The Agent-Based Model of Pluralistic Ignorance. Sincerity, Conformity and Truth.

Abstract

This thesis investigates the phenomenon of pluralistic ignorance, clears up some conceptual confusion regarding its definition, and explores ethical problems connected to it. A primary contribution lies in the development of an agent-based computer simulation of pluralistic ignorance. Observations of this simulation lead to tentative recommendations about how to deal with pluralistic ignorance in social networks. Pluralistic ignorance involves a systemic discrepancy between individuals' private and publicly stated beliefs. This happens because individuals mistakenly perceive widespread acceptance of a proposition within their social group. This leads them to conceal their sincere beliefs and to conform to the perceived majority belief in public. Since everyone does so, it looks like the whole society accepts a certain proposition, when in fact no one (or almost no one) does. Through agent-based simulations, this thesis investigates interventions to dispel pluralistic ignorance. It explores the impact of introducing into pluralistically ignorant social networks invariably-sincere agents (sincerians) and contrarians, revealing their potential to change the behavior of insincere agents and, under certain conditions, help others to share their beliefs sincerely.