Identifying Classes of Actors Through Variable Network Analysis

• Autor(a) principal: Rossoni, Luciano
• Data de Publicação: 2022
• Outros Autores: Gonçalves, Clayton Pereira, Gonçalves, Alex
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Mediações - Revista de Ciências Sociais
• Texto Completo: https://ojs.uel.br/revistas/uel/index.php/mediacoes/article/view/45723
• topic: análise de classes correlacionais; análise de redes sociais; redes de variáveis; significados compartilhados; relacionalidade; análisis de clase correlacional; análisis de redes sociales; redes de variables; significados compartidos; relacionalidad; Correlational class analysis; Analysis of social networks; network variables; shared meanings; relationality

Social network analysis involves identifying actors (people, companies, countries, among others) who are represented by nodes, and their respective relationships (friendship, communication, economic exchanges, etc.), represented by edges. The network can be segmented based on the presence or strength of relationships, forming groups of similar or cohesive actors. In many cases, we seek the similarity of actors not through social relationships but through sharing attitudes, beliefs, or opinions, which leads to the use of conventional variables. In this article, we propose using Correlational Class Analysis (CCA) to assess the Brazilian voter's political belief system, which, in turn, is captured through attitudinal variables. In our study, the method was used to identify classes of similar actors by evaluating their responses regarding four dimensions of political beliefs based on Converse's model. Based on a survey of 1,417 respondents, we identified three schematic classes (Group Interest, Ideologues, and Quasi-Ideologues) representing actors who share common beliefs about party politics. The method we present here contributes to a new way of assessing the similarity between actors considering shared beliefs or opinions, something limited to conventional methods of analyzing social networks.