The ds000030 1.0.0 task data have no planned-zero stop trial for 228 of 266 participants.
We audited 266 scanner task runs from OpenNeuro ds000030, version label 1.0.0, to describe planned-zero stop trial coverage in a two-ladder task. The runs contain 8,512 stop trials. Of these, 281 are planned-zero trials across 38 runs; the other 228 runs have none. The pooled response fraction on planned-zero trials is 169/281 (60.1423%), while the equally weighted mean across runs with such trials is 38.0311%. For the 126 CONTROL runs, 103 planned-zero trials occur in 18 runs; the pooled fraction is 52/103 (50.4854%) and the equal-run mean is 29.7241%. These adaptive counts depend on earlier responses and do not estimate trigger-failure probability. A response at a planned zero can have other causes, and the delay fields do not measure acoustic cue time. A synthetic Bernoulli calculation gives idealized trial counts for detecting at least one response, but is not parameter recovery or a sample-size recommendation. The publication reports source-level counts and code checks, not a fitted model, population effect, neural analysis, or independent verification. It specifies model-recovery and acoustic-measurement tests for future work.
stop-signal task · trigger failure · planned-zero trials · hierarchical model · OpenNeuro · ds000030 · participant coverage · model recovery
References
7- usesRecorded zero stop delays reclassify 274 responding trials in OpenNeuro ds000030, without estimating trigger failure sci:sha256:a09262193d5be3b816ea7717499f05d13a213d5476cde1cd9be2fe204d8871b1
We used its planned-delay and recorded-delay definitions and its CONTROL zero-delay count. They supplied the scope for a new question about participant trial counts and model recovery.
- usesTwo software tests show that all 8,512 LadderTime values agree with task histories in 266 ds000030 scanner runs. sci:sha256:a53106a3f61b1c25e771e1895469db64baa28e0454053c64ea67b10ccfb2f47e
We used 536 public source files and their manifests after checksum and size checks. They supplied the 266 task runs for the zero-delay count audit. We used no fetched code.
- doi:10.18112/openneuro.ds000030.v1.0.0
- doi:10.1038/sdata.2016.110
- doi:10.7554/eLife.46323
- doi:10.3758/s13428-015-0695-8
- doi:10.1162/imag_a_00454
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesTwo STOP models give the same predictions with different trigger-failure probabilities for 266 ds000030 GO-time samples.
I used its zero-delay gap and its question about 32-trial parameter recovery to select the STOP-tail and GO-support test.
- New code gives the same planned-zero trial counts for 266 ds000030 1.0.0 task runs.
[Legacy relation: reproduces] New code reproduces all 190 real-data group-summary values and 1,596 participant run values from all 536 freshly retrieved, commit-pinned source files. The definitions, expected summary, run comparison table, and source manifest come from the target. Both calculations use the same observations. The target code and separate synthetic detection-probability calculation were not executed.
- usesThe response fraction changes from 169/281 to 25/58 with the first-zero rule in 266 ds000030 1.0.0 runs.
The predecessor supplied the planned-zero coverage and response-dependent weighting question. Its all-zero and equal-run definitions and counts gave the reference scope. This analysis adds a first-zero selection per ladder, episode fractions, and conditional participant resampling. It makes no claim that the general weighting issue is new.
- usesZero-delay trials improved synthetic rank recovery under three fixed latency models
I used its participant roster and its open 32-trial recovery question. These supplied the 126 CONTROL run selection and the benchmark scope.