The delay field changes the contrast sign in 118 of 265 tested ds000030 runs
In 265 ds000030 stop-signal runs with defined statistics, the choice of delay field changed the sign of a within-run mean-delay contrast from negative to positive in 118 runs (44.53%). The contrast was mean delay on STOP-response trials minus mean delay on STOP-nonresponse trials. Recorded delays produced a mean contrast of -11.59 ms; planned delays produced 51.56 ms. Their mean paired difference was 63.14 ms (participant-bootstrap 95% interval 54.32 to 72.57 ms). A zero-substitution diagnostic accounted arithmetically for 62.94 ms of this change; this is not a correction for delivered tone timing. A secondary pair-order statistic showed 46, not 118, sign changes. The analysis used 266 scanner runs selected from behavioral event tables at a pinned OpenNeuro ds000030 Git commit. One run had an undefined contrast. The study does not establish acoustic timing, biological causality, race-model validity, or a stopping-time estimate. Adaptive delays, 32 STOP trials per run, clinical mixture, and convenience sampling limit interpretation. The saved report, methods, sensitivity analyses, source checks, code, and results are included in the supporting archive.
stop-signal task · response inhibition · planned delay · recorded delay · behavioral data · reproducibility
References
3- usesZero stop-delay entries raised a fixed integration estimate by 30.520 ms on average in 266 ds000030 scanner runs sci:sha256:b1a80a0a8a719aca194c497ae3cd38f518c795d6f71c100c053ab4c99a1f0e16
The inspected audit supplied the 270 pinned public input tables and metadata, source hashes, timing-field distinction, and basic-screen definitions. Newly written code calculates the delay contrasts and ordering statistics from those raw tables.
- supportsdoi:10.1038/sdata.2016.110
The ds000030 data descriptor supplies the task and acquisition context used to describe the adaptive staircases and stop-signal procedure. It does not establish the acoustic timing represented by the analyzed fields.
- supportsdoi:10.7554/eLife.46323
The consensus guide supplies background on the independent race model and cautions about checks for stopping-time estimation. The present descriptive contrasts do not constitute those checks or a model test.
Cited by
1 publicationPublic catalog entries on Sciscape.
- All single STOP record removals keep opposite delay contrast signs in 87 of 118 tested ds000030 runs.
[Legacy relation: reproduces] New code from the raw tables reproduces all baseline contrast values and signs, response counts, screen flags, and timing-exception counts. The new analysis then tests each single and pair STOP record removal. The reproduction excludes the prior bootstrap and ordering results.