Designing Like a Human: Simulating Cognitive Style in Teamwork with an Agent-Based Model

Abstract

This thesis describes the development of an agent-based model for simulating cognitive style in the context of collaborative problem solving. Cognitive style describes the diverse ways in which people solve problems. Individuals’ cognitive styles can impact the success or failure of a design team. However, the effects of cognitive style in collaborative problem solving are not well understood. To address this gap, this thesis presents KABOOM (KAI Agent-Based Organizational Optimization Model), the first agent-based model of teamwork to incorporate cognitive style. In this thesis, experiments using KABOOM investigate the interacting effects of a design team’s communication patterns, specialization, and cognitive style composition on a team’s performance. Testing the model with a race car design problem reveals that teams can strategically leverage diversity of cognitive style to improve performance. By simulating cognitive style and team problem solving, KABOOM lays the groundwork for the development of team simulations that reflect humans’ diverse problem-solving styles.