情感计算技术如何推动犯罪风险评估工具的发展？

摘 要 随着近年来人工智能深度学习技术的发展,情感计算与人格计算技术日渐成熟,在许多实际应用场景中取得了良好的表现,当前人工智能情感计算技术应用于犯罪风险评估领域,能够有效解决目前主流的风险评估工具难以解决的个体内差异性的预测因子评估问题以及被测评参与者因社会赞许性而导致结果失真的问题。本文在详细阐述目前主流评估工具的局限性基础上,详细阐述了以情感计算技术为支撑的动态风险评估工具的设计思路、目前已有的技术方案以及设计细节的理论依据,在此基础上最后提出以人工智能技术为支撑的新型评估工具的未来发展方向。     

Open risk assessment

As criminal justice actors increasingly seek to rely on more evidence-informed practices, including risk assessment instruments, they often lack adequate information about the evidence that informed the development of the practice or the tool. Open science practices, including making scientific research and data accessible and public, have not typically been followed in the development of tools designed for law enforcement, judges, probation, and others. This is in contrast to other government agencies, which often open their processes to public notice and comment. Lack of transparency has become pressing in the area of risk assessment, as entire judicial systems have adopted some type of risk assessment scheme. While the types of information used in a risk tool may be made public, often the underlying methods, validation data, and studies are not – nor are the assumptions behind how a level of risk gets categorized as “high” or “low.” We discuss why those concerns are relevant and important to the new risk assessment tool now being used in federal prisons, as part of the First Step Act. We conclude that a number of key assumptions and policy choices made in the design of that tool are not verifiable or are inadequately supported, including the choice of risk thresholds and the validation data itself. Unfortunately, as a result, the federal risk assessment effort has not been the hoped-for model for open risk assessment.     

Integrating base rate data in violence risk assessments at capital sentencing

Prediction of violence in capital sentencing has been controversial. In the absence of a scientific basis for risk assessment, mental health professionals offering opinions in the capital sentencing context are prone to errors. Actuarial or group statistical data, known as base rates, have proven far superior to other methods for reducing predictive errors in many contexts, including risk assessment. Actuarial follow-up data on violent recidivism of capital murderers in prison and post release have been compiled and analyzed to demonstrate available base rates for use by mental health experts conducting risk assessments pertaining to capital sentencing. This paper also reviews various methods for individualizing the application of base rates to specific cases. © 1998 John Wiley & Sons, Ltd.     

Construct Measurement Quality Improves Predictive Accuracy in Violence Risk Assessment: an Illustration Using the Personality Assessment Inventory

Much of the risk assessment literature has focused on the predictive validity of risk assessment tools. However, these tools often comprise a list of risk factors that are themselves complex constructs, and focusing on the quality of measurement of individual risk factors may improve the predictive validity of the tools. The present study illustrates this concern using the Antisocial Features and Aggression scales of the Personality Assessment Inventory (Morey, 1991). In a sample of 1,545 prison inmates and offenders undergoing treatment for substance abuse (85% male), we evaluated (a) the factorial validity of the ANT and AGG scales, (b) the utility of original ANT and AGG scales and newly derived ANT and AGG scales for predicting antisocial outcomes (recidivism and institutional infractions), and (c) whether items with a stronger relationship to the underlying constructs (higher factor loadings) were in turn more strongly related to antisocial outcomes. Confirmatory factor analyses (CFAs) indicated that ANT and AGG items were not structured optimally in these data in terms of correspondence to the subscale structure identified in the PAI manual. Exploratory factor analyses were conducted on a random split‐half of the sample to derive optimized alternative factor structures, and cross‐validated in the second split‐half using CFA. Four‐factor models emerged for both the ANT and AGG scales, and, as predicted, the size of item factor loadings was associated with the strength with which items were associated with institutional infractions and community recidivism. This suggests that the quality by which a construct is measured is associated with its predictive strength. Implications for risk assessment are discussed. Copyright © 2013 John Wiley & Sons, Ltd.     

Risk assessment in sentencing and plea bargaining: The roles of prosecutors and defense attorneys

Research on risk assessment in sentencing has focused heavily on the role of judges. Ignoring the role of other courtroom actors in the sentencing process, however, leaves unexamined the potentially significant effects on judicial decision making of arguments made by prosecutors and defense attorneys at sentencing hearings. Unduly focusing on judges also overlooks the vast majority of sentences arrived at through negotiated guilty pleas. We explored the extent to which considerations of risk are made among prosecutors and defense attorneys when advocating for given sentences in open court or during plea negotiations. We surveyed all prosecutors and defense attorneys in 14 judicial circuits in Virginia and found that most prosecutors and defense attorneys at least “sometimes” explicitly invoked actuarial risk estimates both at sentencing hearings and during plea negotiations. However, defense attorneys were much more likely than prosecutors to be averse to the use of risk assessment in either form of case disposition.     

Stakes and risk: Incapacitative intent in sentencing decisions

Measures of seriousness of the conviction offense and prior criminal record are strong correlates of sentences. This is consistent with a desert orientation to sentencing. Competing sentencing theories emphasize concepts of risk and incapacitation, but their roles in sentencing decisions have received less attention. This study focused on judicial risk assessments and on the concept of stakes. Both variables plausibly are related to incapacitative intents of judges. Our hypotheses that the risk and stakes measures used are relatively independent and correlated with, and have an interaction effect on, decision outcomes were supported. Both judges' assessments of risk and our stakes scale accounted for substantial variation in the decision to incarcerate. Time actually served in confinement also was related substantially to stakes, risk, and their interaction. Implications of the stakes and risk concepts for future guidelines development and decision study in criminal justice are discussed.     

履历数据测评的效度分析

履历数据是人员测评的重要手段, 但其组织情景性限制使得国外同类工具不能适用于国内人事选拔, 国内也缺乏实证研究验证其信度、效度。本研究结合某通信企业的招募选拔工作, 开发履历数据分析工具, 选取250名应聘者的履历数据、一般认知能力、大五个性和面试结果数据, 通过效标关联效度、增量效度思想, 分析履历数据有效性。结果表明, 以面试结果为效标, 履历数据具有良好的效标关联效度, 与其他测评工具组合使用时有良好的增量效度。     

A matter of life or death: special considerations and heightened practice standards in capital sentencing evaluations

Mental health evaluations at capital sentencing represent a complex and specialized arena of practice. The moral culpability focus of capital sentencing is distinct from guilt-phase considerations of criminal responsibility, and has a specialized literature. Capital violence risk assessment is uniquely oriented to a prison context, relying on past adjustment to incarceration, as well as group statistical data specific to capital offenders and other inmate groups. Personality testing is a more complex consideration in capital sentencing evaluations. The implications of interviewing the defendant, as well as the parameters and documentation of an interview, make full disclosure and informed consent of particular importance. Defense- and prosecution-retained experts are subject to specific ethical vulnerabilities. These are examined in this paper through the lens of current professional standards.     

Artful Dodgers: The Role of Unreliable Narrators in the Production of Authorized Histories and Assessments of Young People in Conflict with the Law

This article argues that pre-sentence reports for youth courts are co-authored by at least two unreliable narrators: the defendant and the youth justice worker. The production of an objective professional assessment for sentencing purposes is therefore a convenient fiction in which key actors collude. Such reports do not only seek to provide narrative explanations of past actions, but also shape future sentencing decisions and other outcomes that can either be helpful or harmful to service users. Full case files, meanwhile, contain documents that reveal less orderly and more challenging narratives from the margins.     

Sentencing Convicted Juvenile Felony Offenders in the Adult Court: The Direct Effects of Race

While research indicates that Black and Hispanic adults sentenced in the criminal court tend to be rendered more severe punishments than their White counterparts, only one prior study has examined whether this finding holds for juveniles tried in the adult system. The findings from this sole study need replication, however, since the effects posed by trial type were not taken into account and it is likely that the results are confounded by measurement error resulting from overlap in criminal sentencing. The current study addressed these issues by assessing whether race has a direct impact on waived juveniles being criminally sentenced to restitution, probation, or jail. Data were derived from a secondary, cross‐sectional national dataset on felony juvenile offenders convicted in the adult system. Three hypotheses were tested. After controlling for a number of important legal and extra‐legal predictors of sentencing, race differences in sentencing outcomes were observed and the findings yielded partial support for the hypotheses. The implications of the research are noted. Copyright © 2012 John Wiley & Sons, Ltd.     

The sexist algorithm

Algorithmic risk assessment tools are informed by scientific research concerning which factors are predictive of recidivism and thus support the evidence‐based practice movement in criminal justice. Automated assessments of individualized risk (low, medium, high) permit officials to make more effective management decisions. Computer‐generated algorithms appear to be objective and neutral. But are these algorithms actually fair? The focus herein is on gender equity. Studies confirm that women typically have far lower recidivism rates than men. This differential raises the question of how well algorithmic outcomes fare in terms of predictive parity by gender. This essay reports original research using a large dataset of offenders who were scored on the popular risk assessment tool COMPAS. Findings indicate that COMPAS performs reasonably well at discriminating between recidivists and non‐recidivists for men and women. Nonetheless, COMPAS algorithmic outcomes systemically overclassify women in higher risk groupings. Multiple measures of algorithmic equity and predictive accuracy are provided to support the conclusion that this algorithm is sexist.     

Risk assessment savvy: The imperative of appreciating accuracy and outcome

This article focuses on two aspects of actuarial risk at sentencing, the accuracy of the instrument and the outcome it predicts. For theoretical reasons rooted in the cognitive decision-making and sentencing literature, there is a danger that judges and other practitioners might come to overly rely on a “high risk” label or designation without appreciating the accuracy of the prediction or the actual outcome being predicted. Using sentencing and recidivism data from Pennsylvania (n = 10,000), two simple risk instruments are constructed to illustrate the critical importance of understanding accuracy and outcome before relying on the risk tool information.     

Environmental Decision Making on Acid Mine Drainage Issues in South Africa: An Argument for the Precautionary Principle

This paper examines the issue of acid mine drainage in South Africa and environmental decision making processes that could be taken to mitigate the problem in the context of both conventional risk assessment and the precautionary principle. It is argued that conventional risk assessment protects the status quo and hence cannot be entirely relied upon as an effective tool to resolve environmental problems in the context of South Africa, a developing country with complex environmental health concerns. The complexity of the environmental issues is discussed from historical and political perspectives. An argument is subsequently made that the precautionary principle is an alternative tool, and its adoption can be used to empower local communities. This work, therefore, adds to new knowledge by problematising conventional risk assessment and proposing the framing of the acid mine drainage issues in a complex and contextual scenario of a developing country—South Africa.     

Examining Predictors Which May Influence Sentencing Outcomes for Individuals Convicted of Offenses against Children: A Brief Report

Numerous studies examine sentencing decisions, yet little attention has been given to sentencing of child maltreatment and, more specifically, on variables that could impact sentencing outcomes for this form of criminality. Using Pennsylvania Sentencing Data for 2006, this research adds to the existing sentencing research by exploring predictors that might influence sentencing decisions for individuals convicted of crimes against children. The findings indicate that offender sex significantly affects the sentencing decision and offender age significantly affects sentencing length. In addition, all offense characteristics significantly influence the sentencing length decision. Implication of these results and directions for future research are discussed.     

Can the criminal law ever be therapeutic?

This paper addresses the question of whether criminal law, especially in its sentencing and penological aspects, can be adjusted in its workings to incorporate the perspective of therapeutic jurisprudence. A conceptual framework developed by Wexler, Winick, and their colleagues is employed to provide a basis for reflection on the operation and impact of sentencing processes. Of considerable importance in this exercise is a mounting volume of evidence concerning the outcomes of work with adjudicated offenders, which illustrates the potential of psycho-educational, behavioral-cognitive skills-training, and therapeutic programs for reduction of recidivism. Such intervention programs are currently offered within the limits set by, but are not intrinsic components of, judicial decisions. The psycho-legal implications of these findings are discussed and some possibilities surveyed for tentative exploration of actively therapeutic departures within crimino-legal decision-making.     

The Assessment of Normal Fear in Children and Adolescents

Published research into normal fear now spans more than one century. During this time, a large number of papers have been published in the area. The resulting literature has led to a detailed understanding of normal fear experiences and, in particular, the ways in which they change with maturation. Of central importance, when evaluating the documented outcomes of this work, is the soundness of the methods and tools used in assessment. It is not surprising given the large number of researchers that have been involved in this area, that the assessment methods used have varied substantially. These have ranged from the methodologically problematic technique of obtaining retrospective adult reports to the administration of psychometrically validated fear survey schedules. An extensive review of this literature reveals that, for the last two decades, the fear survey schedule has been the most widely used technique for the fear assessment. The preference that has and is being demonstrated for the fear survey schedule as an assessment tool is most likely due to its many advantages including ease of use, objectivity in scoring, and provision of a substantial amount of information in a short period. However, despite its advantages, researchers and clinicians need to be cognizant of its potential limitations and, depending on the questions being asked, may need to consider using it in combination with alternative assessment strategies.     

A Screening Tool to Assess and Manage Behavioral Risk in the Postoperative Bariatric Surgery Patient: The WATCH

Bariatric surgery is increasingly recognized as a highly effective treatment for individuals who are severely obese. Amount of weight loss and resolution of comorbidities surpass those of nonsurgical approaches; however, suboptimal weight loss and weight regain are not uncommon. These outcomes, though not fully understood, are likely at least partially explained by failure to make long-term behavioral and/or cognitive changes. We are unaware of any established clinical tools to guide providers in assessing postoperative behaviors and identifying those who may require specialized treatment. The goal of this paper is to introduce a brief screening tool, The WATCH, to help clinicians assess and identify patients who may be at risk for poor or untoward outcomes post bariatric surgery. We first review the literature on postoperative outcomes, including weight loss, resolution of comorbidities, suboptimal outcomes, and development of problematic eating behaviors. We then provide an easily-recalled, five-item tool that assesses outcomes, and discuss patient responses that may necessitate further intervention or referral.     

Preventative Detention Decisions: Reliance on Expert Assessments and Evidence of Partisan Allegiance within the Canadian Context

The purpose of this study was to examine judges’ written reasons for sentencing in preventative detention hearings and the expert risk assessment reports presented, to determine the level of reliance placed on expert risk assessment reports and to examine the presence of partisan allegiance within the Canadian context. Results demonstrated that judges’ decisions were consistent with expert assessments in terms of risk, treatment amenability, and risk management. Experts’ ratings of treatment amenability and risk management were also significant predictors of the designation outcome, indicating that judges rely on this information in making their final decision. Finally, there was evidence of partisan allegiance, with prosecution-retained Psychopathy Checklist-Revised scores being significantly higher than defense-retained experts’ scores. The results have implications for the development of consistent guidelines for the communication of risk, treatment amenability, and management information. Copyright © 2015 John Wiley & Sons, Ltd.     

Professional judgement in the assessment of risk: is there a role for systemic practice?

Risk assessment is an example of professional decision‐making pared to its stark essentials. Political pressures towards accountability and the need for defensible decisions encourage a ‘tick‐box’ approach to risk assessment, but this can create unrealistic expectations of certainty. In practice, as technological approaches produce ever more complex formal tools for assessing risk, their effectiveness remains dubious while our human decision‐making apparatus is marginalized. This article examines whether we should respect our ability to apprehend complex multi‐stranded narrative realities intuitively, with the hope that such intuitions might contribute to professional decision‐making. This idea is explored with reference to a case study.     

In a search for meaning: challenging the accepted know-how of working with suicide risk

This opinion piece considers the current predominance of assessment tools and strategies in working with people at risk of suicide, and questions their efficacy and how they are privileged in day to day mental health practice. While such tools and an evidence-based ‘scientific’ approach to assessment clearly has its place, the author instead asserts that the modus operandi of therapy – a discursive based exploration – has much more to offer and should be the primary intervention in understanding suicide potential. Helping the client to gain insight into the meaning of their suicidality helps position the client – and practitioner – in the best possible place to reduce risk.