The Predictive Validity of the Structured Assessment of Violence Risk in Youth in Secondary Educational Settings

Mark R. McGowan, Robert A. Horn, Ramona N. Mellott

Research output: Contribution to journalArticle

14 Scopus citations

Abstract

Current developments in violence risk assessment warrant consideration for use within educational settings. Using a structured professional judgment (SPJ) model, the present study investigated the predictive validity of the Structured Assessment of Violence in Youth (SAVRY) within educational settings. The predictive accuracy of the SAVRY scales was assessed using a retrospective file review to gather data on 87 adolescents ranging in age from 12 to 18 years. Receiver-operating characteristic analyses were used to gauge the predictive accuracy. With an area under the curve of .72 (p = .001), the accuracy of the SAVRY total score in correctly identifying violent youth exceeds the accuracy of identifications based on chance predictions in this sample. Logistic regression analyses assessed the relative contribution of the SAVRY subscales, whereas the omnibus equation using all subscale scores correctly classified 82% of those adolescents who were nonviolent and 45% of those adolescents who were violent. These results build on previous research and provide support for the use of the SAVRY in educational settings for identification as well as directing intervention efforts. Practical implications and areas for future research are also discussed.

Original languageEnglish (US)
Pages (from-to)478-486
Number of pages9
JournalPsychological Assessment
Volume23
Issue number2
DOIs
StatePublished - Jun 1 2011

Keywords

  • Adolescence
  • Predictive validity
  • SAVRY
  • School violence
  • Violence risk assessment

ASJC Scopus subject areas

  • Clinical Psychology
  • Psychiatry and Mental health

Fingerprint Dive into the research topics of 'The Predictive Validity of the Structured Assessment of Violence Risk in Youth in Secondary Educational Settings'. Together they form a unique fingerprint.

  • Cite this