Filter Dissertations and Theses By:
Perceptions of Community Corrections: Understanding how Women's Needs are met in an Evidence-Based/Gender-Responsive Halfway House
Year of Dissertation:
This dissertation presents a qualitative study on how women perceive and experience services at an evidence-based, gender-responsive halfway house. The primary focus was to understand how the halfway house helps women address their needs as they prepare to reenter the community. The secondary focus was to understand how the halfway house implements evidence-based principles and gender-responsive strategies. This study analyzed in-depth qualitative interviews with 33 women. Data from these interviews were triangulated with observations of treatment groups and daily interactions, review of program documents, review of participant case files, and informal conversations with staff. Findings suggest that many positive and negative features of the halfway house - including social context, relationships with staff, and program policies - contribute to women's ability to address their needs and prepare for reentry. Findings also draw attention to the influence of external factors including outside resources, social networks, housing availability, the stigma of a criminal record, systemic policies, geographic boundaries, and program length of stay. The interconnections between ecological systems also influence the transitional process and were highlighted in this study. Recommendations for improving community correctional services for women were discussed.
Extremist Networks and Lethality: A mapping of violent white supremacist group networks and an investigation of the relationship between network location and ideologically motivated murder
Year of Dissertation:
Empirical evidence indicates that domestic extremists in the United States pose a greater risk to the American public than international terrorism (Carlson, 1995; Hewitt, 2003; Blewas, Griggs, and Potok, 2005; LaFree, Dugan, Fogg & Scott, 2006). This dissertation attempts to further our understanding of domestic extremists by employing Social Network Analysis (SNA) methodology to investigate the network of white supremacist groups associated with - via formal members - extreme ideologically motivated violence (homicides). SNA focuses on how actors (i.e. people, organizations) are linked in patterns of interaction and the meaning of those connections. The general hypothesis of SNA is that entities, like people or groups, are interdependent, and therefore more likely to network with those who share common interests, goals, belief systems, etc. Ultimately, choices are influenced by the company one keeps (Wassserman and Faust, 2006). The data for this study comes from the Extremist Crime Database (ECDB), created by Joshua Freilich of John Jay College and Steven Chermak of Michigan State University. The first objective of the study is to analyze and measure overall network structure (e.g. density, cohesion) as well as actor level characteristics (e.g. centrality, constraint) in an effort to ascertain which groups are most popular and/or important to the flow of information within the network. A second objective of this study is to determine, via regression analysis, whether certain actor level characteristics are significantly related to an increased threat of ideologically motivated homicide. If so, then a white supremacist group's role or location within the network may serve as a predictor of lethality. While SNA has been used to study international extremist networks (e.g. global jihadists), this research is novel in its approach to the study of domestic extremists and the threat they pose.
Inmate-Perpetrated Harassment: Exploring the Gender-Specific Experience of Female Correction Officers
Year of Dissertation:
Perhaps no other work environment is as mysterious to non-participants as is the institute of corrections in America. Other than the officers and inmates of a correctional institution, few people understand, or even care to venture into, the dynamics of a prison or jail. Because of this disregard, many important sociological studies have failed to explore the correctional institution as a legitimate workplace. In particular, sexual harassment studies have all but neglected the correctional workplace. Nevertheless, as in any work environment, sexual harassment is a significant issue in jails and prisons, even more commonplace and more often ignored than that harassment in other workplaces. Furthermore, a significant portion of harassment against female correction officers is perpetrated, not by their coworkers, but by the inmates for whom these officers are responsible. This exploratory research effort utilizes a questionnaire to examine the nature and consequences of inmate-perpetrated sexual harassment against female correction officers. The respondents to the study's questionnaire indicate the existence of a significant amount of sexual harassment perpetrated against female correction officers by male inmates. Yet, the majority of the sample group were reluctant to identify either their experiences as sexual harassment or themselves as victims of such harassment. From the combined analysis of the quantitative and qualitative responses of 21 female officers of the Lafayette Parish Correctional Center in Lafayette, LA, the study proposes the existence of a labeling disjunction among female correction officers with respect to their perceptions of the sexual misconduct of male inmates. This study concludes with recommendations for both future research and policy developments that may enhance general understanding and management of inmate-perpetrated sexual harassment against female correction officers.
Post-treatment drug use, recidivism, analogous behaviors, and perceptions of fairness: Examining whether parolees with low self-control will benefit from the Collaborative Behavioral Management intervention
Year of Dissertation:
Dr. Jeremy Porter
This dissertation tested Gottfredson and Hirschi's (1990) low self-control theory and its relationship with post-treatment outcomes by conducting a secondary-data analysis of a randomized controlled trial on parolees (n=569) called the Step'n Out study (2005). The Step'n Out study (2005) compared the results of a control group (standard parole) with an experimental treatment for parolees called the Collaborative Behavioral Management (CBM) intervention which was designed to improve substance-use treatment outcomes, reduce drug use, and reduce recidivism for parolees participating in the study. Low self-control theory states that individuals with character traits that are impulsive, risk-seeking, self-centered, and display volatile temper have a high likelihood of engaging in criminal and analogous (i.e. risky sexual practices) behaviors. Gottfredson and Hirschi's (1990) theory makes the assumption that these traits are the result of parental socialization practices, are not able to be changed after the age of 8 or 10, and are stable across time. In order to measure low self-control for the present study, an exploratory factor analysis was conducted on 20 self-report items collected at intake from the parolees in the study and a unidimensional measure of low self-control was constructed. Based on low self-control theory, this study hypothesizes that parolees who self-report engaging in substance use, recidivism, and analogous behaviors after the end of the treatment intervention at the 3 and 9 month follow-up periods will have low self-control traits (measured at intake). Also based on the theory, this study hypothesizes that the treatment condition (control group vs. CBM group) will not moderate the relationship between low self-control traits and post-treatment outcomes even when controlling for demographic, risk-factors, peer-associations, and treatment dosage. The exploratory results from this study were reported using univariate, bivariate, and multivariate statistics. Also a confirmatory factor analysis was conducted to measure the direct and indirect effects of low self-control, peer-associations, and perceptions of fairness on post-treatment outcomes. The results from this dissertation study largely indicate that parolees across the self-control spectrum (low to high levels of self-control) are engaging in post-treatment outcomes (substance use, recidivism, and analogous behaviors) at the 3 and 9 month follow-up periods even when controlling for age, gender, race, age at first arrest, education status, dosage levels, and treatment condition. Therefore, based on the findings from this study, low self-control theory does not allow researchers to understand the causal mechanisms by which post-treatment outcomes occur for parolees. More theoretical refinement of the theory or alternative theories are needed in order to explain the post-treatment outcomes of parolees participating in the Step'n Out study. However, a particularly interesting finding that also has strong public policy implications indicates that parolees that self-reported physically or verbally threatening someone at both the 3 and 9 month follow-up periods had statistically significant lower mean levels of self-control compared to parolees who did not physically or verbally threaten someone.
Religious Establishments, Public Housing, and Liquor Stores: Their Prediction of Juvenile System Behavior
Year of Dissertation:
The following dissertation examines the role of ecological structures in juvenile justice systems, specifically during risk assessment, prosecution, and sanctioning. This analysis of system behavior considers religious establishments, public housing, and liquor stores as the ecological indicators and views them as stigmatizing. Quantitatively, the following examination sought to (a) determine associations between social ecology and risk assessments, prosecutions, and residential sanctioning, and to (b) determine if juvenile probation officers and judges are more stringent and judgmental toward delinquents from neighborhoods that have greater concentrations of religious establishments, public housing, and liquor stores. All adjudicated juvenile delinquents whose cases have been decided by the Florida Department of Juvenile Justice in calendar years 2006, 2007 and 2008 were included in the analysis. Secondary data from the Florida Department of Juvenile Justice, the Division of Alcoholic Beverages and Tobacco Bureau of Licenses, the United States Department of Housing and Urban Development, American Church List and the United States Census were used to address the research questions. The databases were used to support the researcher's overall tenet that certain areas are perceived as disorganized, which leads to stricter expressions of risk assessment, prosecutions, and residential sanctions. It is hypothesized that, (1) risk assessment levels are higher in areas with more religious establishments, public housing, and liquor stores; (2) zip codes with more prosecutions will consequently be those with more of the stigmatized ecological structures; and (3) an increase in religious establishments, public housing, and liquor stores will generate an increase in residential sanctions. It was expected that the relationship between the independent and dependent variables would be significant, over and beyond demographic and legal factors. In the analysis, area demographics of population density; and juvenile demographics of age, race, ethnicity, and gender, along with current and prior legal history, were controlled for to determine the predictive value of the independent variables of religious establishments, public housing, and liquor stores on the dependent variables of risk assessment, prosecutions, and residential sanctions. Prior to statistical analysis, the data was merged and aggregated by zip code to reflect area composition, resulting in a dataset of 298 zip codes and 21 variables. To examine these relationships, analyses were done on a bi-variate and multi-variate level. Multi-variate analysis was performed using hierarchical regression. Three models were designed, considering demographics, and then adding legal variables, followed by ecological structures, to make the complete model.
Reduction of Observable Robbery and Larceny-Theft in the Twelve Largest Cities in the United States from 1980 to 2009
Year of Dissertation:
The reduction in crime rates that occurred in large cities across the United States (US) over the course of the past two decades has been the subject of much speculation and research. However, there have been no definitive empirical studies that conclusively determine the causes for this phenomenon. The goal of this study is to identify the impact of certain factors to the reduction of crime in large US cities that occurred over the past two decades by examining data over a thirty-year period (1980-2009). The identification of contributing factors may allow government officials, both on a local and national level, to focus their efforts on the implementation of policies that, based on empirical study, are likely to reduce crime. This study focuses on Observable Crime, which is operationalized as robberies and larcenies reported in the Uniform Crime Report (UCR) Part II Offenses that were likely to be visible to the police. Those crimes likely to be visible police are determined to be all robberies that were not committed in residences and larcenies that were committed in public areas excluding stores. Law enforcement strategies that were examined in this study include Quality of Life (QOL) Enforcement and Police Presence, which is operationalized as arrests for drug offenses as reported in UCR Part II Arrests and Police Officers per 100,000 residents as reported in the UCR, respectively. The findings of this research supports the hypothesis that Quality of Life Enforcement significant in reducing crime in the twelve largest US cities from 1980-2009.
LONG-TERM INCARCERATION AND PUBLIC SAFETY: PREDICTING THE RECIDIVISM RISK OF LONG-TERM PRISONERS
Year of Dissertation:
Since the 1970s, the number of people incarcerated in the United States has grown exponentially. The United States has now reached a historical moment as it incarcerates more of its citizens that it ever has before. Moreover, the rate at which it does surpasses all other nations. Increased length of sentences and time served have contributed substantially to America's prison growth, calling into focus the need for research that examines the impact long prison sentences have on an individual's likelihood of recidivism. To date, little is known about the relationship between long prison sentences and public safety outcomes. Unfortunately, in recent years, long-term incarceration has received minimal attention from the academic world. Despite the gap in research, questions surrounding long-term incarceration are as critical now as ever. Today, people are spending more time in America's prisons than ever before. Individuals, who at one time would have been released, are collecting in our nation's prison system, creating a larger and longer-term prison population. Today's economic reality forces criminal justice administrators all over the country to consider ways to cut budgets without compromising public safety. This study outlines the importance of considering the impact length of stay has on recidivism when making sentencing and release decisions. Using data from New York, this study examines the impact that each additional month served has on the likelihood that a person will be re-arrested or re-incarcerated within two years of release. Put another way, this study seeks to identify the point at which individuals pose no heighten risk to public safety. Results indicate that this point occurs years before people serving long prison terms are released. The relationship between long prison sentences and recidivism are explored further as factors associated with positive release decisions and criminal justice system involvement are controlled for in the models, helping to identify factors that are significant in predicting recidivism outcomes. Implications for sentencing practice and parole release decisions are discussed. Additionally, future research opportunities are identified.
Sexual Victimization, Disclosure, and Accountability: Organizational Responses of the Boy Scouts of America to Child Sexual Abuse
Year of Dissertation:
Child sexual abuse has always been a highly publicized and controversial topic due to the vulnerability of children. However, there have been several recent child sexual abuse (CSA) scandals garnering even more attention and concern. With the exposure of the Catholic Church abuse scandal in 2002, public attention became focused on the Archdiocese’s poor and prolonged response. However, this crisis was not the sole case of widespread sexual violence and institutional failure. In October of 2012, Boy Scouts of America files from 1959 to 1991 containing reports of child sexual abuse were released. These reports feature detailed information about the abuse incident, the offender, and the Boy Scouts’ response to the allegations of abuse. While there appears to be some research on organizational child sexual abuse, these studies focus primarily on prevention of abuse, and not on organizational responses to abuse. This dissertation addressed the factors of the abuse and the organization that influenced the Boy Scouts of America's response. Data for this study comes from roughly 2379 publically released files of sexual abuse reports from the Boy Scouts of America. These reports, dating from 1959 until 1991, feature detailed information about the abuse incident, the offender, and the Boy Scouts’ response to the allegations of abuse. Multivariate analysis, survival analysis and intervention analysis was conducted on the 2379 cases as well as qualitative content analysis. In addition, in-depth case studies were conducted of a few select cases to gain a deeper understanding of how the BSA responded to abuse.
The Lost Children of New York City: Population estimate, network attributes and the role of social capital in the commercial sexual exploitation of children in New York City
Year of Dissertation:
Abstract The Lost Children of New York City: Population estimate, network attributes and the role of social capital in the commercial sexual exploitation of children in New York City by Meredith Dank Advisor: Richard Curtis, Ph.D. The number of youth that are commercially sexually exploited in the United States is unknown. Additionally, the characteristics that make up the commercial sexual exploitation of children (CSEC) population, as well as the CSEC market, are widely disputed among researchers, child advocates, and professionals in the field. Past research has relied on police arrest records and used interviews with a limited, purposive sample to generate a prevalence rate and describe the characteristics and needs of the CSEC population. As a result, the findings from these studies do not provide a full picture of the issue. In order to gain a better understanding of the CSEC population in New York City, this dissertation employed the Respondent-Driven Sampling (RDS) method to recruit and interview a sample of 249 eligible teenagers. RDS is a sampling strategy that is dependent on a number of assumptions concerning the social networks that are sampled, but it has been successfully used in a wide variety of research studies to recruit hard-to-reach populations. Based on the RDS data collected, a prevalence rate was generated, in addition to an explanation of the attributes of the CSEC networks and its members. Social capital theories were used to explain how youth become involved in the CSEC market and why it is difficult for them to leave it. The findings from this study suggest that youth who possess social capital before entering the market, and are able to maintain it, are more likely to seek help from individuals who possess higher levels of normative social capital, which can ultimately lead to their leaving the CSEC lifestyle. Those youth who do not have pre-existing normative social capital are more constrained by their choices and the alternatives available to them. They feel more compelled to remain in the market in order to survive. Based on the richness of the data collected and resultant findings, this research will add value to the extant body of knowledge, inform policy, and bring much needed attention to this issue.
A STUDY COMPARING THE EYEWITNESS ACCURACY OF POLICE OFFICERS AND CITIZENS
Year of Dissertation:
Police officers are not only responsible for administering lineups and interviews where citizens' eyewitness memories are tested, they are also called upon to make arrests, write reports, prepare warrants and testify in court based on their own memories. Do police make better eyewitnesses than citizens? This study hopes to partially answer that question while contributing to the body of eyewitness research on weapon focus effect (WFE) and the new area of inquiry on trying to understand witness accuracy for multiple perpetrator crimes. This study investigates the effects of WFE and the presence of multiple perpetrators on eyewitness memory in two separate experiments. Two groups, police officers and citizens, were tested and compared in each experiment. One of the reasons that have been theorized to explain the presence of WFE is that weapons might hold a certain amount of contextual relevance or novelty that draws the attention of the observer when the weapon is present in a scene. Since police officers are commonly exposed to weapons and receive training on de-escalation of multi-person conflicts, the current study attempted to determine whether citizens and police would perform in similar or dissimilar ways to situations which might inhibit the observation and encoding of crime scene elements into memory. Two experiments were conducted to measure the arousal level of participants and assess the accuracy of police and citizen identification decisions in situations that potentially divert attention from the perpetrator in a simulated crime. Experiment 1 examined the effects of the presence of a weapon and the "weapon focus effect." The results showed that police officers tested lower on certain factors associated with stress and arousal than citizens but both police and citizens made more errors when a weapon was inferred or present. This is the first time that the inferred weapon condition has been experimentally explored. In addition, police made fewer filler identifications when the lineup target was present than when absent. Experiment 2 tested whether the presence of two culprits instead of one culprit affected identification rates. Both police and citizens experienced an increase in two factors that are associated with stress and arousal. Additionally, both police and citizens' identification accuracy was lower in the presence of two culprits; no accuracy differences were found between the police and citizen group. The results will add to the body of literature in eyewitness identification and contribute to the understanding of how stress or anxiety may or may not affect identification accuracy. In addition, it is hoped that elements of the study will be useable in police training contexts to help understand and improve the way that eyewitness evidence is processed and used by law enforcement agencies.