Blind computational reproduction of published statistical analyses with MerQur — Study 3: eleven further analysis families and recurrence of corrected errors (preregistration)

> This document preregisters Study 3 of a blinded computational reproduction programme for MerQur, a free > statistical software package. Twenty-two articles published in 2025-2026 with openly licensed raw data (CC0 or > CC BY 4.0), two in each of eleven analysis families not or only marginally tested in Studies 1 and 2, were > selected before registration. Their analyses will be re-run, blind to the published results, by three > independent analysts using MerQur 2.4.1 (frozen at code commit b35ff79da354f93fb2b0d64f9ed6778ff4eb3e62), R and > SAS, and every reported value will be classified with a predefined decision tree. Study 3 follows the design and > rules of Studies 1 and 2 (OSF https://osf.io/bj43h; Zenodo https://doi.org/10.5281/zenodo.23858471). A > secondary, exploratory question asks whether errors confirmed in the earlier studies and corrected in 2.3.2 > recur on independent data. No analysis had been run on the study data at the time of this deposit; SHA-256 > hashes of the frozen analysis specifications are given in Appendix A.