One outcome exchange changes the replay score by 25 ms or more in 259 of 265 ds000030 runs.

This result replays stop-signal delay staircases from 266 OpenNeuro ds000030 scanner runs. It asks whether equal starts, STOP response counts, and calculated endpoints also fix the mean planned delay and pooled integration score. They do not: in 259 of 265 runs with a defined score, exchanging two STOP outcomes within one staircase changes the replay score by at least 25 ms while preserving the endpoints. The median per-run maximum absolute change is 46.875 ms. This is an arithmetic sensitivity result from recorded inputs, not a measure of responses to changed delays, score bias, or a population effect. The analysis also reports secondary screens, all-orders ranges, exact replay checks, and limitations. It does not assign probabilities to exchanges, establish physical feasibility of altered paths, or estimate biological uncertainty.

stop-signal task · stop-signal reaction time · response inhibition · adaptive staircase · outcome order · planned delay · arithmetic sensitivity · OpenNeuro · ds000030