The ofness and aboutness of survey data: Improved indexing of social science questionnaires

Abstract

In this paper we adopt a user-centered indexing perspective to propose a concept for the subject indexing of social science survey data. Operationalization processes in survey development mean that the constructs being studied are mostly hidden in the verbalization of the questionnaire (latent subject content). Indexable concepts are therefore found at two different semantic levels that, inspired by research on the indexing of pictures (Shatford, Cat Classif Q 6(3):39–62, 1986), we treat as the ofness and aboutness of survey data. We apply a syntax of term linking and role indicators, combining directive terms (e.g., attitude, experience, perception) with subject terms (e.g., corruption, foreigners). Each directive and subject term combination represents a retrievable unit of interest to the secondary researcher.