Stalking and violence: A 10-year follow-up of stalking offenders

Date created: 
Guidelines for Stalking Assessment and Management
Risk assessment
Predictive validity
Structured Professional Judgment
Interrater reliability

Stalking is a form of targeted violence which most often results in psychological harm, but can also include acts of physical harm, and is associated with victims experiencing a wide range of psychological difficulties. This study examined the psychometric properties of the Guidelines for Stalking Assessment and Management (SAM; Kropp, Hart, & Lyon, 2008), including the interrater reliability, predictive validity, and concurrent validity with the Screening Version of the Hare Psychopathy Checklist-Revised (PCL:SV; Hart, Cox, & Hare, 1995), and the Violence Risk Appraisal Guide (VRAG; Quinsey, Rice, Harris, & Cormier, 1998). Another objective of this study was to examine reoffending rates and patterns in time-to-reoffending among a sample of offenders originally convicted of stalking-related offences. The sample consisted of 106 offenders referred to a Canadian forensic psychiatric clinic or hospital for pre-trial or post-sentence assessment or treatment. Recidivism was coded over the follow-up period of approximately 10 years, and was categorized into four types of recidivism (i.e., any new stalking with physical harm, any new stalking, any new violence, & any new recidivism). Overall, the SAM performed well in terms of interrater reliability. The interrater reliability of numerical SAM total and domain scores was good to excellent (ICC2 range = .73 to .75) while Summary Risk rating agreement was fair (ICC2 range = .50 to .57). When examining the concurrent validity of the SAM, the Nature of Stalking, Perpetrator Risk Factor, and SAM Total numerical scores displayed good concurrent validity with the with the PCL:SV and VRAG, as did the Case Prioritization and Risk of Physical Harm ratings, correlations all significant p < .05. The Risk of Continued Stalking rating was significantly correlated with the PCL:SV, p < .05, but not the VRAG. As would be expected, Victim Vulnerability Factors and Reasonableness of Fear Summary Risk rating were not significantly correlated with either measure. In terms of predictive validity, none of the measures (SAM numerical total scores, VRAG scores, PCL:SV scores) were associated with stalking recidivism. The SAM Summary Risk rating, Risk for Continued Stalking, significantly predicted violent recidivism within the first year.

Document type: 
This thesis may be printed or downloaded for non-commercial research and scholarly purposes. Copyright remains with the author.
Stephen Hart
Arts & Social Sciences: Department of Psychology
Thesis type: 
(Thesis) Ph.D.