None of 126 ds000030 CONTROL runs contains a third-GO and reference pair with the same arrow pattern and nominal gap triplet.

We tested whether the third-GO and reference reaction-time (RT) groups in OpenNeuro ds000030 have a within-participant comparison with the same three-arrow-transition pattern and all three nominal onset gaps. The source comprises 126 CONTROL runs, with 96 GO, 32 STOP, and 128 NULL rows per run. The target selection rule gives 125 participants, 1,763 third-GO responses, and 5,247 reference responses. No comparison cell contains both groups in any of the 125 participants, even before selecting successful current responses across all 126 runs; therefore the specified primary RT estimate is not estimable. The code reproduces the target's raw and arrow-only participant contrasts for all 125 participants, with maximum difference 0.0 ms; the arrow-only mean is 13.547484105158857 ms. A secondary rule that matches arrow pattern and total gap, but not each gap, retains 57 participants (184 third-GO and 191 reference responses) and gives a mean contrast of 22.020047396575098 ms (simultaneous 95% interval [-8.098776739461318, 52.13887153261152] ms). This interval includes zero and does not establish absence of an RT difference or agreement between methods. The result identifies a support limit for this comparison and these source files. It does not show that timing explains the target estimate. The schedules are nominal and do not measure response-to-next-arrow time; STOP outcome, STOP delay, prior responses, and task position may affect the comparison.

stop-signal task · reaction time · trial history · nominal timing · support overlap · OpenNeuro ds000030