Quantitative analysis of group decision making for complex engineered systems

Abstract

Understanding group decision-making processes is crucial for design or operation of a complex system. Unfortunately, there are few experimental tools that might contribute to the development of a theory of group decision-making by committees of technical experts. This research aims to fills this gap by providing tools based on computational linguistics algorithms that can analyze transcripts of multi-stakeholder decision-making entities. The U.S. Food and Drug Administration medical device approval committee panel meetings are used as a data source. Preliminary results show that unsupervised linguistic analyses can be used to produce a formal network representation of stakeholder interactions.