Researchers at the Institute for Futures Studies and Uppsala University have developed a new method for studying complex social processes. The method makes it easy to discover dynamical patterns and relationships in data. The software, implementing this method, is already available for free, and an article about the method is published today in the scientific journal PLoS ONE.
Social science aims to explain phenomena such as segregation, democratization, economic development and cultural change. In recent years, a lot of data describing these kinds of global changes have become available for research. Social science research increasingly requires systematic analysis of such data to identify important dynamics and interdependencies. Automating the process of discovering allows more potential theories to be tested.
The method, Bayesian Dynamical Systems Model, automatically identifies potentially complex relationships in how different phenomena co-evolve and effect each other. This makes it possible to find new and perhaps entirely unexpected patterns where data formerly looked as chaos.
All kinds of social systems can be studied using this method. You can for example look at the relationship between a country's GDP, infant mortality and education levels using World Bank data, or dynamic patterns of female employment within companies, using register data.
"The method is currently being applied at the Institute for Futures Studies to find important tipping points in diverse areas, such as countries' development towards democracy and the emergence of segregation in schools," says David Sumpter, professor of applied mathematics at Uppsala University and research director at the Institute for Futures Studies.
- Shyam Ranganathan, Viktoria Spaiser, Richard P. Mann, David J. T. Sumpter. Bayesian Dynamical Systems Modelling in the Social Sciences. PLoS ONE, 2014; 9 (1): e86468 DOI: 10.1371/journal.pone.0086468
Cite This Page: