Investigating the effects of memory on agreement for homogeneous and heterogeneous social networks

Abstract

A collective is composed of interacting individuals that update their respective states, often to reach a decision. While social validation strategies such as majority game and minority game are widely known, memory and other self-validation strategies have not been studied as much. In this work, we use the Local Acceptance Model to study the effects of memory: (a) by varying the degree of memory in homogeneous populations, and (b) varying the proportion of agents with memory in heterogeneous populations. In both cases, increasing memory decreases agreement on the population, neighborhood, and individual levels. The optimum parameter values would depend on the context of the situation being modeled: for example, high agreement is desirable in voting while high memory is desirable in preserving culture.