Practical Implications of Current Domestic Violence Research: For Law Enforcement, Prosecutors and Judges

Published June 2009

Chapter 3. Offender Characteristics

Section 15 — Are victims accurate predictors of reabuse?

Victim perception of risk has been found to significantly improve the accuracy of prediction over other risk factors [44], increasing sensitivity — the proportion of true positives that are correctly identified by the test — from 55 to 70 percent. [112] However, the same researchers found that women's perceptions have to be interpreted. Women who felt very safe were less likely to be repeatedly reassaulted than those that felt somewhat safe. However, women who were uncertain or felt somewhat unsafe were more likely to be reassaulted repeatedly than those who felt they were in great danger. The reason for this apparent contradiction is that women who felt in greatest danger took effective countermeasures during the study. In other words, the research suggests that if women are not certain they will be safe, they err by giving the benefit of the doubt to their abuser. For these reasons, these researchers concluded that the best predictions of repeated reassaults were obtained by using risk markers, including women's perceptions. [44, 112] The researchers' concern for victims with regard to assessed risk of abuse is borne out by a study of more than 1,000 women who sought protective orders or shelter, or whose abusers were arrested in Los Angeles or New York City. Almost a quarter of the victims who thought their risk of reassault was low were, in fact, reassaulted within one year. [190]

Victims' perception of risk also affects their reaction to criminal justice intervention. Arrest research finds that victims who were not revictimized for more than two years were twice as likely to have opposed arrest, compared to those who were revictimized. Those victims who thought police and court intervention did not go far enough were also accurate. Those who said police actions were too weak were three times more likely to experience revictimization, and those victims who said courts failed them were seven times more likely to experience revictimization. [23]

Implications for Law Enforcement

Asking victims if they fear reassault or severe reassaults provides one of the best ways to predict reabuse or potential lethality — and requires the least resources and time commitment — but cannot be relied on exclusively as a predictor. Although women are unlikely to exaggerate their risk, they often underestimate it. (Research basis: A national homicide study involving hundreds of victims of attempted homicides, as well as the general reabuse studies, confirms these findings.)

Implications for Prosecutors and Judges

Victim input should be an important part of any risk calculation considered by prosecutors and judges. If victims are in doubt as to their safety, prosecutors and judges should assume the worst. (Research basis: Extensive examination of multiple domestic violence risk studies shows agreement on this point.)

Date Created: June 5, 2009