Copy or collaborate? How networks impact collective problem solving

Area of Science:

• Organizational behavior
• Network science
• Computational social science

Background:

• Effective collaboration is crucial for complex problem-solving in diverse fields.
• Optimal communication network structures for group problem-solving remain debated, with conflicting findings on exploration vs. exploitation strategies.
• Existing models often overlook active collaboration, focusing solely on information copying.

Purpose of the Study:

• To investigate the impact of network topology on group problem-solving performance.
• To differentiate between passive copying and active collaboration strategies within group dynamics.
• To determine the optimal balance between network efficiency and information diversity for problem-solving.

Main Methods:

• Development of a computational model simulating group problem-solving.
• Inclusion of both copying and active collaboration mechanisms within the model.
• Analysis of how different network structures influence outcomes based on the prevalence of copying versus collaboration.

Main Results:

• Groups perform worse when relying solely on copying compared to independent work.
• The optimal network structure is contingent on the balance between copying and collaboration.
• Exploration-focused (inefficient) networks are superior when copying dominates, whereas efficient networks excel when collaboration is primary.

Conclusions:

• Active collaboration is a key factor in enhancing group problem-solving beyond simple information copying.
• A 'collaborate first, copy later' heuristic in efficient networks can leverage benefits of both strategies.
• Findings provide insights for organizational design to optimize problem-solving performance.