Zero exported stop-signal delays co-occur with early responses in 266 CNP scan sessions
In 266 scan sessions from OpenNeuro ds000030, 611 of 8,512 stop trials (7.18%) had an exported StopSignalDelay of zero. Every such trial had a positive LadderTime, a recorded response before that delay, and an UnsuccessfulStop outcome. The trials occurred in 180 sessions (67.67%). Using exported delays instead of LadderTime in otherwise identical integration SSRT calculations increased the mean by 30.51994 ms across all sessions (bootstrap 95% interval 25.69038–35.64580 ms), and by 31.43638 ms in the 248-session basic-screen subset (26.50112–36.67994 ms). A secondary hybrid comparison gave mean differences of 32.71852 ms and 33.64415 ms, respectively. Positive exported delays also differed from LadderTime, partly offsetting the zero-delay contribution. These results show material sensitivity to field choice; they do not establish which field reflects delivered sound timing, validate individual SSRTs, or imply biological slowing. The analysis used one scan run per participant, with 32 stop trials per run, and did not analyze training events, MRI, or clinical scores.
stop-signal task · SSRT · reaction time · data provenance · OpenNeuro · behavioral data
References
5- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The analysis uses the scan stop-signal event tables from this exact OpenNeuro dataset version.
- doi:10.1038/sdata.2016.110
- doi:10.7554/eLife.46323
- doi:10.1162/imag_a_00454
- doi:10.1002/hbm.70158
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesRecorded GO omissions give conditional rank limits in 266 stop-signal sessions
We used its 270 source files, source version identifiers, and rank reference table. They enabled the completion ranges and baseline checks.
- usesThe two CNP model scripts do not show the source field for stop-signal delay.
We used its raw scanner event files, participant table, dataset metadata, and version-pinned source manifest. These inputs enabled the 113-control candidate input audit and fresh public source checks. We did not run its programs or recover the later study model input.
- Independent calculations reproduce 266 CNP SSRT results and a 248-session screen
[Legacy relation: reproduces] Independently recalculated all 266 scan-session and 8512 stop-trial values, the specific 248-session screen and reasons, and four seeded bootstrap contrasts; verified 73055 numerical values, all pinned public inputs, and six original-program output hashes.
- usesThe session error field contains the mean of block error counts in 270 ds000030 records.
We used its hash-checked stop-signal dictionary, dataset metadata, README, CHANGES, and one scanner event table. These inputs and its version records enabled the phenotype field audit and one-subject source-mapping check. Fresh requests to all seven version-pinned public S3 URLs then confirmed the same input bytes. We did not reproduce its cohort delay result or execute its code.