The scheduled delays agree with ladder histories in 266 ds000030 1.0.0 scanner runs.

Across 266 selected OpenNeuro ds000030 scanner runs labeled 1.0.0, all 8,512 scheduled STOP delays match the corresponding values at indices 16–31 in 532 JSON ladder histories. These are the only full 16-value sequence matches; one-entry offsets and ±50 ms shifts do not match. All 7,980 within-run event updates and 532 final stored updates follow the 50 ms outcome rule with a zero floor. The floor is needed for 169 updates. Five scalar comparisons to the target agree: 266 participants, 8,512 STOP trials, 611 recorded-zero trials across 180 participants, and a 32.71851503759398496240601504 ms mean increase when only recorded zeros are replaced by planned delays (exactly 69625/2128 ms). All 55,151 audit checks pass, including the retained synthetic method checks. The comparisons establish consistency of stored fields in the hash-pinned files, not physical sound onset or independent timing. Acquisition-code and converter identity, the tag-to-commit mapping, and actual sound delivery remain unknown. The dataset includes only 32 STOP trials per participant; no absolute SSRT, model fit, population interval, or biological effect is estimated.

stop-signal task · scanner runs · scheduled delays · ladder histories · metadata audit · OpenNeuro ds000030