Bounded personality diversity boosts collective intelligence: a swarm optimization approach.

Introduction: This study presents an initial computational test of bounded diversity advantage theory (BDAT), which posits that bounded personality diversity enhances emergent collective problem-solving, using a Particle Swarm Optimization (PSO) testbed. Methods: The current PSO instantiation includes a novel monitoring approach to quantify group agreement dynamics. Focusing on conscientiousness-like behavior, operationalized as industriousness and persistence in swarm agents through adaptive inertia, six diverse and non-diverse swarm compositions were compared across 11 two-dimensional mathematical search functions. Swarm outcomes were evaluated using a Performance Efficiency Score (PES), which balances error (fitness) and iterations (time). Results: Diverse compositions outperformed non-diverse compositions, consistent with the present operationalization of BDAT-derived predictions concerning bounded diversity and premature closure. Monitoring asymmetry in polarized swarms at the first convergence threshold (50%) further indicated that highly industrious agents disproportionately drove early convergence while low-industriousness agents remained more exploratory, a process pattern also consistent BDAT predictions. Discussion: These findings suggest that PSO can serve as a useful computational testbed for examining some BDAT mechanisms, while broader validation of BDAT will require additional research. The findings bridge personality psychology, group processes, and evolutionary perspectives, offering a framework to explore how personality diversity fosters collective intelligence.