A 1 s GO response limit changes sample selection in ds000030 v1.0.0.

This sensitivity analysis asks whether a 1 s response limit changes stop-signal reaction-time (SSRT) estimates or sample selection in public ds000030 v1.0.0 data. All 266 selected runs have 96 GO and 32 STOP trials. Under the baseline rules, 241 runs are selected. A GO-only 1 s limit leaves every one of those 241 SSRT estimates unchanged, but changes sample selection for 17 runs. Applying the limit to both GO and STOP responses changes 18 baseline estimates; the mean paired change is -1.240659 ms (95% participant bootstrap interval [-2.058421, -0.599786] ms), and 12 absolute changes are at least 10 ms. Different policy-specific means use different selected samples. Each run has only 32 STOP trials, which limits individual estimates. The deployed task version and original hardware timing are unknown, and the analysis does not establish a biological cause or stopping time. The response limit tests analysis rules; it is not a correction to the recorded data.

stop-signal task · reaction time · response limit · SSRT · sample selection · sensitivity analysis · ds000030 · fMRI