Statistical methods for studying the evolution of networks and behavior

Studying longitudinal network and behavior data is important for understanding social processes, because human beings are interrelated, and the relationships among human beings (human networks) on one hand and human behavior on the other hand are not independent. The complex nature of longitudinal network and behavior data calls for probability models. Probability models are considered which assume that the evolution of networks and behavior is governed by continuous-time Markov processes. The continuous-time Markov models, which are implemented in the computer program Siena and which are becoming more and more widespread for studying longitudinal network and behavior data, are extended in important directions, and advanced statistical methods are proposed to infer from longitudinal network and behavior data to models. The model extensions and statistical methods are implemented in Siena