Do You Believe Your Partner Is Capable of Killing You?: An Examination of Factors That IPV Survivors Associate with Fatality Risk

Background: Within the field of intimate partner violence (IPV), there has been debate as to whether survivors of IPV are best equipped to assess their risk for future violence. Studies examining the predictive validity of survivors’ appraisal of their own risk have been positive, suggesting that survivors are often able to make accurate predictions of their risk for future violence. Less is known about which specific factors survivors are most likely to associate with fatality risk. As such, the purpose of this study was to examine risk factors that survivors of IPV associate with fatality, and more specifically, whether their abuser is capable of killing them.

Methods: Data utilized in this analysis came from a larger study in which a risk assessment tool was developed for use with survivors seeking waivers to welfare requirements as part of the Family Violence Option. The risk assessment tool included questions about survivors’ experiences with IPV, risk of future abuse, perceptions of safety, and emotional health. A total of 237 risk assessments were completed by risk assessors during the pilot between June and December 2016. Classification and regression tree (CART) analysis was utilized to determine which factors survivors of IPV associated with fatality risk. Logistic regression analysis was used to confirm CART findings.

Results: Overall, 78% of survivors believed their abuser is capable of killing them. CART results showed that these beliefs were significantly impacted by five risk factors: a) abuser has access to a gun, b) abuser has threatened to kill survivor previously, c) moderate or high levels of emotional abuse risk, d) abuser previously threatened to commit suicide, and e) abuser used or threatened to use a weapon against the survivor. Logistic regression analysis confirmed the CART findings. Survivors were more likely to believe that their partner is capable of killing them if their partner has access to a gun (b=2.27, SE=.76, Wald=8.84, p=.01), had previously threatened to kill them (b=1.16, SE=.41, Wald=8.15, p=.01), and if the survivor had moderate emotional abuse risk levels as compared to low levels (b=-.68, SE=.70, Wald=5.66, p=.05).

Conclusion: Findings from this study illustrate that survivors of IPV associated some risk factors with fatality more than others. Interestingly, survivors did not associate certain high risk behaviors, such as stalking and strangulation, with whether their abuser was capable of killing them. The presence of emotional abuse in the model, as opposed to physical or sexual abuse, is particularly important for guiding organizational policy and practices as it highlights the insidious, but less visible, effects of non-violent forms of abuse. In order to be effective practitioners, social workers providing services to survivors need to be cognizant of the relationship between emotional abuse and women’s fear of lethality and ensure that women experiencing non-violent abuse are not overlooked in a system designed to primarily address overt aggression from abusers.