Stop-signal event tables match across ds000030 tags 00016 and 1.0.0, and participant delay arithmetic replicates

All 266 scanner stop-signal event tables bundled with an earlier delay audit are byte-identical to the corresponding tables at original-provider Git tags 00016 and 1.0.0. Their dataset descriptions and change logs differ, so this establishes correspondence for the event tables only, not for the whole datasets, imaging, or acquisition histories. An independent Decimal calculation on the same observations reproduces the earlier audit’s mean planned-minus-logged delay arithmetic (30.5199385519 ms; median 17.5308509376 ms), zero and positive-delay components (+32.7185150376 and −2.1985764857 ms), all participant values within 4.62×10⁻¹⁴ ms, and its 10,000-resample participant-bootstrap interval (25.7010181610–35.6247553402 ms). The inputs contain 8,512 stop trials; 611 rows from 180 participants have zero recorded delay despite a positive planned delay. A separate exploratory analysis and behavioral-screened sensitivity results are also preserved. This is computational reproduction on the same participants, not an independent biological replication. Exact deployed task and conversion versions and physical tone onset remain unknown; delay replacements are sensitivity scenarios, and no absolute SSRT, causal, clinical, or neural conclusion is established.

stop-signal task · response inhibition · SSRT · event tables · computational replication · dataset provenance · OpenNeuro · ds000030 · bootstrap · delay encoding · neuroimaging · reproducibility