Answers are grounded in this paper and cite their sources.
This paper attempts to rectify false suspicions that have developed among some regarding ratio correlations. If researchers are theoretically concerned with relationships between ratio variables that contain common components (e.g., population density and suicide rate) or between a ratio variable and one of its components (e.g., percent urban and population size of nations) correlation analysis will not normally yield spurious results. Spuriousness becomes a problem primarily when the researcher is interested in relationships between component variables which are standardized by a common term for reasons extraneous to the hypothesis. Even here, though, it is not the empirical correlation that is spurious, but possible inferences one might draw about the relationship between the component variables from the observed correlations between the ratio-standardized variables. Assessment of suggested alternatives to ratio correlation indicates that ratio measurement sometimes provides conceptual, interpretive, and statistical advantages over alternatives critics have proposed.
No takes yet. Share an insight, caveat, or question.
Kasarda et al. (1979) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: