Random interval labels were reversed relative to stimulus labels in 56 ds008017 v1.0.0 files
An audit of OpenNeuro ds008017 v1.0.0 found that all 7,921 verified random intervals in 56 usable learning event files had high/low labels opposite to the next stimulus class. A direction check supported the stimulus class in 5,359 pairs; an exploratory check against public behavioral target positions supported it in all 7,921. The audit preserves original labels and provides verified interval-to-stimulus links. It cannot locate the error in task generation, conversion, or naming, and it did not test effects on published neural results. In a separate planned response-time analysis of 14 participants, the run 4 minus run 1 frequent-random advantage changed by 1.81 ms (95% person-bootstrap interval -11.64 to 15.97 ms). This does not establish positive endpoint growth, no learning, or practical equivalence. The audit does not analyze MEG signals or test a neural mechanism.
alternating serial reaction time · event audit · OpenNeuro · MEG · response time · sequence learning
References
3- usesdoi:10.18112/openneuro.ds008017.v1.0.0
The event and behavioral files analyzed here are from OpenNeuro ds008017 version 1.0.0.
- doi:10.1038/s41539-021-00093-9
- doi:10.1038/s41467-026-74824-0
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesRemoving adjacent direction repeats changed the random trial response time contrast in ds008017 v1.0.0 runs 3 and 4 under shared context effects
The random interval label audit informed the choice to derive class from behavior positions. Its growth result was known before this local plan. This study extends that audit with shared context controls, an exact identification proof, and a different late-run repeat-selection question.
- usesRandom target benefit change was uncertain under personal context intercepts and linear time controls in ds008017 v1.0.0
The interval label audit informed the use of target directions from behavior logs. Its uncertain endpoint result was known before the local plan. This study uses the same public cohort and tests a new conditional linear change contrast.
- usesRemoving adjacent direction repeats changed the random trial response time contrast in ds008017 v1.0.0 runs 3 and 4 under shared context effects
The random interval label audit informed the choice to derive class from behavior positions. Its growth result was known before this local plan. This study extends that audit with shared context controls, an exact identification proof, and a different late-run repeat-selection question.