Publication Type: Poster
Abstract: A large body of literature has been dedicated to understanding re-offending after release from prison – a phenomenon known as recidivism. Such research is critical both to understanding the determinants of crime and optimizing the correctional system. A meta-analysis was conducted using 49 published studies on actuarial predictors of both violent and non-violent recidivism in men and women. The current analysis resulted in 65 effect sizes comparing violent recidivists and non-recidivists, and 60 effect sizes between comparing recidivists with non-violent recidivists. Several variables (drug/alcohol use, age, and marriage) were predictive of recidivism in general. Gender differences were also observed. In men, increased violent criminal history was associated with increased violent recidivism. In women, a younger age at first offence and longer sentences were predictive of violent re-offending. Despite limited data on women, unique predictors of recidivism were observed in each gender..

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: One widely-acknowledged criticism of use of UCR rates to assess long-term trends in “crime” is that they are based on crimes reported to police, which can change over time in response to factors other than temporal shifts in offending. This problem is most apparent when UCR murder rates are compared with other violent crime rates, and UCR violent crime rates are compared with rates based on the NCVS. Two explanations for the upward and downward movement of UCR violent crime rates over time exist in the literature. One posits changes in police efficiency in recording crime (O’Brien 1996, 2003; also Jensen and Karpos 1993), while the other emphasizes reductions in violent crime lethality due to improvements in trauma care (Harris et al. 2002). Using annual data from various sources for the years 1960-2004, this study examines these two explanations for long-term trends in UCR homicide, rape, robbery, and aggravated assault rates, along with the alternative which claims that these rates reflect real changes in offending behavior. Preliminary results of time-series models show that all three factors play a role in generating long-term trends in UCR violent crime rates.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Sociologists have long been cognizant of the high cost of adolescent violence to individual and societal well-being. Violent youth rarely limit their delinquency to specifically violent actions. They are often involved in other kinds of delinquency, such as school-related conduct problems (including truancy), drug and alcohol use, and sexual promiscuity (Ellickson, Saner, & McGuigan, 1997; Sommers & Baskin, 1994). Using data from the National Longitudinal Study of Adolescent Health, I compare non-violent and violent adolescent girls in three areas of life: relationships with parents, number of hours spent viewing TV and videos, and self-esteem and depression. Analyses using an ANOVA showed that 1) girls who report a better relationship with their parents also report lower levels of violence, 2) girls whose parents report better relationships with them report lower levels of violence, 3) girls who watch less TV and fewer videos, report less smoking and/or drinking, and have fewer sex partners also report lower levels of violence, and 4) girls who report higher levels of self-esteem and lower levels of depression also report lower levels of violence. This study helps fill some gaps in the literature by examining more direct factors that may influence adolescent female violence as well as comparing the child reports to the parent reports.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: In recent years, it has been proposed that active police response to disorder has impacts on crime rates, especially in regards to violent crime (Kelling and Sousa, 2001; Sampson and Raudenbush, 1999; Sampson and Cohen, 1988). Some studies have shown that there is a relationship between crime and disorder; however, the direction of the relationship is still debatable (see Kelling and Sousa, 2001 and Sampson and Raudenbush, 1999 for opposite views). Additionally, the lack of evidence from longitudinal studies severely challenges the causal claim made by the Broken Windows thesis. The current study utilizes longitudinal methods to examine the relationship between police response to disorder and violent crime incidents across 16 years in the city of Seattle. The availability of longitudinal data will help to differentiate whether the relationship between disorder policing and violent crime is causal or spurious. Some criminological theories predict causal link between violent crime and disorder (i.e., Broken Windows thesis) and others suggest that this relationship is spurious (i.e., Social Disorganization by Sampson, 2002). In this study, I test the validity of some major criminological theories based on their predictions regarding the relationship between crime and disorder trends.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This study looks at environmental factors that explain the speed, temporary setbacks and degree of success in transitions from violence to non-violent behavior by movements that historically retain both a violent and peaceful political character.