By Saoirse ’20

---

Abstract

Professionals within the criminal justice system will likely encounter victims and/or perpetrators of Intimate Partner Violence (IPV) within the course of their professional duties, given the prevalence of IPV. It is important that these professionals take a proactive and not a reactive approach to IPV and play an active role in preventing further instances of the crime. This report is designed to introduce professionals working within the criminal justice system to the two risk assessment instruments with the highest average predictive validity, the ODARA and SARA (Messing and Thaler, 2013) to assist in achieving accuracy in violence risk assessment and support decision making based on said assessment.

Keywords:  Intimate Partner Violence, Violence Risk Assessment, Ontario Domestic Violence Risk Assessment, Spousal Assault Risk Assessment, Criminal Justice System.

Recommendations for Risk Assessments of Intimate Partner Violence

Introduction

Violence against women has been identified as not only a human rights issue but a public health concern. Intimate Partner Violence or IPV was believed to be a relatively minor form of such violence presumably due to its nature as a private form of violence. However, that view changed in 2005 when the World Health Organisation (WHO) released a multi-country study on women’s health and domestic violence revealing that IPV is one of the most common forms of violence against women and that it occurs in all settings and among all socioeconomic, religious and cultural groups (WHO, 2005). The WHO defined IPV as physical violence, sexual violence, emotional abuse or controlling behaviours from an intimate partner (WHO, 2005). With a total sample of approximately 24,000 women from 10 participating countries, the prevalence of IPV in one of these forms ranged from 15 to 71% (WHO, 2015). Recent data has indicated that 6.5% of women and 2.8% of men in England and Wales, the equivalent of approximately 1.3 million women and 600,000 men have reported a specific type of partner abuse in the year 2014-15 (Office of National Statistics, 2016). Thus, it has become important for the criminal justice system to take a proactive, as opposed to a reactive stance towards this crime. The form this proactive approach takes is violence risk assessment and management. 

Over the past few years, there has been a proliferation of specific risk assessment tools that assist the criminal justice system in preventing first instances of IPV or reducing recdivism rates. Some of the most prominent risk assessment instruments include the Ontario Domestic Assault Risk Assessment or ODARA, the Spousal Assault Risk Assessment or SARA, The Brief Spousal Assault Form for the Evaluation of Risk or B-SAFER which is derived from SARA, the Danger Assessment or DA and the Domestic Violence Screening Inventory or DVSI. Although violence risk assessment is a widely-used global phenomenon (Singh, et al., 2014), the validity of the different IPV risk assessment instruments in various law enforcement contexts has largely not been examined. This report provides an understanding of the development, scoring system, validity and applicability of the two risk assessment instruments with the highest average predictive validity, the ODARA and SARA (Messing and Thaler, 2013) to assist in achieving accuracy in violence risk assessment and support decision making based on said assessment by various members of the criminal justice system based on their roles and the circumstances of the case.

Ontario Domestic Assault Risk Assessment (ODARA)

The Ontario Domestic Assault Risk Assessment (ODARA) is an actuarial risk assessment tool designed specifically to predict recidivism in male domestic violence offenders against female partners. It is designed for use by front-line police officers. It is constructed using follow up research and consists of 13 dichotomous items (Hilton, Harris, Popham, Lang, 2010).

Development

The ODARA was created as a follow-up to research involving 589 men who had committed assault against a female partner or ex-partner. The police files of the male perpetrators were statistically analysed to identify factors that acted as predictors of repeated domestic violence. The ODARA only included items that independently strongly predicted recidivism of IPV and excluded common characteristics of offenders that did not predict recidivism (Hilton, 2010).

Scoring

The ODARA calculates the likelihood that a man who has assaulted his female partner in the past will assault a female partner again in the future. It predicts violence in current and former marriages, common-law relations and cohabitation (Hilton, 2010). It relies on police, probation and criminal record files.

The ODARA score is based on 13 items which include prior domestic assault in police records, prior nondomestic assault in police records, prior sentence for a term of 30 days or more, failure on prior conditional release, threat to harm or kill anyone during index incident, confinement of victim during index incident, victim’s fear of future assault, more than one child with the victim, victim having a biological child with a previous partner, more than one indicator of a substance abuse problem, assault on pregnant victim and the victim’s barriers to support. Each item is scored either 0 or 1based on the absence or presence of that item as a risk factor and the final score is the sum of all items (Hilton, 2010). 

Based on the final scores, the offenders are placed in one of seven categories of risk. Placement in the highest categories indicates a higher risk of both frequency and severity of future incidents of IPV while placement in the lower categories indicates a low risk of recidivism.

Validity

The ODARA is intended to predict the frequency and severity of future violence from known IPV offenders (Hilton, 2010). According to a metadata study by Messing and Thaler (2013) that compared predictive validity studies for five risk assessment tools (ODARA, SARA, DA, DVSI and K-SID), the ODARA has the highest predictive validity using the Receiver Operative Characteristic Area Under the Curve (AUC) at AUC=.666 and k=5.

Application to Criminal Justice System

The ODARA was designed specifically for use my front-line police officers and for male perpetrators assaulting female victims. It is based on data from the criminal justice system. There are several aspects of this risk assessment tool that make it particularly useful for members of the criminal justice system but evaluators must also keep certain elements of it in mind while making their decisions.

One of the most convenient aspects of the ODARA for the Criminal justice system is that the risk assessment is based entirely on police record files which are easily accessible to evaluators. However, the downside of this is that any relevant incidents that might affect the results but were never reported would not be incorporated into the analysis. 

Another important facet to note about ODARA is that it is based almost entirely on static risk factors like criminal history and past events which are unchangeable and part of the permanent record. It does not account for change in the offender’s personality, outlook or circumstances. Hence, while it might be useful for assessing immediate risk by front-line police officers and the like, ODARA is not the best option for anyone who needs to account for intervention or rehabilitation over time like parole officers.

Spousal Assault Risk Assessment (SARA)

The Spousal Assault Risk Assessment (SARA) was developed in Canada specifically for use within the criminal justice system. It is designed to evaluate the risk of spousal violence based on 20 risk factors valuated using structured professional judgement (Kropp and Hart, 2000).

Development

The Spousal Assault Risk Assessment Guide was a manual created to train evaluators to document and communicate the nature, severity, likelihood, frequency and imminence of risk based on a list of 20 risk factors that were identified based on a review of relevant scientific literature and after taking into account pertinent clinical and legal issues (Kropp and Hart, 2000). 

Scoring

The SARA guide focuses specifically on spousal violence and has been designed specifically to prevent future incidents of violence. It incorporates information from case files, interviews with the offender and the victim, and any other factors that the evaluator deems relevant (Kropp and Hart, 2000).

The SARA manual recommends that risk be evaluated based on a minimum of 20 factors but also include additional case-specific factors that are up to the evaluator’s discretion. The 20 listed factors are divided into two parts. Part 1 involves general violence risk factors including past assault of family members, past assault of strangers and acquaintances, past violation of conditional release or community supervision, recent relationship problems, recent employment problems, the offender as a victim of and/or witness to family violence as a child or adolescent, recent substance abuse/dependence, recent suicidal or homicidal ideation/intent, recent psychotic and/or manic symptoms and personality disorders (Kropp and Hart, 2000). Part 2 specifically focuses on spousal violence risk factors including past physical assault, past sexual assault/jealousy, past use of weapons and/or credible threats of death, recent escalation in frequency or severity of assault, past violation of ‘no contact’ orders, minimization or denial of spousal assault history and attitudes that support or condone spousal assault with three additional items focusing on the most recent incident – severe and/or sexual assault, use of weapons and/or credible threats of death, and violation of ‘no contact’ order (Kropp and Hart, 2000). Based on the 20 factors, a trained evaluator uses structured professional judgement and assigns either a low, moderate or high risk rating to the offender. 

Since the SARA scores are calculated based on specific algorithms, there is an actuarial element to the assessment. However, it is important to keep in mind that the conclusions based the summary risk ratings are not made according to fixed guidelines but are reliant on the professional discretion of the evaluator. Hence, the SARA IPV risk assessment is based on structured professional judgement. 

Validity

Although the creators of the SARA intended for the summary risk ratings to be interpreted using structured professional judgement, the predictive validity studies for SARA generally utilize the sum of all ratings in order to do statistical analyses. According to Messing and Thaler (2013), the SARA has the second highest validity (after ODARA) using the Receiver Operative Characteristic Area Under the Curve (AUC) at AUC=.628 and k=6. It is also important to note that three of the six studies included in Messing and Thaler’s meta-analysis only used case files and did not incorporate interview data.

Application to Criminal Justice System

Just like ODARA, SARA is also designed with the criminal justice system in mind to evaluate the risk of recidivism. However, it is meant to be administered by trained clinicians, not police officers. Due to its nature as a structured professional assessment, the requirement for educated and trained evaluators makes it less accessible. Additionally, the data set required for the SARA includes not only criminal files of the offender but also access to clinical files and interviews with the offender and victim which can consume time and resources, and decrease the accessibility of the assessment.

However, the advantage of the SARA is that it takes into account dynamic or changeable factors such as recent relationship and employment concerns or suicidal/homicidal ideation which may change over time and are necessary factors for accurate risk assessment and can be used to track recidivism risk over time.

Conclusion

Intimate Partner Violence is often a preventable form of violence since there are several warning signs before an incidence, specifically a fatal incidence, of IPV. It is also a common human rights and public health issue. Thus, it is important for the criminal justice system to implement policies to prevent IPV before it occurs and risk assessment and management strategies lie at the core of such policies.

The Ontario Domestic Assault Risk Assessment and the Spousal Assault Risk Assessment are both designed specifically with the criminal justice system in mind and have been tested for predictive validity in a number of settings. However, they both have distinctive strengths to offer and conversely, they both have their limitations. The decision regarding which risk assessment would be most appropriate for a professional in the criminal justice system would depend on a number of factors including, but not limited to, the time-frame, the amount of information available, the purpose of the assessment, the decisions that need to be made and the individual circumstances of the case. 

Overall, the recommendations of this report are as follows: in situations where the primary purpose is to provide intervention and rehabilitate an offender, it is essential to take the dynamic risk factors into account and use a risk assessment that tempers the past record of the offender with any changes made following the offense. Thus, for officials such as parole officers, judges considering early release pleas or evaluating safety plans, SARA is the more appropriate risk assessment tool. However, when the purpose of the assessment is to evaluate imminent danger, or when there is limited access to either the perpetrator or the victim for interviews, the actuarial approach that can be more readily administered is best. ODARA is a good choice for police officers evaluating a heated situation or judges considering the length of a sentence. 

It is important to note that both of these risk assessments are based in a rather narrow understanding of intimate partner violence as that of a male perpetrator and a female victim. Neither of them takes into account LGBTQ+ relationships or female-on-male violence. They have both been designed with male-on-female IPV in mind and their predictive validity has been tested in accordance with that. There is a need for research into the validity of these assessment tools in other types of relations or the constructions of other tools to account for other types of IPV. Thus, when professionals within the criminal justice system employ these tools, it is important for them to acknowledge this gap and be judicious in their usage of ODARA and SARA when they are not working with a case of male-on-female violence.

References

Hilton, N. Zoe. (2010). Risk assessment for domestically violent men: Tools for criminal justice, offender intervention, and victim services. 240.

Hilton, N., Harris, G., Popham, S., & Lang, C. (2010). Risk Assessment Among Incarcerated Male Domestic Violence Offenders. Criminal Justice and Behavior, 37(8), 815-832.

Kropp, P. R & Hart, S. D. (2000). The Spousal Assault Risk Assessment (SARA) Guide: Reliability and Validity in Adult Male Offenders. Law and Human Behavior, (1), 101. Retrieved from https://washcoll.idm.oclc.org/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=edsjsr&AN=edsjsr.1394431&site=eds-live

Messing, J. T. and Thaller, J. (2013). The average predictive validity of intimate partner violence risk assessment instruments, Journal of Interpersonal Violence, 28 (7), pp. 1537–58.

Office of National Statistics. (2016). Compendium: Intimate personal violence and partner abuse. Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/compendium/focusonviolentcrimeandsexualoffences/yearendingmarch2015/chapter4intimatepersonalviolenceandpartnerabuse

Singh, Jay P., et al. (2014). International Perspectives on the Practical Application of Violence Risk Assessment: A Global Survey of 44 Countries, International Journal of Forensic Mental Health, 13:3, 193-206, DOI: 10.1080/14999013.2014.922141.

World Health Organization. (2005). WHO multi-country study on women ‘s health and domestic violence against women: Initial results on prevalence, health outcomes, and women’s responses. Retrieved from http://whqlibdoc.who.int/publications/2005/924159358X_eng.pdf?ua=1

---