An Evolution Model of Group Opinions Based on Social Judgment Theory (August 2018)

We integrate social judgment theory and bounded confidence to build an agent-based model for exploring how group opinions form and evolve. Each agent in our model, which is an extended Hegselmann–Krause model, has its own personality (openness and vacillation) and emotion (opinion). In a homogeneous case, we find that agents’ opinions will preserve their initial order. Once a crack forms between two agents at time
$t$
, then a crack will exist after that time. Despite the fact that the continuous trust function results in a zero infimum of weight, the opinion evolution system will converge. Simulation shows that an increased level of either openness or vacillation can cause the group to reach a consensus easily. Specifically, the higher the level of openness of an agent, the greater the amount of clusters; and the higher the level of vacillation of an agent, the less time it will take for group opinions to converge. Although the initial opinions are randomly distributed, agents with the same levels of vacillation and openness tend to be concentrated during the evolution process.