Answers are grounded in this paper and cite their sources.
Recent work in the sociology of culture takes a relational approach to social attitudes under the logic that meaning inheres in the relationships among attitudes within a domain. Relational class analysis made this idea tractable by mapping shared construals, or common ways of organizing a domain of beliefs. However, construals are just one part of the relational structure of an attitudinal domain. Respondents occupy positions within construals, and those positions in turn may be aligned, opposed, orthogonal, central, or peripheral to one another. Existing methods capture aspects of this structure, but positions, construals, and their relations are usually identified separately. This article introduces topological relational class analysis (tRCA), a method that recovers all three within a single representation. tRCA identifies positions in the respondent–item space and uses the signed relations among those positions to map both construals and the topology of an attitudinal domain. Simulations show that tRCA outperforms latent class analysis (LCA) in recovering positions, outperforms relational class analysis (RCA) in recovering construals under multigame conditions, and accurately recovers a range of different topologies. A re-analysis of 2016 American National Election Study (ANES) social group attitudes data shows that tRCA recovers two largely independent games, one organized around partisan identity and another around attitudes toward racial groups, while simultaneously decomposing each into distinct positions. tRCA thus builds on Paul DiMaggio’s relational approach to culture and cognition by using survey data to map population heterogeneity, patterns of shared meaning, and the relational topology among attitudes.
No takes yet. Share an insight, caveat, or question.
Ramina Sotoudeh (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: