The arithmetic mean response difference is −26.29 ms in 87 ds000030 runs with planned delays less than 200 ms.

We examined failed STOP responses and the immediately preceding responding GO in 266 stop-signal scanner runs from OpenNeuro ds000030 tag 1.0.0. Under the main rule, 499 pairs from 87 runs had planned delays from 0 ms to less than 200 ms. The equal-weight mean of within-run response-time differences (failed STOP RT minus preceding GO RT) was −26.289567 ms (95% run-bootstrap interval −51.912396 to −3.156531 ms). Sensitivity estimates varied; several intervals included zero. At planned delays of 0 ms and 50 ms, the intervals also included zero. These exploratory results do not establish context or stochastic independence in an independent race model, do not provide a stopping-time estimate, and do not measure physical sound onset. The analysis retains assumptions about source field units and meanings, selection from adaptive delay sequences, and the limited number of STOP trials per run. More failed STOP responses at fixed short delays and verified sound-delivery records would improve this diagnostic.

stop-signal task · response time · independent race model · OpenNeuro · bootstrap · adaptive delay · reproducible research