The Validity of the ODARA in Australian Intimate Partner Violence Cases Without Prior Assault/Credible Threat or Cohabitation

• Author: Troy E. Mcewan,Anne Sophie Pichler,James R. P. Ogloff,Melanie Simmons
• DOI: http://doi.org/10.1177/00938548221139854
• Published date: 01 May 2023
• Date: 01 May 2023
• Subject Matter: Articles

CRIMINAL JUSTICE AND BEHAVIOR, 2023, Vol. 50, No. 5, May 2023, 627 –647.

DOI: https://doi.org/10.1177/00938548221139854

Article reuse guidelines: sagepub.com/journals-permissions

© 2023 International Association for Correctional and Forensic Psychology

627

THE VALIDITY OF THE ODARA IN AUSTRALIAN

INTIMATE PARTNER VIOLENCE CASES

WITHOUT PRIOR ASSAULT/CREDIBLE

THREAT OR COHABITATION

TROY E. MCEWAN

ANNE SOPHIE PICHLER

JAMES R. P. OGLOFF

MELANIE SIMMONS

Swinburne University of Technology and Forensicare

This study investigated whether the Ontario Domestic Assault Risk Assessment (ODARA) is effective in the assessment of

risk in male intimate partner violence (IPV) cases that do not meet the inclusion criteria used in the ODARA’s development

sample (presence of prior assault or threat by perpetrator and previous cohabitation with the victim). Australian police scored

the ODARA in 275 IPV cases without one or both of these characteristics, with results contrasted to performance in 200 cases

meeting both inclusion criteria. The ODARA demonstrated poor discriminant effect over time for both assault and abuse

recurrence in the former group (concordance index [c-index] = 0.56 and c-index = 0.57, respectively), but performed well

in the latter group (c-index = 0.69 and c-index = 0.69). Although subject to some significant methodological limitations and

requiring replication, these findings suggest the ODARA may not provide accurate risk-based classification if applied to cases

missing the inclusion criteria.

Keywords: intimate partner violence; intimate partner abuse; risk assessment; ODARA

Intimate partner abuse (IPA) refers to any actual or threatened physical or sexual violence,

emotional abuse, economic abuse, or stalking that occurs between current or former inti-

mate partners (Barnett et al., 2005). Over one third of American women report being vic-

tims of IPA (Smith et al., 2018), while in Australia in 2019, 30% of homicides, 33% of

sexual assaults, and at least 41% of assaults occurred in the context of intimate partner or

family abuse (Australian Bureau of Statistics, 2020). Given these figures, it is imperative

AUTHORS’ NOTE: The author acknowledge the contributions of Dr Ilana Lauria and Dr Julia Nazarewicz,

who assisted in collating data used in this research, and members of Victoria Police who facilitated the research

project. This research was supported by funding from the Macedon Ranges and North West Melbourne

Medicare Local. Correspondence concerning this article should be addressed to Troy E. McEwan, Centre for

Forensic Behavioural Science, Swinburne University of Technology, Level 1, 582 Heidelberg Road, Alphington

3078, Victoria, Australia; e-mail: tmcewan@swin.edu.au.

1139854CJBXXX10.1177/00938548221139854Criminal Justice and BehaviorMcEwan et al. / The Validity of the ODARA in Australian IPV Cases

research-article2022

628 CRIMINAL JUSTICE AND BEHAVIOR

that police respond effectively to IPA at the earliest opportunity, both because it is associ-

ated with substantial harm and because it is common and frequently persists over time, with

significant implications for criminal justice resourcing (Medina Ariza et al., 2016).

Responding effectively to IPA in ways that hold those using violence accountable and

prevent harm is a major challenge for police organizations. Risk assessment instruments are

increasingly being used to help police identify families at heightened risk and take proactive

steps to prevent further harm in these cases (Medina Ariza et al., 2016; Messing et al., 2020;

Spivak et al., 2020; Radatz & Hilton, 2022; Williams, 2012). Proactive police risk manage-

ment can range from removing access to licensed weapons, initiating civil protective orders,

assisting victims with safety planning and modifications to their home environment, mak-

ing referrals to shelters or domestic violence agencies, social support agencies and/or spe-

cialist risk management panels, and ongoing contact with and monitoring of the aggressor

(Belfrage & Strand, 2012; Medina Ariza et al., 2016; Storey et al., 2014).

A common approach to police risk assessment of IPA cases is to use actuarial, or statisti-

cally derived, risk assessment tools (Kebbell, 2019). This form of risk assessment has been

shown to be somewhat more accurate and reliable than other risk assessment approaches

when examining IPA risks (Hilton et al., 2010; van der Put et al., 2019). An accurate actu-

arial tool can be used to classify cases for further assessment and proactive risk manage-

ment based on their identified probability of further violence over a given period of time

(Hilton et al., 2010; Kebbell, 2019; McEwan et al., 2019). Such tools do not directly inform

what kinds of risk management strategies should be used, only that increased management

is warranted. Some authors have suggested that one outcome from an actuarial instrument

should be further comprehensive assessment by specialists to identify where risk manage-

ment would be best targeted in a given case (Kebbell, 2019; Spivak et al., 2020).

The Ontario Domestic Assault Risk Assessment (ODARA) is an actuarial tool that accu-

rately classifies intimate partner violence (IPV) cases by their risk of future physical assault

(Hilton et al., 2010, 2021). The ODARA was developed for Canadian police to assess the

risk of future police-reported domestic assaults by adult men who had come to police atten-

tion and had a history of credible threat of death or assault against a female marital, com-

mon law or cohabiting partner (Hilton et al., 2010, 2004). It consists of 13 dichotomous

items and is quick and easy to score using information that is readily available to police. The

ODARA has the strongest evidence base among risk assessment tools predicting physical

IPA recidivism. Studies from Canada, Australia, and Europe have consistently found a mod-

erate to large ability to distinguish between those with and without further intimate partner

assaults using the ODARA score (e.g., Gerth et al., 2017; Hilton & Eke, 2016; Hilton &

Harris, 2009; Hilton et al., 2004; Lauria et al., 2017; Rettenberger & Eher, 2013). This was

confirmed in a recent meta-analysis of nine ODARA studies collectively involving 3,319

cases (van der Put et al., 2019). In this sample, there was a 69% probability that a recidivist

would obtain a higher ODARA score than a nonrecidivist, among the strongest effects

obtained for all tools evaluated in the study.

The ODARA is clearly an effective IPV risk assessment tool when used to predict the

outcome that the authors originally intended (i.e., physical assault of a women by her male

intimate partner), in samples similar to the original Canadian development sample (which

included only those with history of cohabitation due to characteristics of the development

dataset; Hilton et al., 2004). This is the intended scope of use as recommended by the author

of the instrument (Hilton, 2021). However, since the ODARA was developed, there have