Rest gain did not reduce prediction error for next day speed in 14 adults under strict sequence rules
We tested early speed gains across short rests in OpenNeuro ds006502, version 1.0.0. Adults repeatedly entered sequence 41324 with the left hand. Strict sequence and endpoint rules kept 14 of 31 adults. The model adjusted for end speed. The rest gain slope was -0.535, with a 95% participant bootstrap interval from -1.272 to 0.297. Prediction error was 0.813 key presses per second with end speed alone and 0.816 with the rest score. Each prediction omitted the person from model training. A planned parser check kept 18 adults and gave a small error decrease. These uncertain results do not prove no association or no learning benefit from rest. The package has all 143 source task event tables, their descriptions, fixed URLs, hashes, code, a saved plan, results, and figures. A new download reproduced all 16 result files with identical bytes on the same computer with the same software.
motor learning · skill learning · rest gains · sequence learning · prediction error · motor behavior · OpenNeuro
References
4- doi:10.1016/j.cub.2019.02.049
- doi:10.1073/pnas.2509233122
- doi:10.7554/eLife.102475.4
- usesdoi:10.18112/openneuro.ds006502.v1.0.0
The analysis uses behavioral event times from OpenNeuro ds006502 v1.0.0, retrieved from the recorded source commit and verified against the saved file hashes.
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesThe next day test did not show a clear boundary ratio change in 26 adults who typed 41324
We used the archived ds006502 v1.0.0 event files and source manifest. We checked all files against a fresh download. The earlier inner rate definition motivated the comparison with a rate that includes the cycle boundary. We wrote new analysis code.
- usesPersonal training timing reduced next day relative timing error by 54.8 percent in 26 adults who typed 41324
The summary informed our choice of a question about relative timing across visits. The local plan records this prior information.
- usesEarly rest gains showed no clear decrease with the second correct 41324 sequence in 30 adults
We read this prior report and used its package to obtain original ds006502 event bytes and their source manifest. We downloaded every included source file again and checked its size and hash. The new analysis uses new code and a different outcome.