CRII: HCC: Temporal dynamics of team emotion and cognition in AI-supported decision-making

Artificial intelligence (AI) is becoming increasingly present and powerful in systems that affect people’s lives, helping people make decisions in contexts ranging from market decisions, military task support, and medical diagnosis. It is thus essential to understand how adding AI aspects to the tools teams use influences team collaboration, and its implications for organizational practices.

This project will address those questions through studying how affective and cognitive processes in teams with AI support intertwine to shape team decision-making performance over time. Insights produced by this research will improve existing knowledge about how AI can be applied to facilitate human collaboration. It will help organizations in various industries make informed decisions about whether and when to adopt AI tools, and develop training and interventions to improve AI-supported team decision-making.

The research will also help AI designers better understand AI’s impact on team decision-making and to design more effective AI tools to support collaboration.

Bridging research perspectives and practices from human-computer interaction (HCI) and team research, this project seeks to develop a human-centered theory on how AI may shape team decision-making processes and outcomes through a set of lab experiments. In the experiments, teams will perform a series of decision-making tasks with AI assistance. The first experiment will vary the communication features of an AI system that provides information and ideas to a team decision-making task, examining how both the presence and the communication style of the AI system affects team interaction processes and decision-making outcomes.

The second experiment will vary when AI assistance is employed in different team collaboration stages – for example, whether AI assistance was introduced at the beginning of the team task or in the middle of the team task. The third experiment will investigate how team turnover during a decision-making task interacts with the use of AI-based tools.

Team information sharing, affective status, learning, and performance during the tasks will be observed and analyzed using computational methods. Understanding these processes will contribute to theory on AI-supported team decision-making and in particular, the emergence of team emotion, cognition, and their inter-relations under different conditions of using AI-based tools.

The analytical approach can help develop unobtrusive measures of team emotion and cognition and create new research paradigms, measures, and materials (such as task designs and customized chatbots) that can be used by the broader scientific community.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.