Assessing Violence Risk in Stalking Cases: A Regression Tree Approach

Advances in the field of risk assessment have highlighted the importance of developing and validating models for problematic or unique subgroups of individuals. Stalking offenders represent one such subgroup, where fears of and potential for violence are well-known and have important implications for safety management. The present study applies a Classification and Regression Tree (CART) approach to a sample of stalking offenders in order to help further the process of identifying and understanding risk assessment strategies. Data from 204 stalking offenders referred for psychiatric evaluation to a publicly-funded clinic were used to develop and assess putative risk factors. A series of nested models were used to generate tree algorithms predicting violence in this sample of offenders. Both simplified and more extensive models generated high levels of predictive accuracy that were roughly comparable to logistic regression models but much more straightforward to apply in clinical practice. Jack-knifed cross-validation analyses demonstrated considerable shrinkage in the CART, although the models were still comparable to many other actuarial risk assessment instruments. Logistic regression models were much more resilient to cross-validation, with relatively modest loss in predictive power.     

The Role of Victim-Related Factors in Victim Restitution: A Multi-Method Analysis of Restitution in Pennsylvania

Mandatory statutes do not always produce change, but a 1995 Pennsylvania statutory change making restitution mandatory dramatically increased the proportion of cases in which restitution was imposed. There are three possible reasons for this generally successful implementation: (a) judges agreed with the victim-centered goals of the statute, (b) there were mechanisms in place to implement the goals of assisting victims, and (c) there was a context supportive of victims that made it easier to follow the law. Two studies investigated these possible explanations. First, a statewide survey of trial court judges suggested that they agreed with the statute's goals of compensating victims. Second, hierarchical logistic models of 55,119 statewide restitution-eligible decisions indicated that a victim-related contextual factor, the nature and location of the victim/witness assistance office, was significantly related to the imposition of restitution, although a more general contextual factor relating to funding for victim programs had only small effects.     

Budget Rhetoric in Presidential Campaigns from 1952 to 2000

We offer a theory to identify the determinants of presidential campaign rhetoric related to the federal budget. The theory builds on the literature dealing with issue ownership, candidate strategy, retrospective voting, and voter preferences to generate eight hypotheses about the use of budget rhetoric. To test these hypotheses, over 800 campaign speeches from the major party presidential nominees from 1952 to 2000 are content analyzed. The content analysis generates measures of both the volume and tone of budget rhetoric. Volume is driven primarily by the objective balance of the budget and subjective importance given to it by voters and a conditional effect involving budget balance, incumbency, and partisanship. Tone is more complex, with positive rhetoric determined mostly by the budget balance and partisanship and overstated rhetoric shaped solely by the salience of the budget to the electorate. The article concludes with suggestions for future research.     

Recidivism in Stalking and Obsessional Harassment

Despite the rapidly growth of mental health attention focused on the phenomenon of stalking, no empirical research to date has attempted to assess the frequency of repeat offending or attempted to identify predictors of recidivism. A total of 148 stalking and harassment offenders who were court-ordered to undergo a mental health evaluation were followed for a period of 2.5–13 years in order to assess the frequency of repeat offenses and the variables that differentiated high versus low risk offenders. Recidivism data were obtained from a variety of sources, including criminal justice records, mental health records, and reports from probation officers and victims. A number of potential predictor variables were selected on the basis of the existing recidivism literature in other criminal justice populations. Frequency analysis were used to identify variables that significantly differentiated offenders who did and did not reoffend while survival analysis was used to analyze the impact of these covariates on time to reoffense. A total of 49% of the offenders reoffended during the follow-up period, 80% of whom reoffended during the first year. The strongest predictors of recidivism included the presence of a personality disorder, and in particular, a Cluster B personality disorder (i.e., antisocial, borderline, and/or narcissistic). In addition, those offenders with both a personality disorder and a history of substance abuse were significantly more likely to reoffend compared to either of these risk factors alone. Surprisingly, the presence of a delusional disorder (e.g., erotomania) was associated with a lower risk of reoffender. The findings are discussed in terms of the legal system and treatment implications.