Task quarter weights lowered the failed stop minus go time contrast by 55 ms in 177 selected AOMIC PIOP2 runs

We tested whether task position changes a mean response-time check in public stop-signal data. We used OpenNeuro ds002790 version 2.0.0. The planned primary sample had 177 of 226 runs. The same failed stop responses were compared with two go means. Weighting the go mean to the same task quarters changed the mean contrast by -54.6 [-69.6, -40.1] ms. Brackets give the 95% participant bootstrap interval. Thirteen runs changed from a nonnegative to a negative contrast; one changed in the other direction. Two-block and eight-block checks had similar directions. The immediately preceding recorded go check gave a different result: 7.2 [-12.5, 27.8] ms in 174 runs. An exact two-context example shows an aggregate sign change under stated local race assumptions. The evidence shows sensitivity to the go baseline. It does not establish valid stopping-time estimates. The package contains source data, hashes, a saved plan, code, results, figures, and instructions for offline reproduction.

stop-signal task · response inhibition · response time · baseline sensitivity · task position · AOMIC PIOP2