Stephen D. Gottfredson

Bio: Stephen D. Gottfredson is an academic researcher from Virginia Commonwealth University. The author has contributed to research in topics: Criminal justice & Generalizability theory. The author has an hindex of 2, co-authored 2 publications receiving 277 citations.

Statistical Risk Assessment: Old Problems and New Applications

Abstract: Statistically based risk assessment devices are widely used in criminal justice settings. Their promise remains largely unfulfilled, however, because assumptions and premises requisite to their development and application are routinely ignored and/or violated. This article provides a brief review of the most salient of these assumptions and premises, addressing the base rate and selection ratios, methods of combining predictor variables and the nature of criterion variables chosen, cross-validation, replicability, and generalizability. The article also discusses decision makers’ choices to add or delete items from the instruments and suggests recommendations for policy makers to consider when adopting risk assessments. Suggestions for improved practice, practical and methodological, are made.

Clinical Versus Actuarial Judgments in Criminal Justice Decisions: Should One Replace the Other?

Abstract: IN VIRTUALLY ALL decision-making situations that have been studied, actuarially developed devices outperform human judgments. This is true with respect to psychiatric judgments (see, for example, Meehl, 1965; Gough, 1962; Ennis and Litwack, 1974); graduate school admissions (e.g., Dawes, 1979; Dawes and Corrigan, 1974); prognostic judgments made by sociologists and psychiatrists relative to a parole-violation criterion (Glaser, 1955, 1962); parole board decisions (Gottfredson, 1961; Gottfredson and Beverly, 1962; Carroll, Wiener, Coates, Galegher, & Alibrio, 1982); mental health and correctional case worker judgments of offender risk (Holland, Holt, Levi, & Beckett, 1983), spousal assault (Hilton and Harris, 2005) and in other areas (Goldberg, 1970), including the analysis of credit risk (Somerville and Taffler, 1995). Indeed, a recent review and meta-analysis of 56 years’ accumulation of research on the “clinical vs. statistical” prediction “problem” conducted as part of a Festschrift for Paul E. Meehl, a pioneer in the field, again confirms that statistical models outperform clinical decision-makers (Ægisdóttier, White, Spengler, Maugherman, Anderson, Cook, Nichols, Lampropoulos, Walker, Cohen and Rush, 2006).

A conceptual framework for adaptive preventive interventions

Abstract: Recently, adaptive interventions have emerged as a new perspective on prevention and treatment. Adaptive interventions resemble clinical practice in that different dosages of certain prevention or treatment components are assigned to different individuals, and/or within individuals across time, with dosage varying in response to the intervention needs of individuals. To determine intervention need and thus assign dosage, adaptive interventions use prespecified decision rules based on each participant's values on key characteristics, called tailoring variables. In this paper, we offer a conceptual framework for adaptive interventions, discuss principles underlying the design and evaluation of such interventions, and review some areas where additional research is needed.

When clinical description becomes statistical prediction.

Abstract: This article reconsiders the issue of clinical versus statistical prediction. The term clinical is widely used to denote 1 pole of 2 independent axes: the observer whose data are being aggregated (clinician/expert vs. lay) and the method of aggregating those data (impressionistic vs. statistical). Fifty years of research suggests that when formulas are available, statistical aggregation outperforms informal, subjective aggregation much of the time. However, these data have little bearing on the question of whether, or under what conditions, clinicians can make reliable and valid observations and inferences at a level of generality relevant to practice or useful as data to be aggregated statistically. An emerging body of research suggests that clinical observations, just like lay observations, can be quantified using standard psychometric procedures, so that clinical description becomes statistical prediction.

Risk Terrain Modeling: Brokering Criminological Theory and GIS Methods for Crime Forecasting

Abstract: The research presented here has two key objectives. The first is to apply risk terrain modeling (RTM) to forecast the crime of shootings. The risk terrain maps that were produced from RTM use a range of contextual information relevant to the opportunity structure of shootings to estimate risks of future shootings as they are distributed throughout a geography. The second objective was to test the predictive power of the risk terrain maps over two six‐month time periods, and to compare them against the predictive ability of retrospective hot spot maps. Results suggest that risk terrains provide a statistically significant forecast of future shootings across a range of cut points and are substantially more accurate than retrospective hot spot mapping. In addition, risk terrain maps produce information that can be operationalized by police administrators easily and efficiently, such as for directing police patrols to coalesced high‐risk areas.

A Signaling Perspective on Employment-Based Reentry Programming Training Completion as a Desistance Signal

Abstract: Acriminal history is an undeniably excellent predictor of future criminal behavior (Gendreau, Little, and Goggin, 1996). Yet despite stability in offending over time, most offenders, even those with serious criminal histories, eventually desist from crime (Brame, Bushway, and Paternoster, 2003; Kurlychek, Brame, and Bushway, 2006, 2007, in press; Langan and Levin, 2002; Maruna, 2001). Employment-based reentry programs are designed and evaluated with an eye toward hastening the desistance process for individuals with criminal history records (Drake, Aos, and Miller, 2009). In other words, these programs rely on employment and training to causally lower an individual’s likelihood of recidivism. Employment is an obvious starting point in the reentry process because it is the major “routine activity” of most adults, and individuals who are exiting prison look to work as their major source of legitimate income (Bucklen and Zajac, 2009; LeBel, Burnett, Maruna, and Bushway, 2008). Yet the most recent meta-analysis of experimental evaluations of noncustodial employment programs for individuals with a criminal history finds that

A risk index for 12-month suicide attempts in the National Comorbidity Survey Replication (NCS-R).

Abstract: Background. Clinical judgments about the likelihood of suicide attempt would be aided by an index of risk factors that could be quickly assessed in diverse settings. We sought to develop such a risk index for 12-month suicide attempts among suicide ideators. Method. The National Comorbidity Survey Replication (NCS-R), a household survey of adults aged 18+, assessed the 12-month occurrence of suicide ideation, plans and attempts in a subsample of 5692 respondents. Retrospectively assessed correlates include history of prior suicidality, socio- demographics, parental psychopathology and 12-month DSM-IV disorders. Results. Twelve-month prevalence estimates of suicide ideation, plans and attempts are 2 . 6, 0 . 7 and 0 . 4% respectively. Although ideators with a plan are more likely to make an attempt (31 . 9%) than those without a plan (9 . 6 %), 43 % of attempts were described as unplanned. History of prior attempts is the strongest correlate of 12-month attempts. Other significant correlates include shorter duration of ideation, presence of a suicide plan, and several sociodemographic and parental psychopathology variables. Twelve-month disorders are not powerful correlates. A four-category summary index of correlates is strongly related to attempts among ideators (area under the receiver operator characteristic curve (AUC)=0 . 88). The distribution (conditional probability of attempt) of the risk index is : 19 . 0 % very low (0 . 0%), 51 . 1 % low (3 . 5 %), 16 . 2% intermediate (21 . 3 %), and 13 . 7% high (78 . 1%). Two-thirds (67 . 1 %) of attempts were made by ideators in the high-risk category. Conclusions. A short, preliminary risk index based on retrospectively reported responses to fully structured questions is strongly correlated with 12-month suicide attempts among ideators, with a high concentration of attempts among high-risk ideators.