Date of Award

Spring 4-15-2021

Document Type


Degree Name

Master of Arts (MA)


Criminal Justice



First Advisor or Mentor

Frank Pezzella

Second Reader

Heath Grant



Statement of Problem: This research examines the difference in the predictive nature of conventional social disorganization measures on seven (7) different crime types between urban and rural counties in the United States during 2014. There is a well-developed body of criminological research examining the effect of neighborhood disadvantage on crime through the social disorganization framework. However, social disorganization has been tested almost exclusively in urban or metropolitan areas, suggesting the current framework explaining criminal incidents through conventional measures of neighborhood disadvantage may be limited by “urban bias”. The present study seeks to test the conventional measures of social disorganization on crime across different types of communities, thus identifying any potential variant effect in previously identified measures on crime across different communities. In doing so, this study will build on the social disorganization framework argue neighborhood disadvantage alone does not adequately explain the variants in crime, but rather how neighborhood disadvantage manifests in different communities with varying opportunities for criminal behavior.

Research Design and Methods: This study examines the variant predictive effect of conventional social disorganization measures on seven (7) different crime types between urban and rural counties in the United States during 2014. Crime data is collected from the 2014 version of the Federal Bureau of Investigation’s (FBI) Uniformed Crime Report (UCR) and accessed from the Inter-University Consortium on Political and Social Research (ICPSR Study #36399; U.S. Department of Justice, 2014). Data for social disorganization measures are obtained by the United States Census Bureau’s American Community Survey (ASC). Data in the ACS include a wide range of community characteristic estimates such as socioeconomic status, demographics, and educational attainment at various geographic units of analysis. In total, a sample of 2,318 counties was selected for this study. In this sample, 1,136 counties (49.01%) are classified as “urban” and 1,182 counties are classified as “rural” (32.2%).

Analysis: This study will run three sets of negative binomial regression models. This analytic technique was selected given the dependent variable of interest (i.e., crime incidents reported across different county types) is operationalized as count data. The first set of models will examine the variant effects of conventional social disorganization measures on different types of crimes in urban counties, while the second set of models will examine the variant effects in rural counties. According to Osgood and Chambers (2000), negative binomial regression, especially poisson-based models, are most appropriate when examining count data and particularly when involving low counts and smaller populations. Crime incidents in areas with small populations, as is the case in rural counties, tend to be less precise and analyses are skewed relative to higher populated, urban counties. The final set of regression models will include the natural log of the population as an offset variable with a fixed coefficient of 1. In doing so, the skewed distributions of crime counts are standardized relative to the size of the population of a given geography, and thus address the problem of error variance and allow for more precision within the analyses (Osgood and Chambers 2000).



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.