Zero stop-delay entries raised a fixed integration estimate by 30.520 ms on average in 266 ds000030 scanner runs
An audit of 266 ds000030 scanner runs found 611 of 8,512 STOP trials (7.18%) with a literal StopSignalDelay of zero and a positive planned LadderTime, across 180 participants. Each was a failed stop with a response before the planned delay. Holding GO responses and stop-response probability fixed, using literal rather than planned delays raised the illustrative integration estimate by a mean 30.520 ms (subject-bootstrap 95% interval 25.600–35.653 ms); 241 participants passed limited behavioral screens, where the mean difference was 31.078 ms. This is an estimator sensitivity in released behavioral fields, not a validated correction to biological stopping speed: planned times do not prove that a tone occurred, and literal markers are not verified acoustic onsets. The analyzed tables are pinned to the OpenNeuro ds000030 1.0.0 Git tag and version-specific DOI. The source tables, code, full trial audits, methods, limits, and reproduction instructions are preserved in the supporting files.
stop-signal task · stop-signal reaction time · behavioral data · event timing · estimator sensitivity · OpenNeuro · reproducibility
References
2- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The analyzed tables are pinned to Git commit 4070b6eea231517bfeff42527a46d9d166ac4e13, which is the repository's 1.0.0 tag.
- doi:10.1038/sdata.2016.110
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesThe delay field changes the contrast sign in 118 of 265 tested ds000030 runs
The inspected audit supplied the 270 pinned public input tables and metadata, source hashes, timing-field distinction, and basic-screen definitions. Newly written code calculates the delay contrasts and ordering statistics from those raw tables.
- usesRemoval of one STOP record increases SSRT with recorded delays and decreases SSRT with planned delays in ds000030 run sub-10227.
The calculation uses this publication's three delay definitions, GO omission replacement, and baseline sample rules. They specify the 241-run primary sample and its R, H, and L score comparisons.
- A second implementation reproduces 266 stop-signal runs, but delay conventions reverse participant orderings
[Legacy relation: reproduces] A separately written implementation with freshly retrieved matching pinned inputs reproduces every archived participant and STOP audit row, all screen flags, three absolute integration estimates, rank statistics, and both specified participant-bootstrap intervals. This is same-data computational reproduction; cue timing and biological validity are not independently established.
- usesDelay-field choice changes participant ranks and GO-speed associations in 126 ds000030 control runs
Its read all-266 scanner audit reported screened rank disagreement, motivating this control-only extension that counts exact pair reversals and paired changes in a GO-speed association; its cohort and screen are not claimed replicated.