Revisiting Dynamic Social Impact Theory: Extensions and Applications for Political Science
Traditional social science research tends to regard political attitudes as the product of individuals’ internal processes. This approach ignores the existence of external influences on individual attitudes, particularly that of social communication. We argue that political attitudes are neither confined to the individual, nor are they wholly social constructions. This study examines how political information is structured and communicated through a social medium. We employ multi-agent simulation to replicate and extend a previous simulation of attitude formation based on Latané’s (American Psychologist 36:343–356, 1981) theory of social impact. Specifically, our extension of the previous model incorporates motivated reasoning processes, attitude persuasiveness, and agent uncertainty in order to more accurately capture the pervasiveness of social communication. These changes allow us to model effects not previously possible, such as attitude polarization and information cascades. Results from this study show that realistic structures in the distributions of attitudes can be recreated using prevailing aggregate- and individual-level models of attitude formation. By formally modeling the dynamics of social impact, researchers gain insight into social influences on individuals’ attitudes and the consequences for public opinion.
International Journal of Politics, Culture and Society
Digital Object Identifier (DOI)