Doctor of Philosophy / Department of Sociology, Anthropology, and Social Work / W. Richard Goe / Ryan E. Spohn / This study seeks to explain variation in incarceration rates across states. To account for such variation, the study combines approaches: the Rusche & Kirchheimer (1939) thesis, which proposes that incarceration rates rise with unemployment due to potential threat to social order from the unemployed, was merged with social stratification theories to develop a theoretical model explaining variations in state incarceration rates by social class and race. The last 30 years have seen a number of studies dedicated to investigating the validity of the Rusche and Kirchheimer (1939) thesis, but these studies have yielded inconsistent results. This study adheres to and advances Rusche and Kirchheimer’s thesis, exploring the relationship between unemployment rates and incarceration rates utilizing nationwide state-level data. I tested the influence of economic factors on prison rates across the nation interacting with race-ethnicity using time series hierarchical regression, and data indicates mixed support for the Rusche and Kirchheimer thesis. This study found that important predictors related to rising incarceration rates include citizen and governmental political ideology, violent and property crime rates, and percent of population that is African American. Habitual violation of laws, including drug crime, and poverty had small effects on the incarceration rate. Additionally, this study found that inequality, not unemployment, was the most salient predictor of incarceration rates; that is, the differential in employment pay rate factored more significantly than the designation of employed/unemployed. The study revealed that such a relationship between income inequality and punishment differentially impacts citizens in general and African Americans in particular. Since excessive use of prisons exacerbates inequality, understanding the link between economic conditions such as income inequality and punishment has notable policy implications.

Doctor of Philosophy / Department of Sociology, Anthropology, and Social Work / L. Susan Williams / Law enforcement is widely regarded as one of the most stressful and violent professions. Officer stress is associated with a number of negative behaviors and psychological outcomes including high rates of substance abuse, divorce, and violence. Concerns over officer violence include both work-related acts, like use of excessive force or unwarranted deadly force, and non-work related violence such as domestic abuse. Despite interest in the interplay between subcultural attitudes, organizational structure, and high stress events, most research on police violence fails to address a fundamental concern--that of gender. In fact, the majority of research addressing officer stress fails to mention gender or concentrates on gender as a simple control variable. In order to examine how gender, stress, and law enforcement structures predict violent behavior among police officers, this dissertation utilizes both existing data and direct officer interviews. Findings show that law enforcement remains largely a masculine enterprise, and that gender drives images, interaction, and organizational behavior, often at the expense of both men and women officers. Stress, burnout, and use of violence by officers are not simply a response to high stress events, but are embedded in the gender structure and process of policing.

Master of Arts / Department of Sociology, Anthropology, and Social Work / Michelle Bemiller / John Braithwaite developed the theory of reintegrative shaming in 1989. His hope was to develop a theory that better explained the complexities of crime such as the age curve, and high number of male offenders. Building from Braithwaite’s work, this thesis utilizes Reintegrative Shaming theory to explore how attitudes influence the reintegrative shaming process, and whether or not women, are more susceptible to shaming than men, as hypothesized by Braithwaite. It seeks to understand the role attitude plays in the reintegrative shaming process. This thesis hypothesizes that a reintegrative shaming punishment will positively impact an offender’s attitudes towards law, deterrence, law enforcement officials, and so on. Further, Braithwaite hypothesized that interdependent and communitarian individuals will be more susceptible to the impact of punishment (shaming); this relationship is also tested. Using data from the Reintegrative Shaming Experiments (RISE) in Australia, this thesis specifically studies the impact of sex and shaming punishment on attitudes towards the law and deterrence attitudes. In addition, information from RISE is used to test the relationships between interdependency and sex, and communitarianism and sex. The findings suggest mixed support for Reintegrative Shaming theory.

The purpose of this project was to empower the incarcerated women at the Jefferson Correctional Institution in Monticello, Florida, through the use of a faith-based program, 'Empowered to Endure Hardship.' The project consisted of sixteen (16) consecutive weeks of group participation, involving 75 women who were divided into two groups. Group A, the target group, consisted of 45 women who completed the questionnaires and participated in all of the group sessions and activities. Group B, the control group, consisted of 30 women who only completed the questionnaires. The sessions in which the target group participated included video and audio preaching tapes, live preaching, group interaction and discussions, prayer and a short devotional period at each session. All of the sermons contained one common thread: How to overcome or endure hardships in life. Practical examples were given for endurance and overcoming techniques were demonstrated. The overall hypothesis was as a result of Group A's participation in an organized structured group, the participants would receive fewer disciplinary reports, corrective counseling reports, and confinement visitations than those in Group B. While this goal was attained by Group A, there was not enough significant difference in Group B to merit any real attention. This does not mean the project was a failure. For in the ensuing weeks after the project was completed, the members of Group B continued to ask that another group be formed in which they could participate to receive the same empowerment that Group A had received. This model of ministry for the women at Jefferson Correctional Institution is ongoing and allows for additional components of ministry as future needs arise.

This research examines the content and process involved in developing criminal profiles. To assess the material typically included in profiles and the way that the conclusions are developed, a qualitative analysis of the content of profiles was undertaken. This included, among other things, the characteristics offered, their prevalence, and the most common characteristics by profiling method. A further quantitative study was undertaken employing Multidimensional Scaling (MDS) which revealed three general themes in the advice given. These included Crime Scene, Internal Offender and Offender Characteristic facets, each of which was comprised of features suggestive of that facet. For example, the Crime Scene facet typically described offender actions and their interpretation at the physical location of the criminal event, such as staging and methods of control.The study employed a sample of 49 profilers reflective of the four main schools of thought; Criminal Investigative Analysis (CIA), Investigative Psychology (IP), Diagnostic Evaluations (DE) and Behavioural Evidence Analysis (BEA). CIA profiles were greatest in number with a sample size of 19. There were 14 BEA profiles and 8 each of DE and IP profiles. Together, there were in excess of 3000 individual variables ranging from age, sex and race to employment, geography, mental illness and risk.Results indicate that the vast majority of profilers employ either research or personal belief in forming their opinions, and that physical evidence is used the least in justifying a profiler’s opinion. However, when considering the individual methods, BEA placed the greatest reliance on physical evidence and its interpretation according with the forensic science philosophy of the method. CIA, the method used by the FBI, produced the most characteristics of the sample at 312 across 19 profiles, followed by BEA at 166 characteristics across 14 profiles. IP and DE profiles, with a sample size of eight each, had 90 and 61 characteristics respectively.Beyond these findings, the implication of the research will be discussed and some future direction for this research and the field in general will be provided. These areas include education, training, the scientific method and professionalisation. This thesis will also canvass the issue of whether profiling is an art or a science with particular reference to the research results and the existing literature that has bothered to address this issue.

No description available.

This research examines the content and process involved in developing criminal profiles. To assess the material typically included in profiles and the way that the conclusions are developed, a qualitative analysis of the content of profiles was undertaken. This included, among other things, the characteristics offered, their prevalence, and the most common characteristics by profiling method. A further quantitative study was undertaken employing Multidimensional Scaling (MDS) which revealed three general themes in the advice given. These included Crime Scene, Internal Offender and Offender Characteristic facets, each of which was comprised of features suggestive of that facet. For example, the Crime Scene facet typically described offender actions and their interpretation at the physical location of the criminal event, such as staging and methods of control.The study employed a sample of 49 profilers reflective of the four main schools of thought; Criminal Investigative Analysis (CIA), Investigative Psychology (IP), Diagnostic Evaluations (DE) and Behavioural Evidence Analysis (BEA). CIA profiles were greatest in number with a sample size of 19. There were 14 BEA profiles and 8 each of DE and IP profiles. Together, there were in excess of 3000 individual variables ranging from age, sex and race to employment, geography, mental illness and risk.Results indicate that the vast majority of profilers employ either research or personal belief in forming their opinions, and that physical evidence is used the least in justifying a profiler’s opinion. However, when considering the individual methods, BEA placed the greatest reliance on physical evidence and its interpretation according with the forensic science philosophy of the method. CIA, the method used by the FBI, produced the most characteristics of the sample at 312 across 19 profiles, followed by BEA at 166 characteristics across 14 profiles. IP and DE profiles, with a sample size of eight each, had 90 and 61 characteristics respectively.Beyond these findings, the implication of the research will be discussed and some future direction for this research and the field in general will be provided. These areas include education, training, the scientific method and professionalisation. This thesis will also canvass the issue of whether profiling is an art or a science with particular reference to the research results and the existing literature that has bothered to address this issue.

Master of Arts / Department of Sociology, Anthropology, and Social Work / L. Susan Williams / Building upon macrostructural “social threat” (Blalock, 1967) research, the current study develops a theoretical model of judicial decision-making processes that focuses upon racial threats perceived within individual court contexts and the corresponding effects on individual sentencing outcomes. This model recognizes that in the absence of a true-measure of a defendant’s threat to the community (likelihood to re-offend) judicial decision makers often rely upon stereotypical generalizations regarding offender populations to render decisions. Although actors develop biases and stereotypes through interactions with society in general, the most relevant knowledge affecting sentencing decisions is perceptions gained through the course of work. Similar to the influential “Pygmalion in the Classroom” study, biases and stereotypes regarding the criminality of groups of criminal defendants are pervasive in contemporary society, undoubtedly influencing sentencing outcomes. Therefore, the most meaningful measurement of threat, as it pertains to sentencing, is the contextual composition of court caseloads. Using data from the Bureau of Justice Statistics-State Court Processing Statistics (SCPS) program, this study examines court-contextual or caseload level threats and the interaction between courtroom context and individual offense/offender characteristics and the corresponding impact on sentencing outcomes. Findings demonstrate that courts of high minority defendant volume apply more punitive sanctions to (increased sentence length and odds of incarceration) to all defendants within this context, while black defendants receive the greatest sanctions. These findings support assertions regarding the impact of threatening populations within courtroom contexts.

Master of Arts / Department of Sociology, Anthropology, and Social Work / L. Susan Williams / Studies on fear of crime demonstrate that fear of rape controls women’s lives by altering emotions and behavior, though how women construct rape discourse through social networks has not been examined. Further, studies tend to dismiss stranger rape because of its rarity compared to acquaintance rape, but this study argues that research must begin where women are. This study looks to women’s voices to articulate how they talk about fear of rape; specifically, it examines responses to a serial rapist at work in a college town. Framed by feminist methodology, this study establishes the influence of fear on women’s lives and the role of women’s social networks in disseminating information, constructing strategies, and changing behavior as it relates to a local serial rapist. The study utilizes a multi-method approach to quantify levels of fear in the community and to document qualitatively women’s responses to knowledge about the serial rapist. Two surveys, content analysis of local newspapers, and interviews support this research. In particular, group interviews conducted in two environments – campus face-to-face groups and online virtual groups – provide opportunities for young women to voice concerns and report behavioral changes related to the serial rapes. The research demonstrates that women are concerned about insufficient information from formal sources and want more accurate reporting. Women depend heavily on informal networks for information, but it is often incomplete and/or inaccurate and may actually intensify fear. As documented in earlier research, women focus on stranger rape to the neglect of the more common acquaintance rape and tend to strategize in individual terms rather than recognize structural issues. A major finding of this research is that young women actually perceive a change in their own identity as they try to manage fear of rape. However, women’s social networks and, in particular, the increasingly popular online networks, provide a forum from which to try out strategies, build collective discourse, and, in turn, develop greater group consciousness among young women. From the experiences of women in this study, several policy implications are offered for managing fear, including education about the more likely threat of acquaintance rape.

Doctor of Philosophy / Department of Sociology, Anthropology, and Social Work / W. Richard Goe / Criminologists have traditionally studied criminal offending and violent victimization separately. Extant studies, however, demonstrates that criminals and victims overlap to some degree, hinting that a common underlying trait explains both criminal offending and violent victimization. This study tests whether Gottfredson and Hirschi’s self-control theory explains the overlap in criminal offending and violent victimization exposure among gang members. Using cross-sectional survey data from the Evaluation of the Gang Resistance Education and Training (GREAT) Program in the United States, 1995-1999, results from the regression models show that low self-control is to some degree correlated with criminal offending and violent victimization. Gang members were more likely than non-gang members to participate in some forms of criminal activities, but they were not more likely to be victimized. When variables stemming from social learning and social bonding are included in the regression models, results show that associating with delinquent peers had the strongest effect in predicting criminal offending, contradicting Gottfredson and Hirschi’s claim that self-control is the only cause of criminal behavior. In concert with previous studies that have found a link between low self-control and violent victimization, results show that youths with low self-control were somewhat more likely than youths with higher self-control to report being victimized. The results of the study, as well as venues for future research, are discussed.