🛑Stop Signal Test

Can you stop an action after your brain has already started preparing it? Take this free Stop Signal Test to experience one of psychology's most important response-inhibition paradigms.

Most trials are Go trials: respond quickly to the direction or identity of a stimulus. On a smaller number of trials, a Stop signal appears shortly after the Go stimulus. When that happens, try to cancel the response you were about to make.

The key scientific measure is not simply your percentage of successful stops. A properly designed task estimates the hidden speed of stopping using Stop-Signal Reaction Time (SSRT).

How to Take the Stop Signal Test

1. 1Press Start Test.
2. 2Respond quickly and accurately to every normal Go stimulus.
3. 3Keep preparing the normal response unless a stop signal actually occurs.
4. 4If the stop signal appears, try not to press anything.
5. 5Continue through both Go and Stop trials.
6. 6Do not deliberately slow all responses to make stopping easier.
7. 7Review Go reaction time, Go accuracy, stop success, and SSRT if reported.

The task intentionally creates conflict between two goals: respond quickly, and occasionally cancel that response. Both are necessary for a meaningful result.

What Is the Stop Signal Task?

The Stop Signal Task (SST) is a laboratory paradigm for studying action cancellation, usually called response inhibition. The participant first performs a normal reaction-time task — for example, left arrow means press Left, right arrow means press Right. On some trials, a stop signal appears after a short delay, and the participant must then suppress the response.

The challenge is that the stopping instruction arrives after the Go process has already begun. This makes the task different from simply deciding not to respond from the beginning.

Stop Signal vs. Go/No-Go

These tasks are closely related but not identical. In a Go/No-Go task, the stimulus itself tells you immediately whether to respond or withhold — for example, green means Go and red means No-Go. In a Stop Signal task, you initially receive a Go instruction, and only later does a second signal tell you to cancel the action.

That timing distinction matters. Go/No-Go emphasizes response restraint — not initiating a response to a No-Go cue. Stop Signal emphasizes response cancellation — stopping an action that has already been initiated internally.

The Horse-Race Model

The classic interpretation uses the independent race model, developed in influential work by Gordon Logan and William Cowan. Imagine two internal processes racing toward completion: the Go process, triggered by the Go stimulus, which produces a response if it finishes first; and the Stop process, triggered when the Stop signal appears, which successfully cancels the response if it finishes before the Go process reaches completion.

You cannot directly observe the time at which successful stopping finishes because the behavioral result is no response. The race model provides a way to estimate that hidden stopping time.

What Is Stop-Signal Reaction Time?

Stop-Signal Reaction Time (SSRT) estimates how long the internal stopping process takes. It is not simply Go reaction time minus stop-signal delay on one trial. Modern best-practice methods estimate SSRT using the distribution of Go reaction times and the probability of responding on Stop trials.

This is important because SSRT is a model-based latent estimate, not an observable stopwatch reading. A casual web page should avoid pretending SSRT is exact to the millisecond.

What Is Stop-Signal Delay?

The Stop-Signal Delay (SSD) is the time between the Go stimulus and the Stop signal. A short SSD makes stopping easier because the Stop process gets a head start. A long SSD makes stopping harder because the Go response is already closer to completion. A fixed SSD is therefore not ideal for comparing people with different Go speeds — a strong implementation adjusts the SSD dynamically.

Why the Stop Delay Should Adapt

Expert consensus recommends a tracking procedure that changes SSD based on performance: a successful stop makes the next stop harder by increasing SSD, and a failed stop makes the next stop easier by decreasing SSD. This gradually moves the task toward roughly 50% successful stopping.

That may sound strange — why not aim for 100%? Because SSRT estimation works best when the Stop process sometimes wins and sometimes loses. If every Stop trial succeeds, the stop signal is probably appearing too early. If nearly every Stop trial fails, it is appearing too late.

Why You Should Not Wait for the Stop Signal

Once people understand the task, some try the strategy of responding slowly on every trial so they have more time to stop. That undermines the measurement — this behavior is called strategic slowing or waiting.

The Stop Signal Task assumes that participants genuinely perform the Go task and respond promptly. If you wait to see whether a stop signal appears, you change the Go reaction-time distribution and can distort SSRT estimation. The correct strategy is to respond normally and quickly on Go trials, and stop only when the signal actually appears.

Why Stop Trials Should Be Relatively Rare

If almost every trial contained a stop signal, people would constantly expect to cancel, and the task would become a waiting exercise. Standard Stop Signal designs therefore use a minority of Stop trials — often around one quarter of trials in research implementations. This keeps Go responding dominant while still providing enough stop observations for analysis.

The exact number of trials also matters. A very short fun version can demonstrate the paradigm, but reliable SSRT estimation usually requires substantially more data than a handful of stops.

What Does a Lower SSRT Mean?

Within a valid implementation, a lower SSRT generally represents faster estimated stopping. But interpretation requires caution — SSRT can be distorted when participants wait strategically, Go omissions are high, Stop failures are extremely rare or extremely common, the stop signal is missed, there are too few trials, or model assumptions are violated.

A number such as 210 ms should therefore not be treated as a pure biological constant. The quality of the task procedure matters as much as the calculation.

Stop Success Rate Is Not the Main Score

Imagine Person A stops successfully on 90% of Stop trials while Person B stops on 50%. It may look as if Person A has much stronger inhibition, but perhaps Person A simply responds extremely slowly on Go trials, or the Stop signal appeared earlier. That is why success percentage alone is not a fair inhibition measure.

In an adaptive task, the system deliberately changes stop timing so success hovers near a useful range, while SSRT captures the estimated speed of cancellation.

Stop Signal vs. Reaction Time Test

The Reaction Time Test asks how quickly can you initiate a response. Stop Signal asks how quickly can you cancel that response after a late stop cue. Fast going and fast stopping are not the same ability — a person with very fast Go reactions may actually make stopping harder because the Go process reaches completion sooner. That competition is exactly what makes the Stop Signal paradigm interesting.

Does the Stop Signal Test Measure Impulsivity?

The task is widely used in research involving inhibitory control and impulsive behavior. But one SSRT cannot be translated into a statement like "you are 70% impulsive." Impulsivity is a broad construct involving multiple behaviors and psychological processes. The Stop Signal Task specifically measures performance in an action-cancellation paradigm and should not be used by itself to diagnose ADHD, substance-use disorders, OCD, impulse-control disorders, or other clinical conditions.

Frequently Asked Questions

The Stop Signal Task creates a paradox: you must commit to responding quickly while remaining capable of canceling that response at the last moment. That tension is the experiment. A strong result is not produced by simply waiting — it comes from maintaining genuine Go performance while successfully stopping when the late signal enters the race.