Isolating modeling effects in offender risk assessment
- 471 Downloads
Recent evolutions in actuarial research have revealed the potential increased utility of machine learning and data-mining strategies to develop statistical models such as classification/decision-tree analysis and neural networks, which are said to mimic the decision-making of practitioners. The current article compares such actuarial modeling methods with a traditional logistic regression risk-assessment development approach.
Utilizing a large purposive sample of Washington State offenders (N = 297,600), the current study examines and compares the predictive validity of the currently used Washington State Static Risk Assessment (SRA) instrument to classification tree analysis/random forest and neural network models.
Overall findings varied, being dependent on the outcome of interest, with the best model for each method resulting in AUCs ranging from 0.732 to 0.762. Findings reveal some predictive performance improvements with advanced machine-learning methodologies, yet the logistic regression models demonstrate comparable predictive performance.
The study concluded that while data-mining techniques hold potential for improvements over traditional methods, regression-based models demonstrate comparable, and often improved, prediction performance with noted parsimony and greater interpretability.
KeywordsRandom forest Neural network Recidivism Risk assessment
- Andrews, D. A. (1982). The level of supervision inventory (LSI). Ontario Ministry of Correctional Services.Google Scholar
- Andrews, D. A. (1995). No title. Cited in LS/CMI Manual pp. 117–144.Google Scholar
- Andrews, D. A., & Bonta, J. (1995). The LSI-R: the level of service inventory–revised. Toronto: Multi-Health Systems.Google Scholar
- Austin, J., Coleman, D., Peyton, J., & Johnson, K.D. (2003). Reliability and validity study of the LSI-R risk assessment instrument. Washington, D.C. Institute on Crime, Justice, and Corrections at The George Washington University.Google Scholar
- Baird, C. S. (1981). Probation and parole classification: the Wisconsin model. Corrections Today, 43, 36–41.Google Scholar
- Banks, S., Robbins, P. C., Silver, E., Vesselinov, R., Steadman, H. J., Monahan, J., Mulvey, E. P., Appelbaum, P. S., Grisso, T., & Roth, L. H. (2004). A multiple-models approach to violence risk assessment among people with mental disorder. Criminal Justice and Behavior, 31(3), 324–340.CrossRefGoogle Scholar
- Barnoski, R. (2010) Washington State static risk assessment—version 2.0. [Modified to improve reliability and validity, requested by Washington State Center for Court Research].Google Scholar
- Barnoski, R., Aos, S. (2003). Washington’s Offender Accountability Act: An Analysis of the Department of Corrections’ Risk Assessment. #03-12-1202.Google Scholar
- Barnoski, R., Drake, E. (2007). Washington’s Offender Accountability Act: Department of Corrections’ Static Risk Instrument. #07-03-1201.Google Scholar
- Bishop, C. M. (1995). Neural networks for pattern recognition. Oxford: Oxford University Press.Google Scholar
- Breiman, L. (1996). Bagging predictors. Machine Learning Journal, 26, 123–140.Google Scholar
- Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (1984). Classification and regression trees. Monterey: Wadsworth and Brooks/Cole.Google Scholar
- Brennan, T., & Oliver, W. L. (2000). Evaluation of reliability and validity of COMPAS scales: national aggregate sample. Traverse City: Northpointe Institute for Public Management.Google Scholar
- Brennan, T., Dieterich, B., Breitenbach, M., & Mattson, B. (2009a). Commentary on NCCD “A questions of evidence: A critique of risk assessment models used in the justice system”. Northpointe Institute for Public Management, Inc. http://www.northpointeinc.com/files/whitepapers/Baird_Response_060409.pdf
- Hare, R. (1991). The revised psychopathy checklist. Toronto: Multi-Health Systems.Google Scholar
- Haykin, S. (1999). Neural networks: a comprehensive foundation. Upper Saddle River: Prentice Hall.Google Scholar
- Hertz, J., Palmer, R. G., & Krogh, A. S. (1990). Introduction to the theory of neural computation. New York: Perseus Books.Google Scholar
- Latessa, E., Smith, P., Lemke, R., Markarios, M., Lowenkamp, C. (2009) Creation and Validation of the Ohio Risk Assessment System – Final Report. Ohio Department of Rehabilitation and Correction. # 2005-JG-E0R-6269 and 2005-JG-C01-T8.Google Scholar
- Loeber, R., & Farrington, D. (1998). Serious and violent juvenile offenders: risk factors and successful interventions. Thousand Oaks: Sage.Google Scholar
- Skeem, J., & Louden, J. (2007). Assessment of Evidence on the quality of the correctional offender management profiling for alternative sanctions (COMPAS). Prepared for the California department of corrections and rehabilitation (CDCR). CA: Davis.Google Scholar
- Smith, M. (1993). Neural networks for statistical modeling. New York: Van Nostrand Reinhold.Google Scholar
- Wasserman, P. (1993). Advanced methods in neural computing. New York: Van Nostrand Reinhold.Google Scholar
- Wasserman, L. (2014). Rise of the machines. In X. Lin, D. L. Banks, C. Genest, G. Molenberghs, D. W. Scott, & J.-L. Wang (Eds.), Past, present, and future of statistical science (pp. 1–12). Boca Raton: CRC Press. Chapter 1.Google Scholar