Variety-seeking, learning and performance

Area of Science:

• Organizational Learning
• Computational Social Science
• Knowledge Management

Background:

• The variance hypothesis posits that variety-seeking (exploration) is crucial for learning and performance.
• Existing computational models show diverse knowledge sources improve problem-solving.
• Precise guidelines on the optimal breadth of search are lacking.

Purpose of the Study:

• To determine the conditions under which variety-seeking benefits organizations.
• To elucidate the trade-offs between exploration and exploitation in organizational learning.

Main Methods:

• Development of a computational model simulating individual learning through interaction.
• Analysis of how learning dynamics are influenced by the performance landscape.
• Modeling knowledge updates based on interactions with other individuals.

Main Results:

• Variety-seeking is beneficial only when learning targets provide informative signals about the performance landscape.
• Superior knowledge may be found locally, near an individual's current position.
• A critical point exists beyond which variety-seeking sharply declines performance, dependent on landscape and knowledge specialization.

Conclusions:

• The benefits of variety-seeking are contingent on the learning environment's characteristics.
• Understanding the critical point of exploration is key to managing its costs and benefits.
• Findings offer implications for setting optimal boundaries for organizational exploration.