Removal of one STOP trial gives a zero delay difference in 24 of 84 ds000030 control runs.

We recalculated paired stop-signal reaction-time (SSRT) estimates after removing each STOP trial in turn from 126 control runs in OpenNeuro ds000030, snapshot 1.0.0 (96 GO and 32 STOP trials per run). The delay contrast, SSRT(recorded delay) − SSRT(hybrid delay), equals the mean difference between the two STOP-delay series. Among 84 runs with a positive initial contrast, removing the sole row with zero recorded delay and positive hybrid delay reduced the contrast to zero in 24; the other 60 retained a positive contrast after every single removal. In the screened 121-run group without the pilot, 23 of 80 affected runs reached zero. The largest absolute contrast change in that group was 23.4375 ms. The largest omission-contrast change was 22.822657999768012 ms for inverse quantiles and 12.490383902738927 ms for linear interpolation. Both quantile methods gave identical delay-contrast changes. This is an exploratory calculation on the same observed rows, not a new sample or an adaptive replay. It does not establish sound onset, validate the hybrid delay replacement, estimate sampling uncertainty or test–retest reliability, or support a biological cause. The study uses 32 STOP trials per run and retains the stated race-model and data-quality limitations.

stop-signal task · stop-signal reaction time · trial removal · reaction time · quantile · omission handling · reproducibility · OpenNeuro