Attention & control Logan & Cowan, 1984

Stop-signal task

Respond to the arrow, unless it turns red just after it appears. It measures how quickly you can cancel a response that is already on its way.

Measures: Action cancellation (stop-signal reaction time)
Template length: about 5 minutes (48 trials)

An arrow points left or right and you respond with F or J. On a quarter of trials the arrow turns red shortly after it appears, which means stop: do not respond.

Because the stop signal comes after the go stimulus, you are cancelling a response that has already started rather than deciding not to begin it. The time this takes, the stop-signal reaction time (SSRT), cannot be observed directly. It is estimated with the race model, which treats going and stopping as two processes racing each other.

Try it yourself

A shortened run of the same template researchers use. Nothing is recorded and you do not need an account.

How the task works

1A fixation cross appears for 500 ms.
2An arrow appears. Press F for left and J for right within 1.2 seconds.
3On 25% of trials the arrow turns red 250 ms after it appears. Withhold your response on those trials.

What people usually find

People succeed at stopping on some stop trials and fail on others. The longer the delay before the stop signal, the more often they fail. When stopping is tuned to succeed about half the time, SSRTs in healthy adults typically fall around 200 to 250 ms.

Responses on failed stop trials are usually faster than ordinary go responses, as the race model predicts: only fast go responses escape the stop process.

How to score it

Probability of responding on stop trials, and mean go RT.
SSRT by the integration method: take the nth go RT, where n is the probability of responding on stop trials times the number of go trials, and subtract the mean stop-signal delay.
Check that failed-stop RTs are faster than go RTs; if not, the race model does not hold and SSRT is not meaningful.

Before you run it

The template uses a fixed 250 ms stop-signal delay, which is fine for a demonstration. For SSRT, the consensus guide recommends a staircase that moves the delay 50 ms up after a successful stop and 50 ms down after a failure. The prompt below sets that up.
Tell participants not to wait for the stop signal. Slowing down on go trials to catch it inflates go RT and distorts SSRT.

Run it as a study

Use the template as it is

The demo above is the same code as the Stop-Signal Task template in the gallery. Add it to a project, change the settings in config.js, publish it and share the link. Reaction times and responses come back to your dashboard.

Create a free account to use it

Or build a variant: use a staircase so you can estimate ssrt

Paste this into the AI panel in the experiment editor. Read the code it writes before you collect anything.

Build a stop-signal task with 4 blocks of 64 trials, 25% stop trials in random order. Go stimulus: a left or right arrow; respond with F or J within 1500 ms. On stop trials the arrow turns red after a stop-signal delay that starts at 250 ms and follows a staircase: +50 ms after a successful stop, −50 ms after a failed stop, kept between 50 and 900 ms. Record the delay on every stop trial. Tell participants in the instructions not to wait for the stop signal, and show their average go RT after each block.

Common questions

What is a normal stop-signal reaction time?

In healthy adults SSRT is typically around 200 to 250 ms, but it depends on the task settings and estimation method, so compare groups within the same study.

Why use a staircase for the stop-signal delay?

It keeps stopping success near 50%, where SSRT estimates are most stable, and adjusts automatically to fast and slow responders.

Can the stop-signal task run in a browser?

Yes. Keyboard timing in browsers is precise enough for SSRT at the group level. Keep the participant in fullscreen, which the template does by default.