A Computational Approach to Examining Team Coordination Breakdowns During Crisis Situations

Abstract

During crisis situations, teams are more prone to coordination breakdowns that are characterized by a temporary, diminished ability to function effectively as a team. However, team research currently lacks robust approaches for identifying transitions from effective team functioning to coordination breakdowns. With the current study, we aimed to develop such robust approaches, and to deepen our understanding of how team coordination dynamics across various physiological signals reflect coordination breakdowns. Consequently, we used audiovisual data from four-person teams involved in a stressful collaborative game task to manually identify coordination breakdowns. Next, we set out to computationally identify coordination breakdowns by applying continuous measures of team coordination (windowed synchronization coefficient and multidimensional recurrence quantification analysis) to photoplethysmogram and electrodermal activity data obtained during the task, and identifying transitions therein with change point and nonlinear prediction algorithms. We found that our computational coordination breakdown identification approaches can identify up to 96% of the manually identified coordination breakdowns although our results also show that the precision of our approaches falls far behind. Our findings contribute theoretically and methodologically to the systematic investigation of coordination breakdowns, which may ultimately facilitate the support of teams in responding to and mitigating negative consequences of crisis situations.

Examining Physiological Features Underlying Team Coordination Breakdowns

“…The current study is a secondary analysis of an existing dataset 10 . An extensive description of this data, including details less relevant to this study, can be found on the Open Science Framework (OSF; https://osf.io/t9eym).…”

“…Three raters identi ed the TCBs in the audiovisual recordings of participants combined with the game recordings 10 . Interrater reliability was found to be good (Cohen's Kappa = .67.…”

“…Previous studies have shown that TCBs can be identi ed with team coordination dynamics [8][9][10] . The team coordination dynamics approach combines principles from Complexity Science and Cognitive Science, to examine patterns of coordination, and how those patterns form and change over time 5,11 .…”

“…This coordination can be measured across multiple signals, and with a variety of methods. Different combinations of such signals and methods may capture unique team cognitive processes and states 6,10,19 . For example, team coordination of upper-body movements 20 and heart rate variability 21 were found to re ect self-rated performance, while inter beat interval 22 and skin conductance 23 coordination re ected externally rated performance.…”

“…The current study builds on previous work 10 to examine the underlying structure of transitions in physiology-based team coordination that do and do not relate to TCBs. More speci cally, the aim of the current paper is to determine what features are relevant for computationally identifying TCBs more precisely.…”