Planned delays at zero entries reduce mean SSRT estimates by 33 ms in 241 screened ds000030 1.0.0 scan runs
We tested how zero delay values affect stop signal reaction time (SSRT) estimates. We used all 266 scan runs and 267 training runs from OpenNeuro ds000030, snapshot 1.0.0. Every recorded zero on a stop trial had a positive planned delay and a response before that delay. There were 611 such trials in scan and 330 in training. Compatible public task code can leave the delay field at zero when a response ends the tone loop. With response scoring fixed, planned delay substitution reduces mean SSRT by 33.3 ms in 241 screened scan runs. The 95% participant bootstrap interval is 28.4 to 38.5 ms. The screened training change is 16.9 ms. The two contexts share 266 people. The package includes raw event bytes, source versions and hashes, code, result tables, figures, and offline instructions. An independent calculation checked all runs and 17,056 stop trials. An offline rerun reproduced all eight primary outputs. This result measures an estimator change. It does not establish true stopping time or how earlier studies used the fields.
stop-signal task · stop-signal reaction time · SSRT · planned delay · reaction time · cognitive neuroscience · estimator sensitivity · OpenNeuro ds000030
References
4- usesdoi:10.18112/openneuro.ds000030.v1.0.0
This exact OpenNeuro version supplies the 533 raw stop-signal event files and participant metadata analyzed here.
- doi:10.1038/sdata.2016.110
- supportsdoi:10.7554/eLife.46323
The publication supplies the nonparametric integration guidance used for go omission handling, screening context, and limits of short stop-signal tasks.
- doi:10.1162/imag_a_00454
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesUsing planned times at zero delay entries increases SSRT rank correlation between training and scan in 232 screened ds000030 1.0.0 participants
We used unchanged event bytes, the fixed source manifest and complete tree, and prior run estimates from this publication. A new raw estimator reproduced the run estimates and screens. We then tested paired agreement between training and scan. The earlier analysis code was read but was not copied or run.
- usesReplacing recorded zero delays with planned delays increases SSRT rank agreement between training and scan records in 232 screened ds000030 1.0.0 pairs
This audit supplied all 533 raw stop signal event files, seven source metadata files, source hashes, and delay rule definitions. We used these unchanged source bytes for a new paired calculation.