarms, misses, median GO time and the five blocks instead of chasing one number.
Repeated practice can make the rules and display feel more familiar. Improvement on this short browser task should not be taken as evidence that unrelated attention skills, driving ability or mental health have improved. The useful comparison is between attempts made under similar conditions.
Accuracy, Device Effects and Important Limitations
The game times inputs through your browser rather than a calibrated measurement device. Screen refresh behavior, browser workload, keyboard scanning, touch processing and system scheduling can change the recorded milliseconds. A one-second software window is useful for consistent game rules, but it does not make reaction times identical across devices.
The score API checks the expected 40-entry history, the 28/12 GO and NO-GO balance, valid outcome types and whether the reported score agrees with the outcomes. These checks support internal result consistency. They cannot independently prove that a person performed every response unaided or that the browser timing represents a laboratory measurement.
A single Go/No-Go game cannot diagnose ADHD, impulsivity, a neurological condition or any other medical concern. Professional tests may use different stimuli, trial ratios, timing rules and controlled administration. Treat this version as an accessible practice exercise and a record of your own specific session.