Deriving concept categories from data distributions is generally methodologically unsound

Evidence

These four categories are based on quartiles, and thus are a linear transformation based on the empirical distribution of the data. This is something that this volume generally does not recommend, i.e., conceptualization completely based on the empirical distribution of the data. — Gary Goertz, Social Science Concepts and Measurement, 762

Categorization

• QUESTION: If empirical-distribution cutoffs are rejected, what should replace them — theoretical thresholds, expert judgment, or case knowledge? Worth chasing in later chapters.