Fixed-response replay keeps a +50 ms start change through 16 stop trials in 495 of 532 ds000030 scanner staircases.
A fixed-response replay changed scanner staircase start values while holding each recorded response sequence constant. In 495 of 532 scanner staircases, a +50 ms start difference remained through all 16 presented STOP trials; in 37 it was erased by the last presented state. After the final update, the difference remained in 492 staircases. The analysis used 266 training–scanner pairs from OpenNeuro ds000030 v1.0.0 and all 8,512 scanner STOP rows, without diagnosis selection. It reconstructed 17,056 recorded STOP plans and 15,990 within-run transitions exactly. The mean increase in run mean planned delay when both starts increased by 50 ms was 47.955827067669176 ms (median 50 ms; range 9.375–50 ms). A recorded-delay covariance identity was also confirmed for all 266 pairs. The replay holds behavior fixed: changed delays may change responses, and the files contain no randomized delay reset. It does not estimate biological stopping speed, independent session reliability, or effects in a larger population. Physical cue times are unknown, each run has 32 STOP trials, and two training-to-scanner start exceptions remain unexplained. The numerical results are exact for the pinned input files and stated rules; checks were performed by one researcher and do not establish independent verification or correctness.
stop-signal task · fixed-response replay · staircase · planned delay · response inhibition · OpenNeuro · ds000030 · covariance · reproducibility
References
6- usesPost-update training values match 530 of 532 scanner staircase starts in ds000030 sci:sha256:507e8ced7b078092da14c41e94c62c7e534c22b53dc95bc4aac2a20898fa5f6d
The analysis uses its 533 event tables, four metadata files, manifests, and endpoint table. Its 530 of 532 start agreement motivates the changed-start replay.
- usesInitial delay dominated the SSRT agreement covariance term in 232 ds000030 pairs sci:sha256:e3d4735801d8736aeb1516ab1436f8ee6836d736bca0f2ae82e61d80da4a8280
Its score covariance account and missing reset comparison limit interpretation of the replay. The package acknowledges that existing account and does not treat fixed responses as a real delay reset.
- usesHybrid replacement of zero stop delays raises training–scanner SSRT rank correlation in 232 screened ds000030 participants sci:sha256:73e301e494ef180b4c6f8dcdc27c44c025efd5d1dc5785717ddcb0395f3146d9
Its training-scanner score comparison supplies the reason to measure how a shared start value stays in the planned-delay term. This replay does not reproduce its screened SSRT scores.
- usesdoi:10.18112/openneuro.ds000030.v1.0.0
The analysis uses the ds000030 v1.0.0 event files; the archived metadata identify this version and its CC0 license.
- doi:10.1038/sdata.2016.110
- supportsdoi:10.7554/eLife.46323
The consensus guide provides the cited limitations on individual SSRT estimates with few STOP trials.
Cited by
4 publicationsPublic catalog entries on Sciscape.
- The +50 ms start change stays at trial 16 in 495 of 532 ds000030 version 1.0.0 scanner staircase replays.
[Legacy relation: reproduces] New standard-library code agrees with 67,881 numeric or empty target replay cells and all 1,302 flagged row identifiers on the same pinned ds000030 version 1.0.0 data. It reproduces the 495/532 last-presented +50 ms retention count. Covariance and SSRT results are outside this comparison.
- usesThe same GO input gives higher mean staircase score correlation in 241 ds000030 scanner records.
Its public scanner input files and source inventory enabled this score calculation. Source hashes and four new public downloads were checked.
- usesOne outcome exchange changes the replay score by 25 ms or more in 259 of 265 ds000030 runs.
Its 266 public scanner event tables, four metadata files, source inventory, and replay rule enabled this outcome-exchange calculation. All included source hashes and original scanner paths were checked.
- usesThe score difference can be positive or negative in 125 of 232 ds000030 version 1.0.0 pairs under the monotonic response model.
Its fixed-response replay distinction supplied the comparison path. New code verifies that fixed-response reset scores fall inside the response-model bounds.