'The mutual influence between relations on one hand, and individual behavior, performance, or attitudes on the other hand, is two-sided.
(1.) The networks - to - behavior influence can generally be understood from the crucial social opportunities and constraints inherent in networks of relations between social actors, and which can be influential for the behavior and performance of the actors. E.g., the well-being of individuals in their social context is conditioned not only by their individual characteristics and behavior but also by their social ties; attitudes and behavior of individuals often are influenced by their peers; the economic production of goods and services is conditioned by the networks between firms as well as by networks between individuals inside the firms. Empirical analysis is complicated because this influence leads to dependencies between actors which should be expressed in some way in the statistical model.
(2.) The behavior - to - network influence means that the selection of others as interaction partners depends strongly on characteristics (such as perceived abilities and attitudes) and behavior of the actors involved, homophilous selection being just one example. This influence can be understood from the differential costs and benefits of different ties, depending on actors' characteristics. Empirical analysis is complicated because selection of interaction partners often also depends on the existing network structure. The most fruitful way to study this mutual feedback between relational networks and individual performance is in dynamic settings. This will - hopefully! - contribute to explanations of collective phenomena such as clustering of behavior and attitudes in cohesive groups of various kinds, various types of collective action, or the genesis and maintenance of norms. For longitudinal (panel-type) data on complete networks and actor attributes, recent developments in statistical methodology have been incorporated in the SIENA program, which enable the statistical analysis of such types of data. The aim of the session is to have papers treating questions of the following type, preferably in combination: How do characteristics (including abilities and attitudes) and behavior of individual actors influence choices concerning transaction and relation partners, and thereby the network structure? How do the network of relations between actors, and the transactions embedded within this network, influence the actors' characteristics (attitudes, etc.) and behavior? What are the consequences, at the collective level, of this feedback between individual characteristics and behavior and network structure? The session is meant to be on the interface between mathematical modeling, statistical modeling, and empirical sociology, and papers are requested to combine empirical data with mathematical or statistical modeling. Researchers who are interested to present in this session are requested to send me an abstract, or contact me, before November 30, 2004, by email. If you wish to present a paper in the session "Networks and Behavior", send an abstract of no more than one page to the Tom A.B. Snijders: t.a.b.snijders@ppsw.rug.nl.