Most recorded 50 ms delay steps followed the reverse stop-outcome rule in all 171 ds005754 v1.1.1 records
We tested all 171 public stop-signal records from OpenNeuro ds005754 version 1.1.1. Of 18,711 nonzero 50 ms delay steps, 18,515 followed the reverse standard tracking rule under the written outcome code (98.95%). Every record had a majority of reverse steps. All 18,316 such steps within inferred blocks followed the reverse rule. Only 3.21% of stop trials had positive response times. In 160 screened records, a label swap changed fixed integration scores by a mean of 25.84 ms. It reversed the order of 6.69% of record pairs without ties. Most scores were negative under either label rule. The delay pattern alone cannot identify a correct repair. The package preserves 177 source files, hashes, code, results, figures, and instructions. Independent checks passed. A fresh download gave identical numeric results and figures. This audit concerns stored fields and calculation sensitivity. It does not validate human stopping times or test clinical or brain associations.
stop-signal task · inhibitory control · source audit · response inhibition · data provenance
References
4- usesdoi:10.18112/openneuro.ds005754.v1.1.1
We analyzed all available stop-signal behavior files from this exact OpenNeuro version.
- supportsdoi:10.7554/eLife.46323
This consensus guide supplies the standard stop-signal tracking direction, integration approach, and response-time checks used as reference conventions.
- supportsdoi:10.1017/S0954579419000300
This earlier task description reports a 16-trial stop practice and five 88-trial blocks; it provides context for, but does not establish, the inferred row structure in the audited files.
- doi:10.1177/09567976241231574
This is the linked study that reports the mean-go-response-time minus mean-delay estimator. Its clinical and brain findings were not tested here.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesSixty of 171 ds005754 v1.1.1 records failed a timing test under one fixed go law per record and written stop fields
We reused its 177 unchanged source files, hash manifest, and fixed version evidence. Its delay audit informed dataset selection. We downloaded every file again from the fixed public commit and wrote a new timing analysis.
- usesFlanker conflict gave a small and uncertain gain in linear ADHD symptom prediction for 174 ds005754 participants
This audit informed the choice of a separate task. We used its fixed source commit and root tree as retrieval checks.