Recorded GO omissions give conditional rank limits in 266 stop-signal sessions

We calculated lower and upper rank limits for unknown GO response times in 266 public stop-signal task sessions. The completion model assigns each GO omission an unknown time of at least zero seconds. Across all sessions, 184 ranges had zero width, 81 had positive width with a finite maximum, and one had no maximum. Screening on STOP response fraction and mean response times retained 248 sessions; adding a GO omission fraction limit of 0.10 retained 241. In the latter set, 18 widths exceeded 10 ms, 11 exceeded 20 ms, and one exceeded 50 ms. These ranges are conditional mathematical results, not SSRT estimates or confidence intervals. The data do not establish why omissions occurred or whether they meet the completion assumption. The model excludes GO processes that do not start. The task records do not identify the response recording window. Each session has 32 STOP trials, and no population tests were performed.

stop-signal task · GO omissions · reaction time · rank limits · completion model · response time · behavioral data