Application of Bayes’ theorem in assessing recidivism risk in a serial rapist: a case study
Autor/i: Valentina Baić, Milan Oljača, Marija Tasić
Stranice: 31-42
UDK: 343.95:343.541
343.238
DOI: https://doi.org/10.47152/rkkp.64.1.2
Apstrakt: The assessment of recidivism risk among sexual offenders represents one of the most challenging issues in criminological and forensic practice, particularly in cases involving multiple repeat offenders, where available data are often fragmentary and the degree of uncertainty is high. The aim of this paper is to demonstrate the application of Bayes’ theorem in the analysis of criminal behaviour and in the formalised probabilistic assessment of the risk of rape reoffending through a case study of a serial rapist sentenced to a long-term term of imprisonment. The study is grounded in a Bayesian inferential framework, in which the recidivism hypothesis is operationalised as a binary outcome: the reoccurrence of rape within a three-year period following release (2029–2032). The individualised baseline probability (empirical prior) was estimated using a beta–binomial Bayesian model based on the offender’s criminal history, due to the absence of domestic longitudinal recidivism rates for the specified time horizon. The evidence used for Bayesian updating comprised operationalised risk indicators, including repeated rape following previous releases, escalation of violence, the presence of antisocial personality disorder, impulsivity, and stable modus operandi patterns. Additional empirical data from a domestic study on rape recidivism were used to contextualise the findings. The results of the Bayesian analysis indicate that the posterior probability of recidivism within the three-year period remains exceptionally high (≈0.97–0.99), even under conservative assumptions regarding evidential strength. The findings suggest that the potential protective effects of age and long-term imprisonment, even when considered under a conservative scenario, do not outweigh the strong pattern of prior behaviour and the clinical-behavioural risk factors. In conclusion, the paper demonstrates that Bayes’ theorem constitutes a methodologically transparent and practically applicable framework for individualised probabilistic reasoning in forensic risk assessment.
Ključne reči: Bayes’ theorem, rape recidivism, risk assessment, serial rapist, case study, forensic reasoning
Reference:
Andrews, D. A. and Bonta, J. (2010) The psychology of criminal conduct (5th ed.). New Providence, NJ: LexisNexis Matthew Bender. https://doi.org/10.4324/9781315721279
Andrews, D. A. and Bonta, J. (2017) The psychology of criminal conduct (6th ed.). New York: Routledge.
Baić, V. and Lajić, O. (2017) 'Analysis of the criminal behavior of multiple perpetrators of the crime of rape', NBP – Journal of Criminalistics and Law, 22(1), 33-51. https://doi.org/10.5937/nabepo22-13316
Baić, V. et al. (2026) 'Recidivism of rapists: A retrospective analysis of static and dynamic risk factors', Teme (in press).
Donaldson, T. and Wollert, R. (2008) 'A mathematical proof and example that Bayes’s theorem is fundamental to actuarial estimates of sexual recidivism risk', Sexual Abuse: A Journal of Research and Treatment, 20(2), 204-226. https://doi.org/10.1177/1079063208317734
Douglas, K. et al. (2013) HCR-20V3: Assessing risk for violence – User guide. Burnaby, BC: Mental Health, Law and Policy Institute, Simon Fraser University.
Gelman, A. et al. (2014) Bayesian Data Analysis (3rd ed.). Boca Raton, FL: Chapman & Hall/CRC.
Hanson, R. K. and Morton-Bourgon, K. E. (2005) 'The characteristics of persistent sexual offenders: A meta-analysis of recidivism studies', Journal of Consulting and Clinical Psychology, 73(6), 1154-1163. https://doi.org/10.1037/0022-006X.73.6.1154
Hanson, R. K. and Morton-Bourgon, K. E. (2009) 'The accuracy of recidivism risk assessments for sexual offenders: A meta-analysis of 118 prediction studies',
Psychological Assessment, 21(1), 1-21. https://doi.org/10.1037/a0014421
Jaynes, E. T. (2003) Probability theory: The logic of science. Cambridge University Press.
Lussier P., Corrado R. R. and McCuish E. (2016) 'A criminal career study of the continuity and discontinuity of sex offending during the adolescence-adulthood transition: A prospective longitudinal study of incarcerated youth', Justice Quarterly, 33(7), 1123-1153. https://doi.org/10.1080/07418825.2015.1028966
Lussier P. et al. (2012) 'A developmental taxonomy of juvenile sex offenders for theory, research, and prevention: The adolescent-limited and the high-rate slow desister', Criminal Justice and Behavior, 39(12), 1559-1581. https://doi.org/10.1177/0093854812455739
Mann, R. E., Hanson, R. K. and Thornton, D. (2010) 'Assessing risk for sexual recidivism: Some proposals on the nature of psychologically meaningful risk factors', Sexual Abuse: A Journal of Research and Treatment, 22(2), 191-217. https://doi.org/10.1177/1079063210366039
Mokros, A. et al. (2010) 'Assessment of risk for violent recidivism through multivariate Bayesian classification', Psychology, Public Policy, and Law, 16(4), 418-450. https://doi.org/10.1037/a0020939
Rossegger, A. et al. (2014) 'Replicating the Violence Risk Appraisal Guide: A total forensic cohort study', PLoS ONE, 9(3), e91845. https://doi.org/10.1371/journal.pone.0091845
Taroni, F. et al. (2014) 'Bayesian reasoning in forensic science', Forensic Science International, 235, 1-9. https://doi.org/10.1016/j.forsciint.2013.11.017
Wollert, R. (2006) 'Low base rates limit expert certainty when current actuarials are used to identify sexually violent predators: An application of Bayes’s theorem', Psychology, Public Policy, and Law, 12(1), 56-85. https://doi.org/10.1037/1076-8971.12.1.56