Actuarial versus clinical assessments of dangerousness.

Abstract

Cormier (1999) proposed the "complete replacement" of clinical assessments of dangerousness with actuarial methods, such as the Violence Risk Appraisal Guide (VRAG). In this article, the author argues that (a) research to date has not demonstrated that actuarial methods of risk assessment are superior to clinical methods; (b) because most clinical determinations of dangerousness are not "predictions" of violence, as well as for other reasons, it is very difficult to meaningfully compare clinical and actuarial assessments of dangerousness; and (c) even the best researched and validated actuarial tool for assessing dangerousness to date, the VRAG, has not yet been validated in a manner that would make it appropriate for use in determining when individuals should be confined on the grounds of their dangerousness. Therefore, although clinicians who engage in risk assessments certainly should be knowledgeable about arguably relevant actuarial assessment schemes and other assessment guides (e.g., the HCR-20), it is premature, at best, to replace clinical risk assessments with actuarial assessments.A longstanding controversy exists within the field of psychology over the relative merits of actuarial (or statistical) versus clinical predictions. Within the field of dangerousness risk assessment, this controversy was heightened in 1999 by the American Psychological Association's publication of the book Violent Offenders: Appraising and Managing Risk by Quinsey, Harris, Rice and Cormier. In this book, the authors argued explicitly and strongly for the "complete replacement" of clinical assessments of dangerousness with actuarial methods. To be precise, they stated: "What we are advising is not the addition of actuarial methods to existing practice, but rather the complete replacement of existing practice with actuarial methods" (p. 171).This article addresses the position taken by those authors regarding the relative merits of actuarial versus clinical assessments of dangerousness and, more generally, the increasingly widespread academic view that actuarial assessments of dangerousness are superior to clinical assessments. Today, one routinely finds statements in the literature suggesting that actuarial assessments of dangerousness have proven to be superior to clinical assessments, or at least, to unstructured clinical assessments (

De l’évaluation clinique au calcul de probabilité : le recours aux outils actuariels dans les pénitenciers canadiens

“…Cet enthousiasme pour le recours aux grilles actuarielles n'est cependant pas partagé par tous. En réaction à cet emballement pour l'utilisation de grilles statistiques, certains chercheurs vont plutôt soulever des doutes quant à la prétendue supériorité actuarielle, tant sur le plan de l'effi cacité prédictive des outils que des fondements méthodologiques de l'approche (Hart et al, 2007 ;Litwack, 2001) 2 . Dans une perspective plus générale, on peut aussi associer cette montée de l'actuarialisme à l'importance prise par le langage du risque dans la façon de penser désormais l'intervention auprès des populations marginales (Cartuyvels, 2008 ;Kemshall 2003).…”

“…Cet exercice nous apparaît d'autant plus nécessaire que les efforts de catégorisation que l'on retrouve dans la littérature scientifi que souffrent en général d'un manque de rigueur sur le plan conceptuel. On retrouve de nombreux travaux portant sur l'analyse comparative de l'effi cacité des grilles cliniques et actuarielles, mais en général on ne réussit pas à y établir de façon rigoureuse les critères pour distinguer ces deux catégories d'outils (Dawes et al, 1989 ;Grove et Meehl, 1996 ;Grubin, 2004 ;Guay, 2006 ;Litwack, 2001). On réduit souvent le caractère actuariel à l'utilisation d'un outil mécanique, négligeant ainsi l'aspect probabiliste qui caractérise toute mesure actuarielle.…”

The effects of rational and experiential information processing of expert testimony in death penalty cases

“…Actuarial predictions of dangerousness are one of the most common methods of risk prediction that have demonstrated superiority over pure clinical predictions in head to head comparisons (see, e.g., Gardner et al, 1996, andCormier, 2002, for two studies, but see Litwack (2001) for why these comparisons may have been lacking). Actuarial prediction instruments are created from empirically verified risk factors (e.g.…”

Static and Dynamic Patient Characteristics as Predictors of Criminal Recidivism: A Prospective Study in a Dutch Forensic Psychiatric Sample.

“…Several authors have expressed doubts whether clinical risk assessment has been fairly judged in the research that is available (Buchanan, 1999;Hart, 1999;Litwack, 2001), and whether risk assessment instruments that are available are of value to routine clinical practice (Buchanan, 1999). Litwack (2002) encourages further study of "the underpinnings of clinical risk assessment" (p. 174).…”