Intimate Partner Violence Risk Assessment Validation Study, Final Report

• NCJ Number: 209731
• Author(s): Jacquelyn C. Campbell Ph.D.; Chris O'Sullivan Ph.D.; Janice Roehl Ph.D.; Daniel Webster Sc.D.
• Date Published: March 2005
• Length: 92 pages
• Grant Number(s): 2000-WT-VX-0011
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research)
• Award: Risk Assessment Validation Study

Annotation

The major goal of this multisite field test was to assess the predictive accuracy of two instruments--the Domestic Violence Screening Instrument (DVSI) and the Kingston Screening Instrument for Domestic Violence (K-SID)--that assess the risk of repeat violence in intimate relationships, as well as one instrument (the Danger Assessment) and one Threat Assessment Method (DV-MOSAIC) that estimate lethal or near-lethal risk in violent intimate partner relationships.

Abstract

Four New York City sites and two Los Angeles sites were selected for testing the instruments. A total of 1,307 domestic-violence victims were involved in baseline interviews. Follow-up interviews were begun in December 2002 and ended in early January 2004. Although follow-up phone contacts began 6 months after baseline interviews, the length of the follow-up periods varied among the sample for various reasons. Retention rates varied from site to site, ranging from a low of 33 percent to a high of 69 percent. Overall, of the 1,307 enrollees at baseline, follow-up interviews were completed with 782 (60 percent). At Time 1, 82 percent of the women had experienced severe abuse, with all but 6 percent having been physically assaulted by their partner or ex-partner. Approximately one-third had been re-assaulted by the end of the follow-up period (a maximum of 2 years). Repeat assaults continued to be severe, with 11 percent having experienced a potentially lethal attack. None of the instruments or method was impressive in predicting reassault. By most analytical strategies, the Danger Assessment had the strongest psychometric properties, including the predictive statistics. The DVSI and DV-MOSAIC also had significant associations with re-assaults. The K-SID was the weakest of the instruments; however, it did well in predicting re-arrests with the use of criminal justice data. Women's perception of risk did better than the other assessment methods. Extensive tables and 151 references