This dissertation analyzes a study that sought to identify the consequences of variations in how gangs are conceptualized and measured at an ecological level, with the hope of building researchers’ understanding of how gangs link to crime, how community level dynamics work to foster or prevent gang activity, and how to reduce and prevent gang problems.
This dissertation reports on a study to identify the consequences of variations in how gangs are conceptualized and measured at an ecological level; the author designed this study with the idea that variations in gang models such as indicators, spatial scales, and levels of measurement, may generate disparities in empirical results as well as in estimates of gang ecologies without a clear understanding or rational for of the implications of their selection. The author addresses two research questions: if indicators of gang ecologies identify gangs in similar ways and with consistent results across spatial scales; and if indicators can predict the presence or absence of gangs as well as predict continuous gang outcomes, such as the number of gangs present or the geographic size of the gang ecology, with consistent results across spatial scales. The author’s study used arrest and gang data provided by the Philadelphia Police Department (PPD), which included information on 3,996 gang members belonging to 113 gangs. The study examined the validity of each of four indicators when they were used to define gang ecologies. The indicators were: where gang members live; gang arrests; crime incidents involving a gun; and PPD-defined gang set space boundaries. The author’s plan was guided by Messick’s unified perspective of construct validity and included two types of analyses: one employing a series of 60 regression models; and one that developed an algorithm creating gang set space polygons using either the locations where gang members live or the locations of gang-related crime. Results indicated that locations of gang members’ homes and of gang arrests can approximate the PPD reported gang set space locations equally well, and the algorithm was able to identify the general locations where gangs exist using only the home address of gang members or gang arrest data.