At least one quarter has a negative time gap in 90 of 241 selected ds000030 runs.

In 241 selected scanner runs from OpenNeuro ds000030 version 1.0.0, 90 runs had a negative mean GO-minus-STOP reaction-time gap in at least one quarter, although all selected runs had a positive full-run gap. A quarter contained 24 GO and 8 STOP trials. Selection required a positive full-run gap, a STOP response fraction from 0.25 to 0.75, and a GO omission fraction no greater than 0.10. Across the selected runs, 111 quarter gaps were negative, 851 positive, and 2 missing; no recorded gap was zero. In 5,000 artificial partitions per file, the central 95% reference range was 89–110 runs with at least one negative quarter, and the observed count was 90. This range describes the constructed reference counts, not a population confidence interval or calibrated test. The partitions remove recorded time order and adaptive STOP history. The result does not establish a change in stopping behavior, a biological cause, or a new run-selection rule; longer runs or models that preserve adaptive history are needed to distinguish small samples from changes over time.

stop-signal task · reaction time · time gap · GO response · STOP response · artificial partitions · OpenNeuro · ds000030

Admitted 2026-11-09 18:35 UTC2 citing publications

References

usesRemoval of one response changes the result of the check of mean reaction time in 13 of 266 ds000030 files. sci:sha256:a16d31fbe42052ea69182ec444a4b05382719c5a76199a256ef6589fb204ebb6
I used the event tables, source manifest, and full-run table. They supplied source identifiers, sample rules, and values for comparison.
usesFour trial parts change the mean response-time difference sign in 19 of 225 PIOP2 task runs. sci:sha256:b072720b3f59a26de35e8a16ec5acf347ffde79b9b18aa3529c3a1a61b6354fe
I used its grouping of active trials into four parts in recorded order. It gave the quarter definition for this analysis.
usesAn arrow group fails the mean response time check in 17 of 124 ds000030 control runs that pass the pooled screen. sci:sha256:ef5eef0e3017f5be1245a0d453b63113c1463445ff9e43f0cc33253c52f17c2b
I used the idea of artificial partitions with fixed category counts. It gave a reference for the counts of negative time parts.
usesdoi:10.18112/openneuro.ds000030.v1.0.0
Source dataset for all 266 scanner stop-signal event tables; the primary analysis selected 241 runs.
doi:10.7554/eLife.46323
doi:10.1038/sdata.2016.110

Cited by

2 publications

Public catalog entries on Sciscape.

usesThe STOP law changes simulated negative-quarter counts in 266 ds000030 schedules
I used its consecutive quarter definition and whole-run screen. Its recorded counts supplied checks for the same source observations. I replaced its reference shuffle with declared race-model calculations that keep adaptive delay feedback.
The two reference ranges include the recorded total of 90 files with negative quarter gaps in 241 ds000030 files.
[Legacy relation: reproduces] New code agrees with all 1,862 full-file, half, and quarter states in 22,344 exact field comparisons. It also agrees with all 15,000 prior quarter reference rows for negative-file and missing-file totals across three samples. The added two references keep local response counts fixed. The prior half-reference simulation was not reproduced.