How Software Agents Can Help to Coordinate Emergency Response Teams: Adaptive Team Performance Comparing Manual and Automated Team Communication

• Published in: Journal of business and psychology Vol. 38; no. 5; pp. 1121 - 1137
• Main Authors: Müller, Rebecca, Graf, Benedikt, Ellwart, Thomas, Antoni, Conny H.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2023; Springer Nature B.V
• Subjects: Automation; Behavioral Science and Psychology; Business and Management; Communication; Community and Environmental Psychology; Industrial and Organizational Psychology; Knowledge; Original Paper; Personality and Social Psychology; Psychology; Social Sciences; Teams; Team communication; Software agent; Transactive memory system; Team performance; Meta-knowledge
• ISSN: 0889-3268; 1573-353X
• DOI: 10.1007/s10869-022-09858-4

In interprofessional emergency response teams, firefighters, police, and paramedics must communicate efficiently (i.e., request the correct expert) to avoid life-threatening consequences. However, this communication is sometimes inefficient, for example, when a wrong expert is requested due to the lack of meta-knowledge. Team research has shown that meta-knowledge of “who knows what” improves team communication, so that members correctly request each other according to their expertise. Advances in technology, such as software agents holding meta-knowledge, can be used to improve team communication. In this paper, we analyze the effects of meta-knowledge on expert seeking, mistakes in requesting experts, and (adaptive) team performance by comparing manual and automated agent-based team communication. Using a control-center simulation, 360 students in 120 three-person teams had the interdependent task of handling emergencies in three phases. We manipulated meta-knowledge in advance, with 61 teams learning and 59 teams not learning other team members’ expertise. Furthermore, in phases 1 and 3, team members had to communicate manually. In phase 2, communication was automated by a software agent taking over expert requesting. In line with our hypotheses, results showed that software agents can compensate the lack of meta-knowledge, so that there were no performance differences between teams with and without meta-knowledge with automated team communication. Our findings provide implications for research and practice that established team constructs should also be considered in human-automation teams.