Abstract
Objectives
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.
Methods
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.
Results
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.
Conclusions
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.
Notes
We use the term “general” offender recidivism assessments to draw a distinction between those used for a correctional offender population and those used for specific populations, namely: sex offenders, psychopaths, and the mentally ill.
Overfitting is a term used to indicate that a model is trained too closely to the development (construction) sample and loses predictive accuracy when applied to additional (validation) samples.
It should be noted that with 100 bootstrap draws, the 95 % CI for each performance measure is calculated as (m-1.96*SD/ √(100), m + 1.96*SD/ √(100)). For presentation purposes, CIs are not included with model results but may be obtained by contacting the corresponding author.
References
Andrews, D. A. (1982). The level of supervision inventory (LSI). Ontario Ministry of Correctional Services.
Andrews, D. A. (1995). No title. Cited in LS/CMI Manual pp. 117–144.
Andrews, D. A., & Bonta, J. (1995). The LSI-R: the level of service inventory–revised. Toronto: Multi-Health Systems.
Andrews, D., Bonta, J., & Hoge, R. (1990a). Classification for effective rehabilitation: rediscovering psychology. Criminal Justice and Behavior, 17(1), 19–52.
Andrews, D., Zinger, I., Hodge, R., Bonta, J., Gendreau, P., & Cullen, F. (1990b). Does correctional treatment work? a clinically relevant and psychologically informed meta-analysis. Criminology, 28(3), 369–392.
Andrews, D., Bonta, J., & Wormith, S. (2006). The recent past and near future of risk/need assessment. Crime and Delinquency, 52(1), 7–27.
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.
Baird, C. S. (1981). Probation and parole classification: the Wisconsin model. Corrections Today, 43, 36–41.
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.
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].
Barnoski, R., Aos, S. (2003). Washington’s Offender Accountability Act: An Analysis of the Department of Corrections’ Risk Assessment. #03-12-1202.
Barnoski, R., Drake, E. (2007). Washington’s Offender Accountability Act: Department of Corrections’ Static Risk Instrument. #07-03-1201.
Berk, R., Sherman, L., Barnes, G., Kurtz, E., & Ahlman, L. (2009). Forecasting murder within a population of probationers and parolees: a high stakes application of statistical learning. Journal of the Royal Statistical Society A, 172(Part 1), 191–211.
Bishop, C. M. (1995). Neural networks for pattern recognition. Oxford: Oxford University Press.
Breiman, L. (1996). Bagging predictors. Machine Learning Journal, 26, 123–140.
Breiman, L. (2001a). Statistical modeling: the two cultures. Statistical Science, 16(3), 199–215.
Breiman, L. (2001b). Random forests. Machine Learning Journal, 45(1), 5–32.
Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (1984). Classification and regression trees. Monterey: Wadsworth and Brooks/Cole.
Brennan, T., & Oliver, W. L. (2000). Evaluation of reliability and validity of COMPAS scales: national aggregate sample. Traverse City: Northpointe Institute for Public Management.
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
Brennan, T., Dieterich, W., & Ehret, B. (2009b). Evaluation the predictive validity of the comps risk and needs assessment system. Criminal Justice and Behavior, 36(1), 21–40.
Brodzinski, J. D., Crable, E. A., & Scherer, R. F. (1994). Using artificial intelligence to model juvenile recidivism patterns. Computers in Human Services, 10(4), 1–18.
Caulkins, J., Cohen, J., Gorr, W., & Wei, J. (1996). Predicting criminal recidivism: a comparison of neural networks with statistical models. Journal of Criminal Justice, 24(3), 227–240.
Cottle, C. C., Lee, R. J., & Heilbrun, K. (2001). The prediction of criminal recidivism in juveniles: a meta-analysis. Criminal Justice and Behavior, 28(3), 367–394.
Duwe, G. (2013). The development, validity, and reliability of the Minnesota screening tool assessing recidivism risk (MnSTARR). Criminal Justice Policy Review, XX, 1–35. doi:10.1177/0887403413478821.
Gardner, W., Lidz, C., Mulvey, E., & Shaw, E. (1996). A comparison of actuarial methods of identifying repetitively violent patients with mental illness. Law and Human Behavior, 20(1), 35–48.
Gottfredson, S., & Moriarty, L. (2006). Statistical risk assessment: Old problems and New applications. Crime and Delinquency, 52(1), 178–200.
Grann, M., & Langstrom, N. (2007). Actuarial assessment of violent risk: to weigh or Not to weigh. Criminal Justice and Behavior, 34(1), 22–36.
Hare, R. (1991). The revised psychopathy checklist. Toronto: Multi-Health Systems.
Harper, P. R. (2005). A review and comparison of classification algorithms for medical decision making. Health Policy, 71(3), 315–331.
Harrell, F., Lee, K., & Mark, D. (1996). Multivariate prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Statistics in Medicine, 15, 361–387.
Haykin, S. (1999). Neural networks: a comprehensive foundation. Upper Saddle River: Prentice Hall.
Hertz, J., Palmer, R. G., & Krogh, A. S. (1990). Introduction to the theory of neural computation. New York: Perseus Books.
Jung, S., & Rawana, E. P. (1999). Risk-need assessment of juvenile offenders. Criminal Justice and Behavior, 26, 69–89.
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.
Liu, Y. Y., Yang, M., Ramsay, M., Li, X. S., & Coid, J. W. (2011). A comparison of logistic regression, classification and regression tree, and neural networks models in predicting violent Re-offending. Journal of Quantitative Criminology, 27(4), 547–573.
Loeber, R., & Farrington, D. (1998). Serious and violent juvenile offenders: risk factors and successful interventions. Thousand Oaks: Sage.
Monahan, J., Steadman, H. J., Appelbaum, P. S., Robbins, P. C., Mulvey, E. P., Silver, E., Roth, L. H., & Grisso, T. (2000). Developing a clinically useful actuarial tool for assessing violence risk. British Journal of Psychiatry, 176(4), 312–319.
Monahan, J., Steadman, H. J., Robbins, P. C., Appelbaum, P., Banks, S., & Grisso, T. (2005). An actuarial model of violence risk assessment for persons with mental disorders. Psychiatric Services, 56(7), 810–815.
Monahan, J., Steadman, H. J., Appelbaum, P. S., Grisso, T., Mulvey, E. P., & Roth, L. H. (2006). The classification of violence risk. Behavioral Sciences and the Law, 24(6), 721–730.
Neuilly, M.-A., Zgoba, K. M., Tita, G. E., & Lee, S. (2011). Predicting recidivism in homicide offenders using classification tree analysis. Homicide Studies, 15(2), 154–176.
Palocsay, S. W., Wang, P., & Brookshire, R. G. (2000). Predicting criminal recidivism using neural networks. Socio-Economic Planning Sciences, 34(4), 271–284.
Ripley, B. D. (1996). Pattern recognition and neural networks. New York: Cambridge University Press.
Rosenfeld, B., & Lewis, C. (2005). Assessing violence risk in stalking cases: a regression tree approach. Law and Human Behavior, 29(3), 343–357.
Schaffer, D., Kelly, B., & Lieberman, J. (2011). An exemplar-based approach to risk assessment: validating the risk management systems instrument. Criminal Justice Policy Review, 22(2), 167–186.
Silver, E., & Chow-Martin, L. (2002). A multiple-models approach to assessing recidivism risk: implications for judicial decision making. Criminal Justice and Behavior, 29(5), 538–568.
Silver, E., Smith, W. R., & Banks, S. (2000). Constructing actuarial devices for predicting recidivism: a comparison of methods. Criminal Justice and Behavior, 27(6), 733–764.
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.
Smith, M. (1993). Neural networks for statistical modeling. New York: Van Nostrand Reinhold.
Smith, P., Cullen, F., & Latessa, E. (2009). Can 14,737 women be wrong? a meta-analysis of the LSI-R and recidivism for female offenders. Criminology and Public Policy, 8(1), 183–208.
Stalans, L. J., Yarnold, P. R., Seng, M., Olson, D. E., & Repp, M. (2004). Identifying three types of violent offenders and predicting violent recidivism while on probation: a classification tree analysis. Law and Human Behavior, 28(3), 253–262.
Steadman, H. J., Silver, E., Monahan, J., Appelbaum, P. S., Clark Robbins, P., Mulvey, E. P., Grisso, T., Roth, L. H., & Banks, S. (2000). A classification tree approach to the development of actuarial violence risk assessment tools. Law and Human Behavior, 24(1), 83–100.
Thomas, S., Leese, M., Walsh, E., McCrone, P., Moran, P., Burns, T., Creed, F., Tirer, P., & Fahy, T. (2005). A comparison of statistical models in predicting violence in psychotic illness. Comprehensive Psychiatry, 46, 296–303.
Tollenaar, N., & van der Heijden, P. G. M. (2013). Which method predicts recidivism best?: a comparison of statistical, machine learning, and data mining predictive models. Journal of the Royal Statistical Society: Series A (Statistics in Society), 176(2), 565–584.
Van Voorhis, P., Wright, E. M., Salisbury, E., & Bauman, A. (2010). Women’s risk factors and their contributions to the existing risk/needs assessment: the current status of a gender-responsive supplement. Criminal Justice and Behavior, 37, 261–288.
Wasserman, P. (1993). Advanced methods in neural computing. New York: Van Nostrand Reinhold.
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.
Author information
Authors and Affiliations
Corresponding author
Appendix I
Appendix I
Rights and permissions
About this article
Cite this article
Hamilton, Z., Neuilly, MA., Lee, S. et al. Isolating modeling effects in offender risk assessment. J Exp Criminol 11, 299–318 (2015). https://doi.org/10.1007/s11292-014-9221-8
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11292-014-9221-8