Northwestern University researchers have confirmed social balance theory through the use of statistical physics. Fritz Heider introduced this theory in the 1940s, explaining how humans naturally strive for harmony in their social circles. Previous studies attempting to confirm this theory using network science and mathematics fell short as networks deviate from perfectly balanced relationships. The Northwestern team combined two key elements crucial to Heider’s social framework – the fact that not everyone knows each other and that some people are more positive than others. By incorporating both constraints into their network model, the researchers successfully confirmed the famous theory after 80 years.
István Kovács, the senior author of the study, stated that the team needed to figure out the math to understand why social intuition works. By integrating constraints on connections and preferences of entities in a system, the researchers created a framework that could be applied to various systems beyond social networks. This breakthrough could help researchers comprehend social dynamics, including political polarization, international relations, and systems involving a mix of positive and negative interactions like neural networks or drug combinations. The study authors believe that their mathematical approach has broad applications and can offer significant insights into complex systems.
Social balance theory, as proposed by Heider, maintains that humans seek harmonious relationships within their social circles. Balanced relationships involve all individuals liking each other, while imbalanced relationships lead to anxiety and tension. The theory has been linked to extreme polarization, such as the political polarization seen today. Researchers had difficulty obtaining large-scale data that include information about both friends and enemies. By using signed network datasets from social scientists, Kovács and Hao were able to develop a network model that respected both constraints – who knows whom and the varying levels of positivity in individuals.
The network model created by Kovács and Hao utilized statistical models to assign positive and negative values to interactions based on the probability of occurrence. This approach kept the values random but within the constraints of the network topology. The researchers found that social balance theory consistently applied to large-scale social networks and extended beyond individual nodes to larger graphlets involving four or more nodes. By considering both constraints in their model, the researchers demonstrated that social networks align with Heider’s theory of social balance.
Kovács and Hao are exploring future directions for their work, including potential interventions to reduce political polarization. They believe that their model can be applied beyond social networks to understand interactions in complex systems like neuronal connections in the brain or drug combinations for disease treatment. By looking at different types of interactions beyond friendships, the researchers aim to gain further insights into the dynamics of various networks. The code and data behind their paper are available on Github for further exploration and analysis.
