STOP record removal produces larger maximum SSRT changes than GO record removal in 122 selected ds000030 controls
In 122 selected ds000030 control runs, the largest absolute change in the integration stop-signal reaction time (SSRT) after removal of one STOP record was greater than the largest change after removal of one GO record, for both a ceiling-rank quantile and linear interpolation. Median maxima were 7.538 ms (STOP) and 1.391 ms (GO) for the rank rule, and 6.827 ms and 0.899 ms for interpolation. Under the rank rule, 6,000 of 11,712 GO removals increased SSRT and 5,712 left it unchanged; none decreased it. This is a fixed-table calculation, not a replay of the adaptive task or a measure of biological stability, causal effect, confidence interval, or session reliability. Each run had 32 STOP trials, below a published group-comparison recommendation of 50. The analyzed files are identified by Git commit and SHA256 hashes; the recorded mapping of that commit to dataset release 1.0.0 was not independently checked.
stop-signal task · stop-signal reaction time · response inhibition · record removal · integration method · quantile rule · data correction · ds000030
References
4- Removal of one STOP record changes H SSRT by at least 10 ms in 37 of 122 selected ds000030 controls with the rank rule. sci:sha256:f5f62b50e884281b1d42b16ab54e9665e748cd6fba28157d895966c868990043
[Legacy relation: reproduces] New rational-arithmetic code checks all 252 baseline rows and all 8,064 STOP-removal rows for the two quantile rules and three delay definitions. The prior source subset, manifest, numerical references, and actual STOP-removal procedure enable these checks and the new GO-removal comparison. Prior code was read but not run.
- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The source manifest identifies the dataset metadata DOI and records CC0. The analyzed files are pinned to Git commit 4070b6eea231517bfeff42527a46d9d166ac4e13; a Git-to-release mapping check was not made.
- doi:10.1038/sdata.2016.110
- doi:10.7554/eLife.46323
Consensus recommendations contextualize the short run length and integration score choices; the 10% omission screen and ceiling-rank rule remain inherited choices.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesRemoval of one STOP record changes partial Pearson correlation by less than 0.008 in 232 ds000030 pairs.
Its GO removal comparison and maximum replacement for omitted GO responses guide the GO-versus-STOP calculation. This work does not reproduce its 122-control scanner H-score comparison.