Laura S. Guy

Psychosis as a risk factor for violence to others: a meta-analysis

The potential association between psychosis and violence to others has long been debated. Past research findings are mixed and appear to depend on numerous potential moderators. As such, the authors conducted a quantitative review (meta-analysis) of research on the association between psychosis and violence. A total of 885 effect sizes (odds ratios) were calculated or estimated from 204 studies on the basis of 166 independent data sets. The central tendency (median) of the effect sizes indicated that psychosis was significantly associated with a 49%-68% increase in the odds of violence. However, there was substantial dispersion among effect sizes. Moderation analyses indicated that the dispersion was attributable in part to methodological factors, such as study design (e.g., community vs. institutional samples), definition and measurement of psychosis (e.g., diagnostic vs. symptom-level measurement, type of symptom), and comparison group (e.g., psychosis compared with externalizing vs. internalizing vs. no mental disorder). The authors discuss these findings in light of potential causal models of the association between psychosis and violence, the role of psychosis in violence risk assessment and management, and recommendations for future research.

Does psychopathy predict institutional misconduct among adults? A meta-analytic investigation

Narrative reviews have raised several questions regarding the predictive validity of the Hare Psychopathy Checklist-Revised (PCL-R; R. D. Hare, 2003) and related scales in institutional settings. In this meta-analysis, the authors coded 273 effect sizes to investigate the association between the Hare scales and a hierarchy of increasingly specific forms of institutional misconduct. Effect sizes for Total, Factor 1, and Factor 2 scores were quite heterogeneous overall and weakest for physically violent misconduct (r-sub(w) = .17, .14, and .15, respectively). Moderator analyses suggested that physical violence effect sizes were smaller in U.S. prison samples (r-sub(w) = .11) than in non-U.S. prison samples (r-sub(w) = .23). Findings are discussed in terms of the utility of the Hare measures for decision-making in institutional and other contexts.

Historical-Clinical-Risk Management-20, Version 3 (HCR-20V3): Development and Overview

The HCR-20 Version 3 (HCR-20V3) was published in 2013, after several years of development and revision work. It replaces Version 2, published in 1997, on which there have been more than 200 disseminations based on more than 33,000 cases across 25 countries. This article explains (1) why a revision was necessary, (2) the steps we took in the revision process, (3) key changes between Version 2 and Version 3, and (4) an overview of HCR-20V3s risk factors and administration steps. Recommendations for evaluating Version 3 are provided.

Screening for Malingered Psychopathology in a Correctional Setting: Utility of the Miller-Forensic Assessment of Symptoms Test (M-FAST)

Evidence of construct validity and generalizability for a new measure developed to screen for malingered psychopathology, the Miller-Forensic Assessment of Symptoms Test (M-FAST), is presented for a sample of incarcerated males (N = 50) who had applied for mental health services in a maximum-security prison. Participants completed a brief diagnostic interview, the Structured Interview of Reported Symptoms (SIRS), and the M-FAST. SIRS-defined malingerers scored significantly higher on the M-FAST total and scale scores. Receiver operating characteristic analysis yielded an area under the curve of .92 (SEM = .04, p= .01). Consistent with previous M-FAST validity research, utility results indicated accurate classification was best achieved with an M-FAST total cutoff score of 6 (positive predictive power = .78, negative predictive power =.89). Utility analyses across race produced almost identical results indicating preliminary generalizability of the M-FAST for African American, Hispanic, and Caucasian inmates.

The Role of Psychopathic Personality Disorder in Violence Risk Assessments Using the HCR-20

Antisocial and psychopathic traits are essential to evaluate when assessing risk for violence using the HCR-20. The role of the PCL-R on the HCR-20 was investigated using a series of meta-analytic tests. Across 34 samples in which both tools were rated, AUCs for violence were similar (∼.69), and exclusion of the psychopathy item (H7) did not reduce the HCR-20s accuracy. Quantitative synthesis of results from multivariate analyses conducted in 7 raw datasets that used both tools demonstrated that the average probability of observing violence for every point increase on the HCR-20 (without H7), while controlling for the PCL-R, was 23%, whereas for the PCL-R it was -1%. The HCR-20 (without H7) added incremental validity to the PCL-R, whereas the converse was not true, and only the HCR-20 (without H7) possessed unique predictive validity. Results suggest the HCR-20s predictive validity was not negatively impacted by excluding the PCL-R. Areas for future study are discussed, including research on various ways to assess and incorporate into risk assessment personality traits related to violence.

Assessing Risk of Violence Using Structured Professional Judgment Guidelines

Violence risk assessments are conducted routinely in psychiatric and correctional settings. One method with demonstrated reliability and validity for assessing risk of violence is the structured professional judgment (SPJ) model. In this article, we provide an overview of the SPJ model and a brief review of the empirical literature supporting its use. We present a clinical case example to demonstrate the use of the HCR-20, the most well-researched SPJ tool, with a psychiatric patient being considered for increased hospital privileges and discharge to the community. We conclude with recommendations for clinical practice using an SPJ tool when assessing risk of violence.

Does Risk Assessment Make a Difference? Results of Implementing the SAVRY in Juvenile Probation

An effective approach to reducing recidivism is, first, to identify a youths risk of reoffending and then to match the intensity of interventions to that risk level. This pre-post quasi-experimental, prospective study compared 247 (pre) with 217 (post) adjudicated youths to examine the implementation of the Structured Assessment of Violence Risk in Youth (SAVRY) and its effects on case management practices in Louisianas Caddo parish probation office. The results indicated that placement rates dropped by 50%, use of maximum levels of supervision dropped by almost 30%, and use of community services decreased except for high-risk youths, but only after the SAVRY was properly implemented. This shift towards more appropriate allocation of resources that are matched to risk level occurred without a significant increase in reoffending. The implications for implementation and for use of risk/needs assessment in juvenile probation are discussed.

The Personality Assessment Inventory as a proxy for the Psychopathy Checklist Revised: testing the incremental validity and cross-sample robustness of the Antisocial Features Scale

The Personality Assessment Inventorys (PAIs) ability to predict psychopathic personality features, as assessed by the Psychopathy Checklist–Revised (PCL-R), was examined. To investigate whether the PAI Antisocial Features (ANT) Scale and subscales possessed incremental validity beyond other theoretically relevant PAI scales, optimized regression equations were derived in a sample of 281 Canadian federal offenders. ANT, or ANT–Antisocial Behavior (ANT-A), demonstrated unique variance in regression analyses predicting PCL-R total and Factor 2 (Lifestyle Impulsivity and Social Deviance) scores, but only the Dominance (DOM) Scale was retained in models predicting Factor 1 (Interpersonal and Affective Deficits). Attempts to cross-validate the regression equations derived from the first sample on a sample of 85 U.S. sex offenders resulted in considerable validity shrinkage, with the ANT Scale in isolation performing comparably to or better than the statistical models for PCL-R total and Factor 2 scores. Results offer limited evidence of convergent validity between the PAI and the PCL-R.

Scoring Subjectivity and Item Performance on Measures Used to Assess Violence Risk: The PCL-R and HCR-20 as Exemplars.

Although reliability is essential to validity, most research on violence risk assessment tools has paid little attention to strategies for improving rater agreement. The authors evaluated the degree to which perceived subjectivity in scoring guidelines for items from two measures—the Psychopathy Checklist–Revised (PCL-R) and the Historical, Clinical, Risk Management-20 (HCR-20)—were related to indices of interrater agreement reported in the research literature. Nine doctoral students with experience scoring the PCL-R and HCR-20 showed a high level of agreement with respect to the amount of subjectivity required to score items. Decreased subjectivity was associated with increased rater agreement for both tools. For the PCL-R, items on aggregate measures that have the strongest effects in predictive validity research (i.e., Factor 2, Facet 4) were rated as the least subjective. For the HCR-20, items on the Historical scale were rated as the least subjective. Implications for measure development and refinement are discussed.

Sex Offender Risk Assessment: A Cautionary Note Regarding Measures Attempting to Quantify Violence Risk

ABSTRACT In recent years, the field of forensic psychology has witnessed a proliferation of instruments designed to ascertain the degree of risk of sex offenders who may be released into the community and/or who may be subject to civil commitment. Although few would argue against the importance of basing evaluations and recommendations on empirical research, practitioners sometimes draw conclusions that go well beyond what reasonably can be inferred when using some of these new ‘scientific’ risk instruments. In this article, we review some of the limitations of current measures used in sex offender risk assessments and discuss how these limitations place constraints on what can and should be communicated regarding results from these instruments. We conclude with a review of ethical concerns that arise from a failure to acknowledge the limitations of these measures.