Statistical Risk Assessment: Old Problems and New Applications

By S. D. Gottfredson, L. Moriarty. 2006

238 citations recorded by the source corpus.

Subfields: Clinical Psychology, Sociology and Political Science

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Clinical Psychology

1. J. Skeem
2. Cynthia Rudin
3. Richard A. Berk
4. J. Monahan
5. Berk Ustun
6. Jiaming Zeng
7. Justin Bleich
8. S. Prins

In Sociology and Political Science

1. Richard A. Berk
2. Cynthia Rudin
3. Berk Ustun
4. Jiaming Zeng
5. J. Skeem
6. G. Barnes
7. Justin Bleich
8. J. Monahan

Citations in the corpus

• Interpretable classification models for recidivism prediction (2015, background): “These approaches have led to serious issues in practice (see, e.g., Gottfredson and Moriarty 2006, for a detailed overview).”
• Balancing the Costs of Forecasting Errors in Parole Decisions (2010, background): “The results of this enterprise are mixed (Dean and Dugan, 1968; Farrington, 1987; Gottfredson and Moriarty, 2006; Berk, 2008a).”
• Development of the California Static Risk Assessment (CSRA): Recidivism Risk Prediction in the California Department of Corrections and Rehabilitation (2013, methodology): “…instruments utilize these measures, but studies have begun to show that the predictive value of the more concise, purely static instruments is at least as high as the more complex, dynamic tools (Gottfredson and Moriarty 2006; Barnoski and Drake 2007; Pattavina and Taxman 2007; Baird 2009).”
• Machine Learning Risk Assessments in Criminal Justice Settings (2018, background): “Even less is known about the accuracy of similar forecasts at other criminal justice decision points (Ohlin and Duncan 1949; Glaser 1955; Dean and Dugan 1968; (Wilkins 1980); Farrington and Tarling 1985; Gottfredson and Moriarty 2006; Berk 2008; Skeem and Monahan 2011).”
• An impact assessment of machine learning risk forecasts on parole board decisions and recidivism (2017, background): “But, the past few years have seen increased scrutiny and controversy surrounding the use of actuarial risk assessments for a variety of criminal justice decisions (Gottfredson and Moriarty 2006; Conroy 2006; Berk 2012; Tonrey 2014; Monahan and Skeem 2014).”
• It's COMPASlicated: The Messy Relationship between RAI Datasets and Algorithmic Fairness Benchmarks (2021, background): “CJ, psychology, and other disciplines have studied RAIs for decades and developed them to improve 48 upon unstructured, human decision-making [3].”
• Forecasting Domestic Violence: A Machine Learning Approach to Help Inform Arraignment Decisions (2016, background): “For example, there are about 50 kinds of sexual crimes. concerns (Farrington & Tarling 2003; Berk & Sorenson 2005; Gottfredson & Moriarty 2006; Berk & Bleich 2013; Ridgeway 2013), popular intimate partner risk assessment tools typically do not consider the wide range of crimes that are folded into…”
• Statistical Procedures for Forecasting Criminal Behavior (2013, background): “More recent commentaries have documented a number of other problems, sometimes including no evaluation at all (Farrington and Tarling, 2003; Gottfredson and Moriarty, 2006; Berk, 2012).”
• A Question of Evidence: A Critique of Risk Assessment Models Used in the Justice System (2009, background): “It is for precisely this reason that Gottfredson and Moriarty (2006) conclude: If a factor on a risk assessment instrument has no demonstrated relationship to recidivism, how can positive changes in this factor reduce risk?”
• Asymmetric Loss Functions for Forecasting in Criminal Justice Settings (2011, background): “performance of criminal justice forecasts has been mixed at best (Farrington 1987; Gottfredson and Moriarty 2006 ).”
• A Signaling Perspective on Employment-Based Reentry Programming Training Completion as a Desistance Signal (2012, unknown): “An individual's criminal propensity can only be known within fairly wide confidence limits (Bushway, Sweeten, and Nieuwbeerta, 2009), and any risk prediction exercise will necessarily involve considerable error (Gottfredson and Moriarity, 2006).”
• A Comparison of Logistic Regression, Classification and Regression Tree, and Neural Networks Models in Predicting Violent Re-Offending (2011, methodology): “When developing predictive models for use in criminal justice applications (and certain other fields), it is important to consider base rates in the development process and to avoid making predictions or classifications based on criteria that produce larger errors than would occur with the simple us”
• Forecasts of Violence to Inform Sentencing Decisions (2013, background): “Despite jurisprudential concerns and forecasting accuracy that has been difficult to evaluate (Farrington and Tarling 2003; Gottfredson and Moriarty 2006; Harcourt 2007; Berk 2008a, 2009, 2012), there is little doubt that these methods are here to stay.”