Population Wide Attitude Diffusion in Community Structured Graphs

Attitudes are “general and relatively enduring evaluative responses to objects” where objects can be “a person, a group, an issue or a concept” (Visser and Clark 2003, Page 1). Attitudes are shown to have an impact on, and can sometimes predict, the behaviors of individuals (e.g., voting behavior (Krosnick 1988), and consumer purchases (Jones and Fazio 2008; Fazio and Olson 2003)). In addition, attitudes can change based on the influence of others (Visser and Clark 2003). Understanding population wide attitude change is an important step to understanding the behavior of societies. For instance, consider the change in attitudes towards global warming and the environment that has resulted in a significant change in public policy and national priorities (Change 2009). We are interested in understanding the following four aspects of population wide attitude change: • Attitude strength is continuous – individuals can vary from strongly negative, to strongly positive attitudes. • Attitudes are non-progressive – individuals can have a positive attitude, then a negative attitude, then a positive attitude again. • Attitudes are linked – different attitudes influence each other – i.e, if you believe in environmentalism it is also likely you believe in recycling. • Individuals have a level of cognitive effort – there is some level of effort that individuals will exert in changing their attitudes. • Community structured social networks – oftentimes social networks contain communities; strongly connected subgraphs that are only weakly connected to the rest of the graph.