Averaging multidimensional concept scores contradicts the necessary-and-sufficient structure scholars define

Evidence

Wuttke and his colleagues (2020) provide a very nice recent example of this kind of analysis. They discuss the currently very important topic of populist attitudes and more generally the concept of populism. As with democracy they note that it is typically conceptualized as a multidimensional concept using necessary and sufficient conditions: “Most scholars now concur that populist attitudes denote a multidimensional construct, composed of two or more concept components” (p. 1) and that “populism scholars uphold that the distinct conceptual characteristic of populism at the mass level is the concept’s status as an attitudinal syndrome, which is defined in terms of necessary and sufficient conditions, more specifically by the simultaneous presence of the concept’s constituent components.” However, in terms of empirical operationalization, “scholars obtain individual populism scores by computing weighted or unweighted averages across the concept dimensions. More specifically, scholars aggregate the concept attributes manually by computing mean values or using data-driven approaches such as factor analysis” (p. 2). They then explore the nontrivial empirical and statistical consequences of constructing a populism measure faithful to the semantic structure of the concept. — Gary Goertz, Social Science Concepts and Measurement, 939

Categorization

• XREF: Connects to work on measurement validity in survey research, e.g. composite indices of democracy or populism that weight components equally by default.
• SEED: Angle: ‘Your Index Is Lying About Your Theory’ — how routine mean-scoring of multidimensional constructs silently discards the concept’s own logic.