The exported delay gives higher mean trigger-failure estimates in two race models for 112 CNP runs.

In 112 candidate control runs from the Consortium for Neuropsychiatric Phenomics (CNP), a working race model gave higher mean trigger-failure probability (q) with the exported stop-signal delay than with the planned delay. For STOP gamma shapes 1 and 4, exported-minus-planned differences were 13.26426363 and 13.98142576 percentage points. Paired run-bootstrap 95% intervals were 10.05484691 to 16.67776620 and 10.97865913 to 17.21823585 percentage points. The model estimated q and mean STOP latency together. Mean STOP latency was lower by 25.46168198 and 24.37104777 ms with exported delay. The result is conditional on the tested delay fields, trial rules, parameter bounds, and two working distributions. The original model input table and verified tone-onset times are unavailable, so this is not a reproduction of the original hierarchical model and does not establish true biological trigger-failure rates or which delay is correct.

stop-signal task · trigger failure · race model · response inhibition · delay sensitivity · gamma distribution · CNP · behavioral modeling