Modelling the structure and dynamics of network-based social systems

Abstract

The development of policy interventions to enhance the sustainability of our future often depends on the capacity to understand and model relevant social systems and to predict the consequences of policy change. To this end, we describe in this paper a suite of statistical models for the structure and dynamics of network-based social systems. In these models, global network structure is hypothesised to arise as the outcome of dynamic, interactive processes occurring within local neighbourhoods of a network. A hierarchy of such processes and model specifications are described. Models can be estimated from a number of possible data structures, including single or panel observations, and partial data structures obtained through certain types of network sampling schemes. Using several illustrative applications, we describe how the models can be used to extend our understanding of social systems and build theoretically plausible and empirically-grounded simulation models. Advantages of the approach include a capacity to quantify uncertainty and its implications, and the facility to assess the extent to which a model captures features of the social system not parameterized in the model. We conclude with a sketch of future work designed to improve our understanding of ways to enhance sustainability within network-based social systems.