Title

An Agent-Based Simulation For Investigating The Impact Of Stereotypes On Task-Oriented Group Formation

Keywords

Group formation; Multi-agent social simulations; Stereotypes

Abstract

In this paper, we introduce an agent-based simulation for investigating the impact of social factors on the formation and evolution of task-oriented groups. Task-oriented groups are created explicitly to perform a task, and all members derive benefits from task completion. However, even in cases when all group members act in a way that is locally optimal for task completion, social forces that have mild effects on choice of associates can have a measurable impact on task completion performance. In this paper, we show how our simulation can be used to model the impact of stereotypes on group formation. In our simulation, stereotypes are based on observable features, learned from prior experience, and only affect an agent's link formation preferences. Even without assuming stereotypes affect the agents' willingness or ability to complete tasks, the long-term modifications that stereotypes have on the agents' social network impair the agents' ability to form groups with sufficient diversity of skills, as compared to agents who form links randomly. An interesting finding is that this effect holds even in cases where stereotype preference and skill existence are completely uncorrelated. © 2011 Springer-Verlag Berlin Heidelberg.

Publication Date

3-14-2011

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume

6589 LNCS

Number of Pages

252-259

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/978-3-642-19656-0_36

Socpus ID

79952406100 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/79952406100

This document is currently not available here.

Share

COinS