Community structure and criminal victimization

This research has attempted to better understand property crime victimization by studying the important role of community structures, while controlling for the following demographic variables: age, gender, race, and income. Three different types of analyses were used: (1) bivariate analysis; (2) multivariate analysis, and (3) path analysis. Bivariate analysis was used in order to gain a better understanding of the following zero-order relationships: (1) the relationship between the structural characteristics of communities and property crime victimization; (2) the relationship between the structural variables and the mediating variables--guardianship and neighborhood cohesion; (3) the relationship between guardianship and property crime, and (4) the relationship between neighborhood cohesion and property crime. Most of these relationships were found to be in the expected direction.

The multivariate analysis was conducted by running three separate regression models. Model 1 included only the structural variables of community size, racial heterogeneity, and residential mobility. In Model 2, demographic variables were added in order to see how this addition would impact the effects of the structural variables on property crime victimization. Model 3 included both the structural and the demographic variables, along with guardianship and neighborhood cohesion. This additional regression model was used in an attempt to discover the effects of guardianship and neighborhood cohesion on property crime victimization.

Path analysis was used in order to find out the direct and indirect effects of the structural and demographic variables on property crime victimization.

Many of the findings of this research were not consistent with past research. There appears to be other important factors which were not included. For example, guardianship and neighborhood cohesion did not mediate the effects of the structural and demographic variables. The thesis concludes with alternative explanations for these and other inconsistent findings.