Cultural consensus theory: Aggregating expert judgments about ties in a social network

Abstract

This paper describes an approach to information aggregation called Cultural Consensus Theory (CCT). CCT is a statistical modeling approach to pooling information from informants (experts, automated sources) who share a common culture or knowledge base. Each informant responds to the same set of questions, and the goal is to estimate the consensus knowledge of the informants. CCT has become a leading methodology for determining consensus beliefs of groups in the social sciences, especially cultural and medical anthropology. The paper illustrates CCT by providing a model for aggregating expert judgments about ties in a social network. Expert sources each provide a digraph on the same set of nodes, and the CCT model is used to estimate the most likely digraph to represent their shared knowledge. © Springer Science + Business Media, LLC 2009.