Modeling the joint dynamics of relational events and individual states

Networks evolve at multiple levels; edges or relations may change, and the characteristics of the individuals within the network may change as well. Often these processes are intertwined, and in order to study them, statistical models must be developed that account for coevolution of multiple network components. The increased availability of continuous time network data has prompted new models for network inference such as the relational event framework. While network data may be continuously observable, the characteristics or state of an individual can still only be observed periodically. This type of panel data can be readily analyzed using actor-oriented models, but these methods do not accommodate continuous network data. We propose a model that integrates the relational event framework with actor-oriented models for behavioral change, allowing us to model the joint dynamics of relational events and individual states. This composite model preserves the advantages of each method, while leveraging the richer information available in relational event data. We apply our model to datasets collected from virtual team experiments to highlight the utility of our method.