I have attached an adaptive stop signal experiment that replicates the referenced study. It will be helpful to first read over the parameter and trial level notes.
Please note that SuperLab has built-in Go/No-Go support. However, this experiment uses more of a conditional Go/No-Go, e.g. a trial becomes a “No-Go” only if a stop signal is presented. Therefore, we implement Go/No-Go using rules at the trial level.
To implement the staircase procedure, we must keep track of whether or not the participant responded. We use a text parameter to store one of only two values. When the participant provides a response, the parameter is set to yes; otherwise, it will be set to no by default:
At the end of the trial, check if the stop signal was presented and if the participant responded. If so, then this should be considered a failed inhibition, which will decrease the difficulty of the next stop task. If the stop signal was presented and the participant did not respond, i.e. a successful inhibition, increase the difficulty of the next stop task.
Difficulty is determined by changing the delay between the go and stop signals. To implement this, a time trial level will be created and moved back and forth as a Rules action:
To answer your second question, SuperLab does not calculate stop signal reaction time (SSRT) for you. To do so, look through your data files and average out the most recent stop signal delays. Also, find the median reaction time for trials that did not present a stop signal. Then, subtract average stop signal delays from median reaction time. This will be your SSRT value.
adaptiveStopSignal.zip (271 KB)