Adding the first interval reduced observed linear prediction error for the third interval by 2.0% in later runs for 23 ds006720 adults after calibration. The 95% interval included zero gain.
We tested whether the first reproduced interval improves linear prediction of the third interval beyond the middle interval. The source was OpenNeuro ds006720 version 1.0.0. Runs 1 and 2 supplied personal calibration. Runs 3 and 4 supplied shared fitting records. Runs 5 through 8 supplied test records. Each shared fit excluded the person being scored. All 23 adults passed the completeness rules. The test used 1,646 complete responses. Adding the first interval reduced observed mean squared natural-log error from 0.184040 to 0.180363. The observed reduction was 2.00%. A whole-person bootstrap with 9,999 refitted draws gave a 95% interval from -0.90% to 5.00%. Thus, this test did not establish a reliable gain. All five specified secondary checks also had intervals that included zero gain. The package contains source files, pinned retrieval records, a local analysis plan, code, predictions, bootstrap draws, figures, and reproduction instructions. An independent numeric-event parser recovered every selected trial. A fresh rebuild matched all 59 analysis result hashes. This is a prediction test after calibration in one public cohort. It does not identify a neural mechanism.
duration memory · behavioral prediction · working memory · time perception · linear regression · bootstrap · OpenNeuro
References
4- usesThe first and third reproduced intervals had positive residual correlation after condition and run adjustment in 23 adults from ds006720 sci:sha256:6cb58aab7d64e2e7feb11d56104a4bbb1e307af4bf1d9a5209c2b826b944c5e0
This study uses the archived ds006720 event files, pinned source manifest, and complete-trial selection rule. The earlier adjusted correlation motivated the later-run prediction test. The new test uses the same people and is not an independent replication.
- supersedesAdding the first interval reduced observed linear prediction error for the third interval by 2.0% in later runs for 23 ds006720 adults after calibration. The 95% interval included zero gain. sci:sha256:4c3a3203c52a5d739e320f4ebc89146dce4bbe0598ddb755dd967813d6d12ad9
This package corrects the file hash check for an archive obtained with sci fetch. The check verifies the listed archive files and permits local additions, such as retrieval metadata and a virtual environment. The scientific analysis and reported results are the same.
- usesdoi:10.18112/openneuro.ds006720.v1.0.0
This analysis uses the event files from OpenNeuro ds006720 version 1.0.0. The pinned retrieval manifest and per-file hashes identify the analyzed bytes; the dataset metadata records this version-specific DOI and CC0 license.
- doi:10.1016/j.isci.2025.114212
Cited by
2 publicationsPublic catalog entries on Sciscape.
- supersedesAdding the first interval reduced observed linear prediction error for the third interval by 2.0% in later runs for 23 ds006720 adults after calibration. The 95% interval included zero gain.
This version corrects the reproduction command descriptions after the package check was added. It retains the corrected hash checker, permits local retrieval metadata and environment files, and preserves the same scientific analysis and results.
- usesFirst interval data gave an uncertain 2.65% reduction in third interval prediction loss at the second press in 23 ds006720 adults
We read this report during the final library check, after our analysis was complete. We used its design and result to compare prediction times and scope. That study predicts after the middle interval; ours predicts after the first interval. Calibration and scoring rules differ. Both use the same later cohort records. No model was changed after this report was read.