Replacing zero recorded delays with positive staircase delays lowers mean SSRT by 32.8 ms in ds000030 version 1.0.0

The audit uses OpenNeuro ds000030 version 1.0.0. It found 611 stop trials with zero recorded delays and positive staircase delays. They occurred in 180 of 266 participants with scanner event files. All 611 responses came before the staircase time. Among 265 computable participants, replacing these zero entries lowered mean estimated stop-signal reaction time (SSRT) by 32.8 ms. The participant bootstrap 95% interval was 28.1 to 37.9 ms. This comparison assumes that LadderTime gives the scheduled delay in milliseconds. It does not establish the actual sound time. A basic behavior screen retained 245 participants and gave a similar mean change of 33.6 ms. All 7,980 staircase transitions matched recorded movements. Group comparisons are exploratory and unadjusted. Each file has only 32 stop trials. The package preserves 537 source files, versions, hashes, code, the fixed plan, trial flags, tables, figures, and reproduction instructions. Six synthetic tests passed. An independent program verified all 795 A/B/C estimates. An offline rerun reproduced all scientific tables exactly.

stop-signal task · response inhibition · stop-signal reaction time · SSRT · fMRI · OpenNeuro · dataset audit · reaction time · data quality