Skip to main content
Infrastructure Resilience Conference 2018

Full Program »

Managing Information Overload: Simulation of Information Sharing Strategies in Groups

Good decision making relies on high-quality, context-specific information, where information is data that is well framed and meaningful, such that it can support a decision. The available sources and streams of data have been growing exponentially, which raises the question how individuals and organizations aggregate data that is relevant and create a group intelligence that is more than the sum of its parts. It is clear that groups vary in their ability to extract and utilize unique information from individual group members, but the underlying processes that allow groups to cultivate this ability are poorly understood. Because these underlying processes are not easily manipulated in laboratory or field experiments, we use an agent-based model to systematically explore the interaction between individual-level cognitive factors (how people weight new information relative to old information) and group-level social factors (what information people in the group share, and how equal sharing is within the group). Furthermore, this paper identifies the relation between (1) number of interactions, (2) interaction dominance of one agent, (3) number of information units; based on the assumptions that (a) an agent updates her opinion based on new information, (b) the emergence of group intelligence depends on the updating of individual agents and the degree of information sharing. Preliminary results indicate that when people randomly share information, one unit of information will dominate the communication, which leads to poorer overall decision quality. Future modeling work may build on these insights to explore the effects of inaccurate information, and future experimental work might explore whether groups can be trained to interact in the optimal configurations identified in this paper.

Dirk-Jan van Veen
Singapore-ETH Centre

Ravi S. Kudesia
Singapore-ETH Centre - Future Resilient Systems


Powered by OpenConf®
Copyright ©2002-2016 Zakon Group LLC