Supported degree: a new network measure
Since interaction between people exists at all levels of human activity, understanding how the patterns of interactions shape behavior and performance of network members is a key question across social sciences. This thesis introduces a new measure of individual network positioning which we denote as supported degree that reflects both local centrality of an individual in her network and the cohesiveness of her network neighborhood. We characterize this measure mathematically, propose an algorithm that allows to measure supported degree from the data and compare the ability of supported degree to explain a series of behavioral socio-economics outcomes vis-a-vis standard measures of individual local positioning. We show that supported degree is a good predictor of a series of individual socio-economics characteristics and explains them as well as the degree, a classic measure of local centrality and considerably better than the clustering coefficient, the standard measure of network cohesion.