Current closed-loop audit study with three-author attribution: Muxue Zhou, Zhen Xiao, and Xiaokai Xia. Includes the 3,000-fit observation study, the 2,200-fit baseline/reliability extension, independent Julia/R checks, and de-identified descriptive analysis data. The smaller independent batch did not reproduce the original previous-failure contrast sign; batches were not pooled.
The numerical content is unchanged from computational commit 52eddb7. All 11,180 source-sync entries and 17,792 archive manifest files passed SHA-256 verification. Julia 1.12.7 and R 4.6.1 apply to the current study.
Zenodo DOI reserved for the matching archive: https://doi.org/10.5281/zenodo.22743916 (publication being completed).
Software: MIT. Research products: CC BY 4.0. This is a public code/data release, not an anonymous reviewer package or a PsychArchives deposit. Historical versions are preserved.
Archive SHA-256: a131b4b6d76ca374062d53a8cca44a23ddbf84a3a112bf4f6c44c104df35e2a2