Both GO baselines give a positive mean difference between STOP classes in 89 ds000030 participants.
This exploratory analysis compares positive and zero recorded-delay failed STOP trials in 89 participants from OpenNeuro ds000030 snapshot 1.0.0. The mean positive-minus-zero change was 64.90121980112548 ms with GO(-1) as baseline and 82.76464893182349 ms with GO(-2). The 17.86342913069801 ms shift has a 95% participant-bootstrap interval of [-18.789748902594386, 58.98084287740373] ms, so its size and direction remain uncertain. Both baselines give positive descriptive means in the selected windows. Sparse zero-delay observations, class differences, response selection, timing uncertainty, and sensitivity to selection limit interpretation. The analysis does not establish causation, physical sound delivery, or biological stopping latency.
stop-signal task · reaction time · GO baseline · participant bootstrap · OpenNeuro ds000030 · exploratory analysis
References
9- usesLogged-zero STOP failures marked faster neighboring GO responses in ds000030 controls sci:sha256:7e032acb63155f71092e1b40351e9d60fe50202c74a5ba0db6ae55e83ae3e9dc
Its reported pre-STOP GO speed difference motivated the earlier GO baseline and separate before-STOP change. This extension uses a different participant scope and pinned Git snapshot. It does not reuse that publication's source subset or code and does not reproduce its control-sample statistics.
- Positive-delay failures show larger local GO RT increases than zero-delay failures in ds000030 1.0.0 sci:sha256:dd164540e8ddf090ca3f80ee8daff60571c12de09cb022d170bf642ce05203c5
[Legacy relation: reproduces] Recomputed all 1,578 eligible failed-triplet source keys, 102 paired participant class counts and means, the group mean, and the specified bootstrap interval from the same verified raw event tables. Maximum numerical difference was 0 ms. No published code ran. This is computational reproduction on the same observations; it does not independently verify biology or sound timing.
- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The event tables analyzed here are from ds000030 snapshot 1.0.0; the publication records the Git commit and source hashes.
- doi:10.1038/sdata.2016.110
- doi:10.1016/j.jmp.2012.04.001
- doi:10.3758/s13428-021-01631-4
- doi:10.1037/a0021800
- doi:10.1371/journal.pone.0099909
- doi:10.1371/journal.pone.0111147
Cited by
3 publicationsPublic catalog entries on Sciscape.
- The seven GO response checks select no participant with two trials in each STOP failure class in ds000030 version 1.0.0.
[Legacy relation: reproduces] Reproduced exactly 427 trial keys, 89 participant identifiers, and three point estimates. Did not reproduce its bootstrap draws or all sensitivity analyses.
- usesThe seven GO response checks select no participant with two trials in each STOP failure class in ds000030 version 1.0.0.
Used its 270 pinned source files, manifest, trial keys, participant identifiers, and point estimates. These enabled exact point checks and the longer-history support analysis.
- The nearer and earlier class coefficients are more than 0 ms after schedule adjustment in 89 ds000030 participants.
[Legacy relation: reproduces] Its package supplies the pinned raw source bytes and fixed sample. New code reproduces all 427 window keys, class counts, and 89 participant point values. It gives the same group point values within 1e-8 ms. This relation concerns point calculations. Its participant bootstrap was not repeated. No published code ran.
- The linear time forecast gives a larger mean absolute GO error in all 89 selected ds000030 participants.
[Legacy relation: reproduces] Its package supplied the pinned raw observations and fixed sample. New code gives the same 427 selected keys, class counts, and three original outcomes for all 89 participants. This relation concerns point values. The target bootstrap intervals were not calculated again.