Parole decision-making: A salient factor score

Although considerable effort during the past thirty years has been devoted to the attempted construction of actuarial aids for parole selection, such devices have not generally been adopted by paroling agencies for operational use. One recent exception is the United States Board of Parole which has commenced usage of an actuarial device, termed a “salient factor score,” as a risk assessment aid in conjunction with explicit parole decision-making guidelines. This paper describes the construction, validation, and operational usage of this device.     

Salient factor score validation—A 1972 release cohort

In October 1973, the United States Board of Parole adopted an actuarial device, termed a “salient factor score”, as a risk assesment aid. This device was constructed and initially validated on a sample of 1970 releases utilizing a two year follow-up period for each case. However, as the utility of a predictive device depends upon its a ability to predict to prospective samples, periodic validation is required. This paper describes the application of the salient factor score to a subsequent validation sample of prisoners released during 1972 and discusses the effects of using various outcome measures.     

A comparison of actuarial methods for identifying repetitively violent patients with mental Illnesses

This is a progress report on the development of practical methods for the actuarial prediction of violence. The literature indicates that actuarial prediction is more accurate than clinical prediction, but in practice actuarial methods seem to be used rarely. Here we address two obstacles to the clinical use of actuarial prediction methods. First, clinicians may be averse to actuarial methods that require calculations. To remedy this, we developed a regression tree screen that presents actuarial information about violence in a series of yes/no questions. Second, using actuarial methods to identify the small minority of violent patients in a general psychiatric population may be too costly. To remedy this, we developed a method to prescreen patients for intensive evaluation using an inexpensive assessment. We evaluated regression trees and two-stage screening by comparing their accuracies against conventional actuarial methods. The results showed that actuarial predictions based on regression trees and two-stage screens were as accurate as regression-based methods in identifying repetitively violent patients. These easier-to-use methods may therefore be useful techniques for actuarial predictions.     

A Bayesian approach to the group versus individual prediction controversy in actuarial risk assessment

Recent attempts to indict the use of actuarial risk assessment instruments have relied on confidence intervals to demonstrate that risk estimates derived at the group level do not necessarily apply to any specific individual within that group. This article contends that frequentist confidence intervals are inapposite to the current debate. Instead, Bayesian credible intervals are necessary-in principle-to accomplish what commentators are concerned about: describing the precision of an actuarial risk estimate. After illustrating both the calculation and interpretation of credible intervals, this article shows how such intervals can be used to characterize the precision of actuarial risk estimates. It then explores the legal implications of wide and overlapping intervals. Contrary to what detractors claim, the fact that risk estimate intervals overlap is not a germane to legal (logical) relevance, and therefore actuarial risk estimates cannot be per se "inadmissible" on this basis.     

The Rise of Criminal Background Screening in Rental Housing

This article investigates the recent rise of criminal background screening in rental housing as a case study of the diffusion of actuarial social control. That case study suggests that actuarial techniques have spread more widely through the crime prevention field than sociolegal scholars have recognized, replacing disciplinary efforts to diagnose and alter the behavior of individuals with actuarial efforts to identify and isolate high-risk groups. This actuarial strategy has proliferated not only because new discourses encouraged it but also because new institutional structures facilitated it. That conclusion illustrates the importance of structural (rather than cultural) factors in shaping society's response to crime—particularly the growing availability of the collective institutional capacity that actuarial social control requires .     

Salient factor score and releasee behavior: Three validation samples

This paper provides an opportunity to observe the application of an actuarial device, used by the United States Parole Commission as an aid in parole selection, to three separate validation samples—each exceeding 1000 cases. Follow-up data for this study were provided through the cooperation of the Federal Bureau of Investigation. In addition, evidence is presented which may shed light on the conflict in the literature regarding whether or not the rate of unfavorable outcome of released prisoners at risk actually declines substantially after the first year (i.e., whether or not the first year after release is really the most critical).Adapted from Report Fifteen, United States Parole Commission Research Unit, August 1977.     

Actuarial Sentencing: An “Unsettled” Proposition

This paper discusses the concerns associated with the introduction of, and increased reliance on, actuarial risk tools in sentencing in order to: (1) stimulate cross-disciplinary dialog and research about the impact of incorporating actuarial risk logic into sentencing processes and (2) identify questions requiring further empirical examination. In this article, I recognize that actuarial risk logic offers managerial and organizational benefits, but I also demonstrate that the application of actuarial risk when sentencing offenders is not without important consequences. First, I provide a brief outline of the emergence, logic, and entrenchment of probabilistic reasoning within criminal justice decision-making, and the more recent extension and application of actuarial risk logic to sentencing. Then, I use the following themes to define the limits of using risk sciences in sentencing: (1) the logical structure of risk; (2) the slippage between risk prediction and individual causation; (3) current methodological limits of risk science; (4) the potential for gender and race discrimination; (5) the legal relevance and transparency of risk-based sentencing; and (6) the jurisprudential and organizational impact of various risk technologies. Importantly, the nature and severity of these complications will vary by, and within, the jurisdiction (or sentencing regime) because current sentencing practices are influenced by local jurisdictional needs and sentencing laws.     

Applying a forensic actuarial assessment (the Violence Risk Appraisal Guide) to nonforensic patients

The actuarial Violence Risk Appraisal Guide (VRAG) was developed for male offenders where it has shown excellent replicability in many new forensic samples using officially recorded outcomes. Clinicians also make decisions, however, about the risk of interpersonal violence posed by nonforensic psychiatric patients of both sexes. Could an actuarial risk assessment developed for male forensic populations be used for a broader clientele? We modified the VRAG to permit evaluation using data from the MacArthur Violence Risk Assessment Study that included nonforensic male and female patients and primarily self-reported violence. The modified VRAG yielded a large effect size in the prediction of dichotomous postdischarge severe violence over 20 and 50 weeks. Accuracy of VRAG predictions was unrelated to sex. The results provide evidence about the robustness of comprehensive actuarial risk assessments and the generality of the personal factors that underlie violent behavior.     

Exploring the validity of the Level of Service Inventory-Revised with Native American offenders

The use of actuarial risk/need assessment tools is an increasingly important part of the correctional landscape. Actuarial tools ideally will provide a valid, dynamic assessment of an offender's overall risk/need level, and will identify their most prevalent criminogenic needs. What results is typically a number or score that can be used to assign an offender to a risk level that is associated with an assumed likelihood of recidivism. Testing the predictive validity of actuarial risk/need assessment tools is of paramount concern, particularly when they are utilized with new (and under-researched) populations. The current study assessed the predictive validity of the Level of Service Inventory-Revised using a sample of Native American and White offenders in a northern midwestern state. Results showed the instrument to have modest predictive validity utilizing the entire sample of offenders, with varying results for subsequent subgroups.     

Looking at Prediction from an Economics Perspective: A Response to Harcourt's Against Prediction

This article discusses Bernard E. Harcourt's Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age (2007 ). The book opposes the use of probabilistic methods, such as profiling, on efficiency, equity, and jurisprudence grounds. By contrast I argue that profiling is always efficient, that there is no theoretical flaw in reliance on actuarial methods, as long as they are implemented properly. I also show that the equity-based criticism of reliance on actuarial methods (Harcourt's ratchet effect argument) is based on two questionable assumptions: that profiling is perfectly efficient (as zero deterrence is assumed), and that the police are making an obvious logical mistake, by gradually increasing the extent to which they target the group with the higher offending rate instead of targeting only them in the first place.     

The Influence of Actuarial Risk Assessment in Clinical Judgments and Tribunal Decisions about Mentally Disordered Offenders in Maximum Security

Research has shown that actuarial assessments of violence risk are consistently more accurate than unaided judgments by clinicians, and it has been suggested that the availability of actuarial instruments will improve forensic decision making. This study examined clinical judgments and autonomous review tribunal decisions to detain forensic patients in maximum security. Variables included the availability of an actuarial risk report at the time of decision making, patient characteristics and history, and clinical presentation over the previous year. Detained and transferred patients did not differ in their actuarial risk of violent recidivism. The best predictor of tribunal decision was the senior clinician's testimony. There was also no significant association between the actuarial risk score and clinicians' opinions. Whether the actuarial report was available at the time of decision making did not alter the statistical model of either clinical judgments or tribunal decisions. Implications for the use of actuarial risk assessment in forensic decision making are discussed.     

The Sexual Violence Risk-20: factor structure and psychometric properties

Although the Sexual Violence Risk-20 (SVR-20) is widely used, its psychometric properties have only been investigated in a limited number of studies. This study explored the factor structure of the SVR-20 and examined its psychometric properties. Confirmatory factor analysis (CFA) was used to examine the fit of the original three-domain model of the SVR-20. The CFA showed that the original structure was not satisfactory. Exploratory principal components analysis (PCA) was conducted in search of a more optimal factor structure. Psychometric properties (i.e., internal consistency, predictive value, and convergent validity) of both the original domains and alternative factors were investigated. The PCA and subsequent CFAs pointed in the direction of an alternative, more optimal three-factor solution. The three alternative factors were labeled as Antisociality, Sexual deviance, and Problematic thinking and produced better internal consistency coefficients than the original domains. However, the validity of the SVR-20 was modest and no evidence was found indicating that the alternative factors were better in this regard as compared to the original domains. Despite the overall superiority of actuarial measures in predicting recidivism, the structured professional judgment of the SVR-20 proved to be more predictive of sexual, violent, and general recidivism than its actuarial scoring method.     

How should we understand the effect of age on sexual recidivism?

Abstract

Research into the effect of age on sexual recidivism risk is a relatively new and developing area of interest and is likely to be of great interest for forensic practitioners responsible for the community supervision of sexual offenders. Meta-analytical and follow-up reconviction studies indicate an inverse relationship between age and sexual recidivism risk, where younger sex offenders pose a greater risk of reconviction than older sex offenders. This finding has led to the development of actuarial risk scales which identify younger sex offenders (<25 years) as posing the greatest risk. However, recent research studies have reported contradictory results to this assumption and found a non-linear relationship between age and sexual recidivism risk. Only a small number of studies have investigated the effect of age on sexual recidivism by comparing age bands and rates of sexual recidivism. Researchers have also considered the effect of age on actuarial risk, which risk factors are associated with which age bands and sexual recidivism rates between sex offender subgroups. The purpose of this paper is to integrate this research and to link commonalities between these studies. This paper organizes the effect of age on sexual recidivism into five categories: (i) the effect of age and actuarial risk; (ii) the effect of age on sexual arousal; (iii) the effect of age-at-release on sexual recidivism risk; (iv) the effect of age-at-first-offence on sexual recidivism risk; and (v) the effect of age on child molesters and rapists on sexual recidivism risk. Important differences were found between age bands of sexual offenders in terms of sexual recidivism risk and actuarial risk factors as well as differences between rapists and child molesters. The relative importance of factoring age when assessing risk in sex offenders is discussed.     

Advances in the assessment and prediction of interpersonal violence

This article underscores the weakness of clinical judgment as a mechanism for prediction with examples from other areas in the psychological literature. Clinical judgment has as its Achilles'heel the reliance on a person to incorporate multiple pieces of information while overcoming human judgment errors--a feat insurmountable thus far. The actuarial approach to risk assessment has overcome many of the weaknesses of clinical judgment and has been shown to be a much superior method. Nonetheless, the static/historical nature of the risk factors associated with most actuarial approaches is limiting. Advances in risk prediction will be found in part in the development of dynamic actuarial instruments that will measure both static/historical and changeable risk factors. The dynamic risk factors can be reevaluated on an ongoing basis, and it is proposed that the level of change in dynamic factors necessary to represent a significant change in overall risk will be an interactive function with static risk factors.     

The Effect of Framing Actuarial Risk Probabilities on Involuntary Civil Commitment Decisions

Despite a proliferation of actuarial risk assessment instruments, empirical research on the communication of violence risk is scant and there is virtually no research on the consumption of actuarial risk assessment. Using a 2 × 3 Latin Square factorial design, this experiment tested whether decision-makers are sensitive to varying levels of risk expressed probabilistically and whether the framing of actuarial risk probabilities is consequential for commitment decisions. Consistent with research on attribute framing, in which describing an attribute in terms of its complement leads to different conclusions, this experiment found that the way actuarial risk estimates are framed leads to disparate commitment decisions. For example, risk framed as 26% probability of violence generally led decision-makers to authorize commitment, whereas the same risk framed in the complement, a 74% probability of no violence, generally led decision-makers to release. This result was most pronounced for moderate risk levels. Implications for the risk communication format debate, forensic practice and research are discussed.     

A Time Series Modeling Approach in Risk Appraisal of Violent and Sexual Recidivism

For over half a century, various clinical and actuarial methods have been employed to assess the likelihood of violent recidivism. Yet there is a need for new methods that can improve the accuracy of recidivism predictions. This study proposes a new time series modeling approach that generates high levels of predictive accuracy over short and long periods of time. The proposed approach outperformed two widely used actuarial instruments (i.e., the Violence Risk Appraisal Guide and the Sex Offender Risk Appraisal Guide). Furthermore, analysis of temporal risk variations based on specific time series models can add valuable information into risk assessment and management of violent offenders.     

Unmanaged care: towards moral fairness in health care coverage

Health insurers are generally guided by the principle of "actuarial fairness," according to which they distinguish among various risks on the basis of cost-related factors. Thus, insurers often limit or deny coverage for vision care, hearing aids, mental health care, and even AIDS treatment based on actuarial justifications. Furthermore, approximately forty-two million Americans have no health insurance at all, because most of these individuals cannot afford the cost of insurance. This Article argues that Americans have come to demand more than actuarial fairness from health insurers and are increasingly concerned by what I call "moral fairness." This is evidenced by the hundreds of laws that have been passed to constrain insurers' discretion with respect to particular coverage decisions. Legislative mandates are frequent, but seemingly haphazard, following no systematic methodology. This Article suggests an analytical framework that can be utilized to determine which interventions are appropriate and evaluates a variety of means by which moral fairness could be promoted in the arena of health care coverage.     

A Classification Tree Approach to the Development of Actuarial Violence Risk Assessment Tools

Since the 1970s, a wide body of research has suggested that the accuracy of clinical risk assessments of violence might be increased if clinicians used actuarial tools. Despite considerable progress in recent years in the development of such tools for violence risk assessment, they remain primarily research instruments, largely ignored in daily clinical practice. We argue that because most existing actuarial tools are based on a main effects regression approach, they do not adequately reflect the contingent nature of the clinical assessment processes. To enhance the use of actuarial violence risk assessment tools, we propose a classification tree rather than a main effects regression approach. In addition, we suggest that by employing two decision thresholds for identifying high- and low-risk cases--instead of the standard single threshold--the use of actuarial tools to make dichotomous risk classification decisions may be further enhanced. These claims are supported with empirical data from the MacArthur Violence Risk Assessment Study.     

Accuracy of the static-99 in predicting recidivism in Switzerland

The Static-99 is a widely used actuarial risk assessment instrument. Various international validation studies have found satisfactory to good predictive validity for the Static-99, with the area under the curve (AUC) between 59% and 95%. This study is the first evaluation of the predictive accuracy of the Static-99 among sex offenders in Switzerland. The Static-99 scores of 69 violent/sex offenders in Switzerland were assessed using data from their psychiatric assessments. Recidivism was operationalized as reconviction assessed from penal records. The Static-99 risk levels were predictive for recidivism (AUC = .758) among our population. The results are discussed on the basis of the literature.