Abstract

The overall goal of this dissertation is to examine how the built and natural environment influences how potential criminals identify crime sites to offend within. Guided by the theoretical principles of crime site selection and crime pattern theory, this study specifically focuses on the crimes of street robbery and commercial burglary in three unique study locations—Atlanta, GA, Fayetteville, NC, and Rochester, NY. The data for this study were collected from multiple publicly available data repositories. Of these repositories, criminal incident data for the dependent variables were gathered from the National Policing Institute's Police Data Initiative. Data for the independent variables, which are representative of the built and natural environment, were collected from various open-source public and governmental agencies. To assess the influence of the built and natural environment on crime site selection, several techniques were employed. First, general spatial patterns were mapped using both kernel density estimation (KDE) and directional distribution analysis. Subsequently, temporal trends were identified by separating the data into several temporal units of analysis, including by meteorological season, weekday/weekend, and four-hour block increments. To assess multivariate relationships, two machine learning techniques were used: multivariate clustering and random forest classification. In alignment with prior literature, findings indicate that criminal incidents for both street robbery and commercial burglary cluster spatially and temporally. Of note, there are seasonal trends identified within the data, as well as trends relating to the time of day. Results from the multivariate clustering analysis reveal several unique spatial clusters of variables within each study location. The random forest classification and regression analysis rank ordered the importance of independent variables in their relationship to the criminal incidents in question. This ordering varied considerably depending on the temporal unit of analysis in question, which suggests the spatial predictors of street robbery and commercial burglary differ by season, weekday and weekend, and time of day. These results hold theoretical, methodological, and practical implications within the scope of environmental criminology.

Notes

If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu

Graduation Date

2023

Semester

Spring

Advisor

Moreto, William

Degree

Doctor of Philosophy (Ph.D.)

College

College of Community Innovation and Education

Department

Criminal Justice

Degree Program

Criminal Justice

Identifier

CFE0009506; DP0027510

URL

https://purls.library.ucf.edu/go/DP0027510

Language

English

Release Date

May 2023

Length of Campus-only Access

None

Access Status

Doctoral Dissertation (Open Access)

Share

COinS