Spatializing social networks: making space for theory in spatial analysis

This study is a quantitative and spatial analysis of the gang-related violence in a section of Los Angeles. Using data about the spatial distribution of gang violence in three neighborhoods of Los Angeles, this research first adopts a deductive approach to the spatial analysis of gang violence by spatial regression models that considers the relative location of the gangs in space while simultaneously capturing their position within a social network of gang rivalries. Several models are constructed and compared and the model that seems to best fit the observed geography of violence is one in which both the territorial geography and the social geography of the gangs is utilized in the autocorrelation matrix. Building on the findings of the spatial regression modeling, the concept of social position and associated techniques of structural equivalence in social network analysis is then explored as a means to integrate these different spatialities. The technique of structural equivalence uses the two different spatialities of embeddedness to identify gangs that are similarly embedded in the territorial geography and positioned in the rivalry network which aids in understanding the overall context of gang violence. The importance of theory to guiding spatial regression modeling is demonstrated by these findings and the hybrid spatial/social network analysis demonstrated here has promise beyond this one study of gang crime as it operationalizes spatialities of embeddedness in a way that allows simultaneous systematic evaluation of the way in which social actors??? position in network relationships and spatial settings provide constraints and possibilities upon their behavior