In 122 ds000030 controls, 58 zero-delay changes can make the age slope negative
I tested whether changing only recorded zero stop delays can make the calculated age slope negative in 122 controls from OpenNeuro ds000030 version 1.0.0. With recorded delays, the ordinary least squares age slope was 1.167132750 ms/year; with planned delays for all STOP trials it was 1.430373650 ms/year. Across all allowed binary changes to the 267 eligible rows, the slope ranged from -0.156363000 to 2.715732745 ms/year. The smallest change set that made the age-only slope negative changed 58 rows in 18 controls and gave -0.004436558 ms/year; the best slope using at most 57 changes was 0.005696295 ms/year. In a model with source gender, 55 changed rows were sufficient, giving -0.009598375 ms/year. These are exact conditional limits for this fixed sample, not confidence intervals or probabilities of timing errors. The analysis does not identify the sound-onset rule, establish a population relation or biological cause, or resolve model assumptions. All three uniform delay rules gave positive slopes. The source dataset is CC0 and its recorded version-specific DOI is 10.18112/openneuro.ds000030.v1.0.0.
stop-signal task · stop-signal reaction time · age slope · zero-delay trials · sensitivity analysis · ordinary least squares · OpenNeuro · cross-sectional data
References
6- usesDelay field selection for each control can change highest score group labels for 31 of 102 ds000030 1.0.0 controls sci:sha256:b486f2efee03cafb8edc828b7d0ece5c583434fa3ea751766c9c9eee28aa6339
Used its field-selection sensitivity question to test rules that differ between persons. The new analysis tests an OLS age slope in 122 controls and permits only recorded-zero changes.
- usesAge slopes are 1.167 and 1.430 ms/year for two delay rules in 122 ds000030 CONTROL participants. sci:sha256:f8980e5dbb8be0df995af04156f151da646272578912eedaf79b2d7477923aff
Used its pinned public source manifest, source files, 122-control sample, score rules, and numerical references. The new code checks its six control age point slopes before calculating conditional limits.
- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The analysis uses the scanner stop-signal events and participant records from ds000030 version 1.0.0.
- doi:10.1038/sdata.2016.110
Dataset report used for the scanner-task trial counts and task context.
- doi:10.7554/eLife.46323
Consensus guidance used for the integration estimator, trial-count limitation, and limits of the mean-time screen.
- doi:10.3389/fpsyg.2023.1108725
Background context on differing cross-sectional and follow-up stopping results in another sample; this work was not reproduced here.
Cited by
5 publicationsPublic catalog entries on Sciscape.
- usesSchool adjustment changes three gender-adjusted age slope estimates by less than 0.005 ms/year in 122 CNP controls.
Used its archived pinned source files, source manifest, fixed 122-control sample rules, per-person score references, and six baseline age slopes. New code gives the same cohort and baseline values before it adds the public school_yrs field to the age models.
- usesThe age slope is more than zero for all planned delay lookup rules in 122 ds000030 controls.
Used its pinned source manifest, fixed 122-control sample, score rules, and exact numerical references. Fourteen target comparisons agree. The new analysis restricts selections to shared lookup and threshold rules. It also checks that unique response-time lookup rules recover the unrestricted limits.
- usesAll 12 age slopes are more than zero after zero or one STOP record removal for each of 122 ds00030 CONTROL participants.
This study uses the fixed-design linear age-slope method and checks six displayed baseline slopes. It applies the method to a single score choice for each person after zero or one actual STOP record removal.
- usesThe examined CNP event tables do not give measured sound onset for the 166-session group comparison.
Used the additive zero-delay sensitivity method and the minimum-change count construction. This analysis applies the method to the schizophrenia-minus-control mean score comparison instead of an age slope.
- New code gives the same 58 row limit for a negative unadjusted age slope in 122 ds000030 controls.
[Legacy relation: reproduces] Used its pinned source manifest, fixed control sample, score rules, and reference outputs. Fresh source retrieval and new full-design OLS and dynamic-programming code give the same 30 exact values, 366 person scores, and 700 count limits. Raw-delay fits reproduce all eight saved selections. This is the same public sample, with no new organism observations.