CCPV Logo

Applying Agent-based Simulation Software and Sociological Theory to Decision Making

INTERDISCIPLINARITY: INTEGRATING INSIGHTS

Utilizing diverse methods and bodies of knowledge - such as area studies expertise, sociological theory, computer simulations and insights about human behavior gained in economic experiments - requires true interdisciplinarity: Therefore, in all our working groups people from at least two different backgrounds come together, discussing and translating concepts to reach a joint product that none of them could have reached alone.

For example, in our simulation team, the social scientists provide the theoretical models of cultural change, boundary formation and collective action. This secures that the assumptions underlying our simulation are not arbitrarily chosen, but are based on socio-psychological, sociological and economic theories and findings. Furthermore, to provide our simulation with realistic empirical data, area experts work closely with the computer scientists to make their expertise and qualitative data compatible with the simulation's needs.

Another prime example is the development of our integrated content- and network analysis web crawler: Here, we operationalize social network analysis and content analysis methodology to automatically characterize the structure of virtual communities on the web. As they are social phenomena and the basis of increasing cultural, political and economic activity - just consider their importance in recent election campaigns - using sociological, psychological and cultural theories is of prime importance when we want not only to understand the technical, but the social aspects of these communities. Integrating these theories into a powerful web crawler allows us to answer questions that could not addressed by conventional social science or purely technical analysis alone.

Not only require most to work in interdisciplinary teams, but there are also a lot of interdependencies between the tasks that are illustrated in the diagram below: