Assessing and Improving Team Decision Making
Abstract
This project employed analytical and experimental techniques from signal detection theory to (a) assess the accuracy of team performance, (b) identify sources of inefficiency in team decision making, (c) specify how team members utilize information received from sources having different statistical properties, and (d) model the team deliberation process. The team's task was to decide whether signal-plus-noise or noise-alone had occurred, based on individual graphical displays presented to the team members. Team performance was shown to depend on (1) the signal-to-noise level of members' displays and the efficiency of individual member detection (compared to the statistical optimal), (2) the correlation (common noise) among members' displays, (3) constraints on member communication and interaction, and (4) how efficiently the team combined member judgments to form the team's decision (including mandatory voting rules). The internal correlation, expertise, and bias of different information sources influenced the decision weights that team members gave to these sources, sometimes in a non-optimal fashion. Using a Bayesian network model of team deliberation, the project began to quantify important aspects of the deliberation process such as the interaction protocols used by members of a networked team.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 2002
- Accession Number
- ADA399024
Entities
People
- Robert D. Sorkin
Organizations
- University of Florida