Multidisciplinary learning through collective performance favors decentralization

Area of Science:

• Organizational Behavior
• Team Dynamics
• Learning Sciences

Background:

• Traditional team learning models assume concurrent learning or solution sharing, which is often unfeasible in multidisciplinary settings.
• In multidisciplinary teams, distinct roles obscure individual contributions, complicating performance attribution and learning from collaborators.
• Mediating artifacts are crucial for enabling effective learning in complex team structures.

Purpose of the Study:

• To investigate how team network structures influence learning and performance in multidisciplinary teams.
• To identify mechanisms through which individuals can attribute performance in interdependent tasks.
• To provide design principles for optimizing team learning in specialized environments.

Main Methods:

• Simulated team environments with varying network structures and task types.
• Analysis of individual and collective performance metrics under different network conditions.
• Examination of learning dynamics influenced by exploration and exploitation search strategies.

Main Results:

• Team network structure impacts performance, particularly when contributions are weighted by network properties.
• Dense networks negatively affect performance during exploration (innovation) but positively impact exploitation (refinement).
• Decentralized team structures consistently enhance performance across diverse tasks.

Conclusions:

• Mediating artifacts, such as collective performance assessments, enable effective learning from network neighbors in multidisciplinary teams.
• Optimal team network design depends on the task's emphasis on exploration versus exploitation.
• Decentralization is a robust strategy for improving team performance in complex, interdependent work settings.