# MRI-003 — Findings supported by the current packet

- The 1905 publication separates Discussion from Written Communications. Shaw is a written contributor; his physical attendance is not established. Vickery is in Discussion. The 19-name roster is a publication roster, not an audience register. [S02](https://wellcomecollection.org/works/ds624dzu/items?canvas=11&page=2&shouldScrollToCanvas=true)
- Galton names desired qualities and permits different types. The supplied material does not establish a common validated measurement scale joining those qualities. His proposed family-record eligibility rule is explicit. [S01, opening argument and research program](https://galton.org/essays/1900-1911/galton-1904-am-journ-soc-eugenics-scope-aims.htm)
- The 25 curated seed records preserve different explanations and proposals. Wells’s objection to selecting successes coexists with his sterilization proposal; Hobhouse’s present restraint coexists with conditional future legislation. These examples do not establish frequencies or participant-wide ideological classifications. [Supplied seed records](../../War_of_the_Worlds_Part_I/data/proposition_seed.csv), [S01](https://galton.org/essays/1900-1911/galton-1904-am-journ-soc-eugenics-scope-aims.htm)
- Table I’s ten normal-probability counts reproduce after rounding. Table II’s explicit approximation and regrouping reproduce `[100, 150, 250, 250, 150, 100]` against `[89, 161, 250, 250, 161, 89]`, per 1,000. The distance 0.022 describes agreement after construction; it is not an inferential normality test. [S04, pp. 660–661](https://pdf.galton.org/essays/1900-1911/galton-1901-improvement.pdf), [executed checks](../../War_of_the_Worlds_Part_I/results/arithmetic_checks.json)
- Table III’s transcribed cells sum to its printed descendant margins, totaling 10,004 against 10,000 parental row-label counts. The modern literal implementation differs in 42 of 100 rounded cells, by at most 10 per cell. It is not an exact historical replication; the residuals do not establish deception. The modeled yield ratio approximately 142.857 is conditional on stipulated assumptions, not measured human value. [S04, p. 661](https://pdf.galton.org/essays/1900-1911/galton-1901-improvement.pdf), [residuals](../../War_of_the_Worlds_Part_I/results/table_III_literal_implementation_NOT_exact.csv)

The packet does not independently reconstruct Booth’s upstream data, validate civic worth, estimate later population or policy effects, establish motives, or demonstrate a universal social mechanism. Independent transcription review, complete corpus coding, exact Table III replication, and Figure/Table IV verification remain outstanding. [Packet §§3–4, §6.3](../../War_of_the_Worlds_Part_I/RESEARCH_PACKET.md)