Four-part all STOP weights change the mean response-time difference sign in 6 of 225 PIOP2 runs.
This analysis separates all-STOP allocation from failed-STOP response selection in a four-part response-time contrast. In 225 complete AOMIC-PIOP2 scanner runs, replacing pooled GO weighting with all-STOP weights changed 6 signs (2.67%; bootstrap 95% interval 0.89%–4.89%), while the earlier failed-STOP weighting changed 19 signs (8.44%; 4.89%–12.44%). The mean allocation term was 11.35 ms (3.04–19.53) and the mean failure-selection term was 54.22 ms (39.58–68.14). In 266 CNP training runs, the terms were 0.0615 ms (0.0109–0.1141) and −3.07 ms (−4.40 to −1.82), with zero sign changes for either weighted contrast. In 266 CNP scanner runs, the terms were 0.3678 ms (0.1903–0.5838) and −4.92 ms (−6.24 to −3.68); all-STOP weighting changed zero signs and failed-STOP weighting changed two. The two terms can cancel, so sign-change counts do not add. Results depend on the tested partition. These are descriptive results from different tasks and cohorts; the analysis does not measure stopping latency, establish a valid screening rule, or test differences between datasets or CNP phases. Bootstrap intervals assume independent participants and describe sample composition.
response time · stop-signal task · failed STOP · trial weighting · omissions · bootstrap · AOMIC-PIOP2 · CNP · reproducibility
References
6- usesFour-part centering keeps 8.50% of recorded delay variance in 226 PIOP2 runs. sci:sha256:67b0cd981d07c76156851e59c41532cc2b2554f1756336ab4582e8311bea08fb
Its sparse-part counts and retained delay-variance result supplied a limit on time-part comparisons. This analysis uses that limit to interpret the outcome-selection term. It does not reproduce the delay-variance result or use its code.
- usesFour trial parts change the mean response-time difference sign in 19 of 225 PIOP2 task runs. sci:sha256:b072720b3f59a26de35e8a16ec5acf347ffde79b9b18aa3529c3a1a61b6354fe
This analysis uses its checksum-pinned event files, four-part contrast, sample rules, and numerical reference table. New code reproduces 6,064 selected reference fields. It separates the change into all STOP allocation and failure selection terms.
- usesdoi:10.18112/openneuro.ds002790.v2.0.0
The analysis uses the ds002790 AOMIC-PIOP2 event tables at the recorded 2.0.0 version.
- doi:10.1038/s41597-021-00870-6
- doi:10.1038/sdata.2016.110
- doi:10.7554/eLife.46323
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesFour-part trial boundaries change the weighted response-time difference sign in 14 of 225 PIOP2 runs.
This work uses its all STOP weighting definition for the second boundary analysis. It recovers the six equal-part sign changes. It does not repeat its bootstrap or other-dataset calculations.
- New software gives the same six sign changes in 225 PIOP2 task runs.
[Legacy relation: reproduces] I used its pinned data and reference results. New rational arithmetic gave the same signs, counts, flags, and missing values in all three groups and both samples for 2, 4, and 8 parts. All RT values and 120 bootstrap interval limits agreed in the specified tolerances.
- usesFailed STOP weights for four parts give a lower phase rank correlation than pooled weights in 266 ds000030 pairs.
We used its CNP event copies, pinned source manifest, three weight definitions, and selected scalar references. These enabled an exact paired weight comparison and source checks.