SIFTing through Decisions: A Computational Model on Information Sharing

Abstract

This thesis explores a novel theory of how individuals decide to share information with their team, based on the characteristics of information (certainty, relevance), the team environment (shared knowledge, voice) and individual attributes (competence and warmth). Drawing from hidden profile research, the present work advances our understanding of team dynamics through an agent-based computational model. The SIFT model explores the emergence of knowledge within teams over time, which observes how teams cultivate, grow, and share information necessary to make optimal team decisions toward a task. Findings revealed that information distribution was the strongest predictor of unique information sharing: teams with shared access communicated more fully, while asymmetrical structures limited what was surfaced. Voice dynamics also played a central role in participation. Agents who spoke early, especially in low-voice environments—experienced positive feedback that increased their voice and led to disproportionate speaking opportunities over time. Conversely, others received fewer chances to contribute. While both certainty and voice influenced deliberation time, each had a distinct and sustained effect: low-certainty and low-voice conditions imposed greater cognitive demand throughout. Together, these findings suggest that knowledge emergence is shaped less by individual dispositions than by structural and social feedback loops. The model offers both theoretical and practical insights into when and why team members share, highlighting the importance of early voice development and information symmetry in promoting effective team decisions.