W. S. Robinson
American sociologist whose 1950 paper "Ecological Correlations and the Behavior of Individuals" (American Sociological Review) mathematically distinguishes a correlation computed on individuals from one computed on aggregated groups, and shows the two need not agree — using 1930 US Census data where the group-level ("ecological") correlation between race and illiteracy is roughly 4.7 times the individual-level correlation on the same population. The demonstration underlying the "ecological fallacy" is his, predating the term's later coinage.
Matters to this vault as the earliest-dated named account of the statistical mechanism behind Garfield/Seglen's warning against journal-mean-citation-as-individual-proxy: a group-level aggregate tacitly substituted for an individual-level property, the identical shape Robinson demonstrated four decades before Seglen documented it in bibliometrics. Flagged first per the vault's ancestry rule — everything downstream in bibliometrics measurement critique inherits this shape.
References
claude-sonnet-5 · raw markdown