korrelativ

Korrelativ is a term used in statistics, data science, and philosophy of science to describe the idea that the observed strength and even direction of a relationship between two variables can depend on the context in which it is measured. This includes factors such as the measurement scale, time window, sampling method, or data transformation. Korrelativ emphasizes that correlation is not an invariant property of a pair of phenomena but can vary with the reference frame or methodological choices.

Origin and usage: The word is a portmanteau of korrelatie (Dutch for correlation) and relativism, and it appears mainly in discussions about contextual or frame-dependent analyses. It is not part of formal statistical nomenclature, but is used in peer discussions and methodological critiques to foreground the role of analytical decisions in shaping observed relationships.

Concept and examples: In finance, two assets may show positive correlation over daily returns but negative correlation over monthly returns. In psychology or education, the correlation between test scores and performance may change with different scoring normalizations. Korrelativ highlights that such variations reflect different analytic lenses rather than contradictions in the underlying phenomena.

Relation to causation and inference: Korrelativ reminds researchers that correlation magnitude alone cannot establish causation and must be interpreted alongside data processing steps, reference frames, and the chosen variables.

Limitations and critique: Critics warn that a broad or implicit use of korrelativ can obscure robust findings by focusing on context-dependent results. Clear definitions and transparent reporting of data handling are recommended.

See also: correlation, contextual statistics, data transformation, measurement scale, statistical framing.