W. S. Robinson's 1950 ecological-correlation paper names the statistical mechanism behind Garfield/Seglen's warning: a group-level aggregate wrongly substituted for an individual-level property

ecological-fallacymeasurement-theorygarfieldseglenstatisticsw-s-robinsonsource-discipline

W. S. Robinson's 1950 paper distinguishes an "individual correlation," computed on indivisible units (persons), from an "ecological correlation," computed on aggregated groups. Using 1930 US Census data, he shows the two need not agree: the ecological correlation between race and illiteracy, computed by geographic division, is .946 — roughly 4.7 times the individual-level correlation of .203 computed on the same underlying population. Robinson states "the purpose of this paper is to clarify the ecological" correlation problem, and notes that across the sociological literature of his day, "the substitution is made tacitly" — group aggregates standing in for individual properties without the substitution ever being argued for. He asks directly whether ecological correlations "can validly be used as substitutes for individual" correlations, and answers, in the paper's own two-word verdict: "They cannot."

This is the identical logical shape as Garfield/Seglen's warning, recorded in claim-garfield-seglen-within-journal-variance-undermines-individual-use: a journal's mean citation rate is a group-level (ecological) aggregate; an individual article's citation count is the individual-level property it is popularly, tacitly substituted for. Robinson's demonstration predates Garfield's bibliometrics warnings by two decades and Seglen's by four, and gives claim-garfield-seglen-and-steck-warnings-share-a-scalar-proxy-structure's "Garfield/Seglen leg" a named statistical mechanism — the ecological fallacy — independent of bibliometrics entirely. See entity-w-s-robinson and entity-ecological-fallacy.