Predicting Sexual Recidivism

Turgut Ozkan, Stephen J. Clipper, Alex R. Piquero, Michael Baglivio, Kevin Wolff

Research output: Contribution to journalArticleResearchpeer-review

Abstract

The current study focuses on adolescents with sex offense histories and examines sexual reoffending patterns within 2 years of a prior sex offense. We employed inductive statistical models using archival official records maintained by the Florida Department of Juvenile Justice (FDJJ), which provides social, offense, placement, and risk assessment history data for all youth referred for delinquent behavior. The predictive accuracy of the random forest models is tested using receiver operator characteristic (ROC) curves, the area under the curve (AUC), and precision/recall plots. The strongest predictor of sexual recidivism was the number of prior felony and misdemeanor sex offenses. The AUC values range between 0.71 and 0.65, suggesting modest predictive accuracy of the models presented. These results parallel the existing literature on sexual recidivism and highlight the challenges associated with predicting sex offense recidivism. Furthermore, results inform risk assessment literature by testing various factors recorded by an official institution.

Original languageEnglish
Number of pages25
JournalSexual Abuse: Journal of Research and Treatment
DOIs
Publication statusAccepted/In press - 2019
Externally publishedYes

Keywords

  • juvenile sex offender recidivism
  • machine learning
  • random forests
  • sexual recidivism

Cite this

Ozkan, Turgut ; Clipper, Stephen J. ; Piquero, Alex R. ; Baglivio, Michael ; Wolff, Kevin. / Predicting Sexual Recidivism. In: Sexual Abuse: Journal of Research and Treatment. 2019.
@article{d1da2cc0a2e34930881fd0a9db763902,
title = "Predicting Sexual Recidivism",
abstract = "The current study focuses on adolescents with sex offense histories and examines sexual reoffending patterns within 2 years of a prior sex offense. We employed inductive statistical models using archival official records maintained by the Florida Department of Juvenile Justice (FDJJ), which provides social, offense, placement, and risk assessment history data for all youth referred for delinquent behavior. The predictive accuracy of the random forest models is tested using receiver operator characteristic (ROC) curves, the area under the curve (AUC), and precision/recall plots. The strongest predictor of sexual recidivism was the number of prior felony and misdemeanor sex offenses. The AUC values range between 0.71 and 0.65, suggesting modest predictive accuracy of the models presented. These results parallel the existing literature on sexual recidivism and highlight the challenges associated with predicting sex offense recidivism. Furthermore, results inform risk assessment literature by testing various factors recorded by an official institution.",
keywords = "juvenile sex offender recidivism, machine learning, random forests, sexual recidivism",
author = "Turgut Ozkan and Clipper, {Stephen J.} and Piquero, {Alex R.} and Michael Baglivio and Kevin Wolff",
year = "2019",
doi = "10.1177/1079063219852944",
language = "English",
journal = "Sexual Abuse-A Journal of Research and Treatment",
issn = "1079-0632",
publisher = "Springer-Verlag London Ltd.",

}

Predicting Sexual Recidivism. / Ozkan, Turgut; Clipper, Stephen J.; Piquero, Alex R.; Baglivio, Michael; Wolff, Kevin.

In: Sexual Abuse: Journal of Research and Treatment, 2019.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Predicting Sexual Recidivism

AU - Ozkan, Turgut

AU - Clipper, Stephen J.

AU - Piquero, Alex R.

AU - Baglivio, Michael

AU - Wolff, Kevin

PY - 2019

Y1 - 2019

N2 - The current study focuses on adolescents with sex offense histories and examines sexual reoffending patterns within 2 years of a prior sex offense. We employed inductive statistical models using archival official records maintained by the Florida Department of Juvenile Justice (FDJJ), which provides social, offense, placement, and risk assessment history data for all youth referred for delinquent behavior. The predictive accuracy of the random forest models is tested using receiver operator characteristic (ROC) curves, the area under the curve (AUC), and precision/recall plots. The strongest predictor of sexual recidivism was the number of prior felony and misdemeanor sex offenses. The AUC values range between 0.71 and 0.65, suggesting modest predictive accuracy of the models presented. These results parallel the existing literature on sexual recidivism and highlight the challenges associated with predicting sex offense recidivism. Furthermore, results inform risk assessment literature by testing various factors recorded by an official institution.

AB - The current study focuses on adolescents with sex offense histories and examines sexual reoffending patterns within 2 years of a prior sex offense. We employed inductive statistical models using archival official records maintained by the Florida Department of Juvenile Justice (FDJJ), which provides social, offense, placement, and risk assessment history data for all youth referred for delinquent behavior. The predictive accuracy of the random forest models is tested using receiver operator characteristic (ROC) curves, the area under the curve (AUC), and precision/recall plots. The strongest predictor of sexual recidivism was the number of prior felony and misdemeanor sex offenses. The AUC values range between 0.71 and 0.65, suggesting modest predictive accuracy of the models presented. These results parallel the existing literature on sexual recidivism and highlight the challenges associated with predicting sex offense recidivism. Furthermore, results inform risk assessment literature by testing various factors recorded by an official institution.

KW - juvenile sex offender recidivism

KW - machine learning

KW - random forests

KW - sexual recidivism

UR - http://www.scopus.com/inward/record.url?scp=85067873116&partnerID=8YFLogxK

U2 - 10.1177/1079063219852944

DO - 10.1177/1079063219852944

M3 - Article

JO - Sexual Abuse-A Journal of Research and Treatment

JF - Sexual Abuse-A Journal of Research and Treatment

SN - 1079-0632

ER -