The Danger Assessment (DA) is a brief instrument designed to provide a clinical assessment of dangerousness in violent intimate relationships. Ten prior research investigations by the instrument author and others have established beginning support for reliability and convergent and discriminant group construct validity. Two small recent studies have also provided beginning support for predictive validity. This paper will give a brief overview of instrument development, the prior psychometric data on the DA, instruction on its clinical use and present the data from a 10 city national case control study of intimate partner femicide and attempted femicide supporting the validity of the items. Two hundred and fifty (250) homicide and 200 attempted homicide victims were the cases, and controls were 450 battered women from the same 10 cities. Results to be reported include internal consistency reliability for the Danger Assessment. The mean unweighted DA score for the Attempted Femicide victims was significantly different than the Abused women controls (6.5 and 2.8 respectively; p <.01). All of the factors on the DA were significantly different between cases and controls except for victim suicidality, which although not a risk factor for femicide is still potentially lethal and may be a risk factor for her killing him. Also reported will be the results of ultivariate regression models and examination of ROC analysis to determine appropriate weighting of the factors for more advanced scoring of the instrument. Implications for assessment for lethality risk in violent relationships will also be presented.
Learning Objectives: See session objectives
Keywords: Domestic Violence, Risk Assessment
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.