Consensus Analysis for Populations With Latent Subgroups: Applying Multicultural Consensus Theory and Model-Based Clustering With CCTpack

Abstract

Advanced consensus analyses are made available with the cultural consensus theory (CCT) framework, which is a model-based methodology used to closely derive the consensus knowledge of a group or population (e.g., a culture). The relevant data are their questionnaire responses about a subject domain. CCT has been established as a main methodology for ethnographic studies in social and cultural anthropology, and it has also been incorporated into other areas of the social and behavioral sciences as an effective information-pooling methodology. Recently, there are major advances in CCT, including (a) the multicultural extension, which can detect latent subgroups of a population, each with their own consensus answers to the questionnaire items; (b) the development of new models for several questionnaire designs, true/false, ordered category (Likert) scales, and continuous scales; (c) the estimation of important parameters that affect the response process; and (d) the development of Bayesian hierarchical inference for these CCT models. The joint analysis of these features positions such CCT approaches as some of the most advanced consensus analysis methods currently available. That is, they jointly (and hierarchically) estimate the consensus answers to the questionnaire items, the degree of knowledge (cultural competence) of each individual, the response biases of each individual, the difficulty (cultural salience) of each questionnaire item, and the subcultural group of the individual. In this article, we provide an overview of these major advancements in CCT, and we introduce CCTpack, which is currently the only software package available that handles all these extensions, especially (a) and (b).