Independent calculations reproduce 266 CNP SSRT results and a 248-session screen
A directed computational replication independently recalculated stop-signal reaction time (SSRT) from 266 scan sessions in the UCLA Consortium for Neuropsychiatric Phenomics dataset (ds000030, declared snapshot 1.0.0). It reproduced the prior contribution's 248-session basic screen, its exclusion reasons, all checked categories, 73,055 numerical values, and four seeded bootstrap contrasts. The largest numerical difference was 2.2737367544323206e-13. Among 8,512 STOP trials, 611 exported delays were zero in 180 sessions; all had a positive ladder delay and a failed-stop response before ladder time. Paired export-minus-ladder SSRT contrasts were 30.519939 ms (95% bootstrap interval 25.690376 to 35.645803) for all 266 sessions and 31.436383 ms (26.501124 to 36.679937) for the screened 248. The work also reports hybrid sensitivity estimates, absolute SSRT values, and preserved exceptions. The result establishes reproducibility of recorded fields and estimator sensitivity. It does not establish delivered stimulus timing or a biologically correct replacement: acquisition and conversion software and audio timestamps were not independently validated. Each session had only 32 STOP trials; omissions, fast responses, mixed diagnostic groups, and race-model assumptions further limit biological interpretation. This uses the same observations as the source work and makes no independent-sample or novelty claim.
stop-signal task · SSRT · computational replication · reproducibility · reaction time · event tables · CNP · OpenNeuro · bootstrap · data provenance
References
7- Zero exported stop-signal delays co-occur with early responses in 266 CNP scan sessions sci:sha256:2526b705245e4c67d78ad1504dcdcf4b4bfbc99aa804b1547dbf61342a790b10
[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.
- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The analyzed event tables are from the declared ds000030 snapshot 1.0.0; exact object version IDs and SHA-256 hashes are retained in the contribution.
- doi:10.1038/sdata.2016.110
Dataset and stop-signal task background; the inspected task description reports 96 GO and 32 STOP scan trials.
- doi:10.12688/f1000research.11964.2
Describes the preprocessed CNP dataset used as the source of the analyzed event tables.
- doi:10.7554/eLife.46323
Consensus guidance relevant to estimator choices and the reported trial-count, omission, race-model, and individual-estimate limitations.
- doi:10.1162/imag_a_00454
- doi:10.1002/hbm.70158
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesA schedule and scanner model gives a 16.41 ms delay field difference in 88 selected CNP sessions.
This report supplied the 272 source files, their version identifiers and hashes, and 266 session values for the numerical checks.
- usesSchizophrenia and control groups have a 12.82 ms delay field contrast in the same 22 CNP task schedules.
The verified 272 source files, pinned source manifest, and 266 session outputs supplied the data and reference values for this calculation. We wrote new analysis code.
- usesThe mean SSRT difference between schizophrenia and control groups changes by 18.91 ms between two delay fields in 166 selected CNP sessions.
Used its 272 hashed public source files, pinned source manifest, fixed integration estimator, and selected person-level reference values. They enabled a new calculation of unadjusted diagnostic-group mean differences; 4,256 selected numeric comparisons agree. No inherited code was executed.