Figure 2.

Interruptions in motor planning account quantitatively for observed dips in RT distributions. (a) The paradigm used by Bompas and Sumner⁴⁷ to study saccadic inhibition. Participants were instructed to make a saccade (blue arrow) to the target (black square) as soon as it appeared, and to ignore any distracters. In most trials, a distracter (white square) was briefly flashed (50 ms) opposite to the target. The delay between target onset and distracter onset (SOA) varied randomly across trials. For each trial, the processing time (PT) is equal to RT − SOA, and corresponds to the time interval between distracter onset and saccade onset. (b–g) RT distributions reported by Bompas and Sumner⁴⁷ (observer 1, black traces, 1200 trials per trace) along with our model results (red traces). SOAs are indicated, in ms. In the rise-to-threshold model, the mean interruption occurred from 85 to 115 ms after distracter onset (dark gray shades), with the onset and offset times varying across trials (light gray shades show 1 SD in each direction). (h) PT distributions for the simulated (red trace) and experimental data from Bompas and Sumner⁴⁷ (observer 1, black traces). PT histograms include aggregated data from all SOAs. All experimental data were redrawn from Bompas and Sumner⁴⁷ (Methods).