The 35,882 former offenders released by Virginia's Department of Corrections during the period of July 1, 1998, through June 30, 2002, are profiled using data provided by Virginia's Department of Corrections and Department of Correctional Education as well as the Virginia Employment Commission. Demographic characteristics of recidivating and non-recidivating former offenders released during this period, with sub-groupings to include gender, race, age, employment status, earnings, employment stability, and educational completion, as defined in the operational definitions of the study, are detailed, and the researcher's observations are noted. / Ph. D.

One aim of residential drug treatment programs for drug offenders released from prison is to reduce recidivism rates for these offenders. However, research on this topic has shown mixed results on the effectiveness of such programs, and all of the factors that influence whether an individual will recidivate are still unclear. This study explored the relationship between criminal classification levels of drug offenders and the length of time spent in residential treatment and the effects that this relationship had on drug offenders' odds of recidivism. Data from case history records of drug-involved offenders in Florida from 1991 to 1997 were analyzed to determine whether the length of time in treatment reduced these offenders' odds of recidivism, whether offenders convicted of a felony 1 level offense were more or less likely to recidivate compared to those convicted of a felony 2 or 3 level offense, and whether there was an interaction effect of length of time in treatment and criminal classification level on offenders' odds of recidivism. Binary logistic regression analysis shows that the more time a drug offender spends in residential treatment, the higher the offender's odds of recidivism; drug offenders convicted of a felony 1 level offense were less likely to recidivate than offenders convicted of a felony 2 or 3 level offense; and that there was no interaction effect of length of time spent in treatment and criminal classification level on drug offenders' odds of recidivism. / Master of Science

A conviction is a ramification that extends beyond the correctional facility. An extensive amount of research has explored the barriers the justice-impacted experience once they are released from prison. One of the most immediate and impactful barriers is their ability to secure employment, due to it being quintessential in reducing their likelihood to recidivate and engage in illegal activity post-release. While much research has specifically focused on former prisoners’ ability to secure employment post-release, very limited researched exists that examines how convicted offense impacts employment. Utilizing the Serious and Violent Offender Reentry Initiative (SVORI) multi-site impact evaluation as its secondary dataset, the present study aimed to explore the impact violent offenses (non-sex), sex offenses, white-collar offenses, property offenses, drug offenses, and confounding sociodemographic factors have on securing employment three months post-incarceration. This study hypothesized there is a significant association between employment status and convicted offenses/convicted offense types among the justice-impacted, even when accounting for confounding sociodemographic factors. Through binary logistic regression analysis and multiple imputations, the results from the study reveal statistical significance for the relationship between convicted offenses (assault, car theft, drug dealing, drug possession, and forgery), convicted offense types (drug and white-collar), and confounding sociodemographic factors (age, education, and race) with employment 3 months post-incarceration. It is hoped these results reveal how stifling deficits are to securing employment for the justice-impacted, and the need for further policy and programming application to decrease these challenges.

This thesis explores an approach to predicting recidivism by leveraging crowdsourcing, contrasting traditional judicial discretion and algorithmic models. Instead of relying on judges or algorithms, participants predicted the likelihood of re-offending using the COMPAS dataset, which includes demographic and criminal record information. The study analyzed both quantitative and qualitative data to assess biases in human versus algorithmic predictions. Findings reveal that homogeneous crowds reflect the biases of their composition, leading to more pronounced gender and racial biases. In contrast, heterogeneous crowds, with equal and random distributions, present a more balanced view, though underlying biases still emerge. Both gender and racial biases influence how re-offending risk is perceived, significantly impacting risk evaluations. Specifically, crowds rated African American offenders as less likely to re-offend compared to COMPAS, which assigned them higher risk scores, while Caucasian and Hispanic offenders were perceived as more likely to re-offend by crowds. Gender differences also emerged, with males rated as less likely to re-offend and females as more likely. This study highlights crowdsourcing's potential to mitigate biases and provides insights into balancing consistency and fairness in risk assessments. / Master of Science / Within the criminal justice system, predicting whether someone will re-offend has typically depended on the judgment of judges and computerized systems. This thesis investigates another avenue for predicting re-offending by using crowdsourcing, which gathers input from a group of people. In this study, participants were asked to predict the likelihood of re-offending for several offenders using demographic and criminal record information from the publicly available COMPAS dataset. Participants provided scores, and some also explained their reasoning. Bias, defined as a systematic unfairness that leads to prejudiced outcomes, was a key focus. To understand bias, the study created different groups within the participant crowd based on age, gender, and race, and compared their predictions with COMPAS scores. The analysis revealed important insights into the biases present in both human and algorithmic predictions. A homogeneous crowd, is associated with minimal differences in ratings across genders and races, suggesting a consistent but potentially biased perspective. While a diverse crowd, leads to varied ratings without a clear trend, reflecting a broader range of viewpoints but also increased variability. This suggests that while a diverse crowd may help reduce bias, it can also result in less predictable assessments.

A vigorous debate persists in the literature about the efficacy of clinical judgment and actuarial models of risk assessment. This study was designed to augment those commonly used methods by integrating a variety of factors that produce risk and protective effects among 101 youthful offenders. Adolescents and young adults in a maximum-security facility were interviewed with the Psychopathy Checklist: Youth Version (PCL:YV), and completed self-reports of psychopathy, impulsivity, and perceived parental care and protection. This selection of empirically-supported predictors was enhanced by criminal history and family information obtained through extensive file review. Markedly different prediction models emerged based on age. ADHD and PCL Factor 2 predicted adolescents' institutional maladjustment. In contrast, young adults' institutional behavior was influenced by impulsivity, family substance abuse, and gang membership. Treatment progress also differed depending on age; the absence of certain risk factors predicted success for adolescents, while academic achievement and intelligence facilitated young adults' advancement. Importantly, support was demonstrated for the moderating effects of protective factors on violence. Finally, the predictive validity of newly-developed psychopathy self-reports was examined in relation to the PCL:YV. Both the SALE PS-24 and the APSD were modestly effective at differentiating between high and low levels of psychopathy.

Juvenile homicide offenders pose a significant risk to society considering the severity of the crime, yet this population of delinquents receives little attention in terms of recidivism research. This study examined the recidivism outcomes of a group of 256 determinately sentenced homicide offenders in Texas. Each of these juveniles had be confined to and subsequently release from the Texas Youth Commission, and were followed for three years following release. The aim of this research study was to distinguish recidivists from non-recidivists across demographic, delinquent history, and risk factor measures. These same measures were then used to predict recidivism among the entire sample. Overall, the measures of race, gender, and previous delinquent adjudications emerged significant. First, Black youth were significantly more likely to recidivate than both Hispanic and White youth. Additionally, recidivism among males was significantly greater than non-recidivism among males. In contrast, there were fewer females in recidivist group than the non-recidivist group. Finally, youth with a greater number of previous delinquent adjudications were significantly more likely to recidivate, and the risk of recidivism increased as the number of previous adjudications increased. These findings are consistent with previous literature and indicate that these factors are prevalent in analyzing the recidivist behavior of determinately sentenced juvenile homicide offenders.

Research on the factors underlying sex offender recidivism has not considered the importance of the reintegration process through which the offender rejoins the community after prison. This thesis reports findings from 3 empirical studies designed to explore whether poor release planning might contribute to sex offender recidivism. In Study 1, a coding protocol was developed to measure the comprehensiveness of release planning for child molesters, which included items relating to accommodation, employment, pro-social support, community-based treatment, and Good Lives Model (T. Ward & C.A. Stewart, 2003) secondary goods. The protocol was retrospectively applied to groups of recidivist and nonrecidivist graduates of a prison-based treatment programme, who were matched on static risk level and time since release. As predicted, overall release planning was significantly poorer for recidivists compared to nonrecidivists. Study 2 was a validation and extension of Study 1. The original coding protocol, and some revised items, were applied to matched groups of recidivists and nonrecidivists from a different treatment programme. Consistent with Study 1 findings, overall release planning was significantly poorer for recidivists. Data from Studies 1 and 2 were pooled (total N = 141) and Cox regressions showed that accommodation, employment, and social support planning combined to best predict recidivism, with predictive accuracy comparable to that obtained using static risk models. Study 3 investigated whether release planning was associated with actual reintegration experiences, and additionally explored released child molesters’ good lives plans. Release plans were rated for 16 child molesters, who were interviewed post-release about their reintegration experiences and good lives plans. As predicted, significant positive correlations were found between release planning and reintegration experiences 1 and 3 months following prison release, and results suggested that effective reintegration might help facilitate living a good life. Overall, results from the 3 studies suggest that poor release planning and subsequent reintegration experiences contribute to sex offender recidivism. Implications for researchers, clinicians, policy makers, and community members are discussed.

Juvenile sex offenders represent a serious and violent group of delinquents. Despite the severity of their crimes, the literature focusing on risk factors that influence recidivism and the types of re-arrest after incarceration is lacking. This research study examined 499 determinately sentenced juvenile sex offenders that were released from the Texas Juvenile Justice Department. This sample was then followed for three years upon their release. This analysis revealed that 51.5 percent were re-arrested for any offense while 45.91 percent were re-arrested for a felony offense. This study identified a number of risk factors relative to JSO recidivism. These factors include having a history of emotional abuse, race being African American, being gang affiliated, having a larger number of previous adjudications, and having higher counts of institutional misconduct infractions. Those JSOs older at intake and release, and those who were incarcerated for longer periods of time were less likely to re-offend upon release. Lastly, this study ends with suggestions for future research as well as policy implications geared toward juvenile sex offenders.

This dissertation examines recidivism for domestic violence offenders under both traditional probation and specialized probation conditions. This research compares recidivism at a traditional probation locality (Riverside Criminal Justice Services) and a specialized domestic violence unit (Chesterfield Community Corrections Services) in order to determine the nature of recidivism at both localities. In addition, the research aims to identify individual risk factors which are statistically associated with recidivism. The research is based on secondary data and draws on offender criminal records generated by the Virginia State Police, as well the extraction of information from probation files at both localities. The multivariate models indicate that a higher degree of recidivism occurs at the specialized domestic violence unit, and that the experimental locality, along with a younger age, are consistently statistically significant predictors for recidivism. Less consistent but still statistically significant risk factors included positive drug tests, higher supervision levels, and unsuccessful probation completion. The implications of these findings, as well as policy recommendations and directions for future research are fully discussed.

Introduction: Renewed efforts for controlling criminality in Sweden are demanded by the Swedish Government. Restorative justice conferences (RJC), in Sweden known as mediation, can be described as a meeting mediated by a trained facilitator in which the offender and the victim of a crime meets, sometimes together with their friends and family. RJC aims to promote the offender to take responsibility for his or her actions and repair the damage the victim has suffered. Research on RJC shows promising results in regard of reduced recidivism and increased victim benefits. No Swedish publication studying public opinion on RJC has been found. The theoretical framework of reintegrative shaming stated that a community that offer reintegration is essential for successful crime prevention. Aim: To measure the level of RJC support held by a Swedish community sample in regard of different offender and offence characteristics. Method: Participants were recruited at a Swedish authority located in nine different cities (n = 192) and surveyed using an online questionnaire. The questionnaire was based on two scales measuring benefits of restorative justice and support for participation in restorative justice derived from a previous study. Results: The respondents had an overall positive attitude toward RJC. No respondent demographic was associated with the total level of restorative justice support although differences were found when using age and gender as independent variables. Respondents were more positive of RJC for juvenile offenders and non-violent crime. Discussion: A discussion regarding the public support for RJC in relation to offender recidivism is provided established on reintegrative shaming and previous empirical evidence. Concluding arguments for RJC to be extended in a wider context in the Swedish penalty system are presented. / <p>2016-06-01</p>