Assessing hierarchy and balance in dynamic network models

This paper discusses two algorithms that compute important structural properties of dynamic social networks. The properties are hierarchy in a directed graph, and balance at a node in a signed directed graph. The hierarchy algorithm is used in a model that generates dominance structures in groups. The balance algorithm is part of a model that generates deviant behavior in a group. Both algorithms work with a dynamic network data structure that changes with each social event of the theoretical simulation model. These changes include: adding and deleting nodes; and adding, deleting, and changing the value of ties. To be useful, these algorithms must be efficient because hierarchy and balance occurs continuously in the running of the simulation models.