Go to top of page

Appendix C

Selected evaluations

Selected international evaluations of criminal justice system-focused family violence interventions 1998–2009
Author/year/jurisdiction Scope, sample and method Key findings

Murphy CM, Musser PH & Maton KI 1998. Coordinated community intervention for domestic abusers: Intervention system involvement and criminal recidivism. Journal of Family Violence 13(3).

Baltimore, Maryland, United States

Focus on male perpetrators of domestic violence.

Sampled 235 cases and assessed recidivism over 12–18 month period after cases were first handled by the Maryland State’s Attorney’s Domestic Violence Unit.

The sample was drawn from the Unit’s records of men charged with domestic violence-related offences between January and August in 1994. This data was then combined with follow-up data through to September 1995.

Two different recidivism measures were then applied to the cases—one with a narrow range of offences and one that had a broader range.

Coordinated interventions may have a cumulative effect on the risk of recidivism, in that those offenders whose prosecution was dropped or found to be not guilty had higher rates of recidivism (19% in the narrow measure and 29% using the broad measure) compared with those persons in the sample who were successfully prosecuted (13% and 23%).

Those who were prosecuted and also received domestic violence counselling, or probation, or both, had a reduced recidivism rate of as low as nine percent for the narrow measure and 18 percent using the broad measure.

This study found that the greater the range of sanctions and intervention applied to domestic violence perpetrators, the greater the reduction in the risk of reoffending.

Giacomazzi AL & Smithey M 2001. Community policing and family violence against women: Lessons learned from a multiagency collaboration. Police Quarterly Journal 4(1).

South-west United States

Sought to analyse and evaluate the process of multi-agency collaborative efforts between a large municipal police department and other service providers in combating domestic violence.

Within the context of a community policing framework, this study examines the barriers that may inhibit collaborative multiagency approaches to domestic violence in one unnamed urban location with a population of 500,000 residents in the south-western United States.

Interview with focus groups in addition to archival research was used to collect data. Participants in the focus groups were representatives from randomly sampled domestic violence service agencies. In total, 14 individuals participated.

One of the main barriers to collaborative multi-agency approaches may be an underlying desire by service staff to protect their own agency rather than actively engaging in collaborative efforts.

Poor communication between agencies is also identified as a common barrier in multi-agency efforts, with many participants in this study citing a lack of organisation, scheduling of meetings and unclear expectations of each agency’s role and service coverage as inhibiting collaboration.

Creative responses to domestic violence are best developed through collaborative multi-agency partnerships. However, it cannot be presumed that personnel in relatively autonomous organisations, both public and private, have the organisation capacity and/or willingness to truly collaborate.

Shephard MF, Falk DR & Elliot BA 2002. Enhancing coordinated community responses to reduce recidivism in cases of domestic violence. Journal of Interpersonal Violence 17(5).

Duluth, Minnesota, United States

Examined male domestic violence offenders who entered the DIAP Men’s Nonviolence Program in 1994 (n=261), 1996 (n=217), 1997 (n= 220) and 1998 (n=100), as well as identical numbers of female domestic violence victims over the same period.

The analysis involved comparing those in the pre-intervention period (1994) with those in the pilot year (1996) and again with those in the two intervention periods (1997 and 1998).

Participants were assessed with regard to 25 different risk factors for reoffending.

Those participants who successfully completed an intervention during either 1997 or 1998 were found to have a reduced risk of recidivism.

This study suggests that the sharing of risk assessments of domestic violence offenders between service agencies can enhance coordinated responses to domestic violence, in turn, leading to a reduction in recidivism.

Overall, declines were seen in the risk of recidivism in participants over the three and a half years of this study.

Zweig JM & Burt MR 2003. Effects of interactions among community agencies on legal system responses to domestic violence and sexual assault in STOP-funded communities. Criminal Justice Policy Review 14(2).

United States

To examine whether federal funding of non-profit victims support programs has produced more collaborative and coordinated service delivery in the community.

Two hundred interviews were conducted with staff involved in 90 different programs in 32 different states across the United States.

Data was also collected through questionnaires sent directly to federally funded sexual assault and domestic violence agencies.

This study sought to measure the performance of the 90 different victims support services in four key areas—level of STOP funds and other resources, pre-STOP level of community services, state STOP program support for collaboration and community interaction.

The more agencies work together in communities, including law enforcement and prosecution agencies working with victim services agencies, the more likely services are to improve for victims within the legal system.

Producing positive change in service coverage is partly achieved through more positive interactions between law enforcement, prosecution and their behaviour towards victims.

Interactions among non-profit victims support program staff, law enforcement and prosecution seem to lead to changes in the legal system’s approach to handling domestic violence and sexual assault cases and assists communities in meeting the needs of victims of violent crime.

Pennell J & Francis S 2005. Safety conferencing: Toward a coordinated and inclusive response to safeguard women and children. Journal of Violence Against Women 11(5).

United States

Discusses the effectiveness of safety conferencing in supporting victims and their families overcome the experience of domestic violence and/or sexual assault.

This academic paper develops a theoretical and practical framework for the implementation of safety conferencing, which is a four stage process designed to produce a collaborative and inclusive support network for victims.

The four processes involved in safety conferencing are discussion in an advisory group, focus groups with women who were battered and separately with shelter staff, exchanges with domestic violence counsellors and feedback from the focus group participants.

Attendees at these meetings include representatives from batterer programs, children’s services, police, domestic violence court, correctional services and social work.

Discussions at the focus groups identified a number of challenges that need to be overcome if safety conferences are going to be beneficial to victims of domestic violence and sexual assault, these include—continued connections between victims and batterers, isolation felt by survivors, difficulties in providing services to all cultures and the limitations of a coordinated response that includes only services and not victims or their families.

It was concluded by many participants in the focus groups that victims and their informal support networks need to be at the centre of safety conferencing.

The safety conference, and more broadly a coordinated and inclusive response, is an effective way to reshape connections in the community, in turn promoting the safety of women and children from all backgrounds.

Hovell MF, Seid AG & Liles S 2006. Evaluation of a police and social services domestic violence program—Empirical evidence needed to inform public health policies. Journal of Violence Against Women 12(2).

San Diego, California, United States

This evaluative study examines the effectiveness of the Family Violence Response Team pilot program for victims of domestic violence who have children.

Using an experimental cohort study design, 307 families who had experienced domestic violence, had children and who received FVRT intervention, were compared with a control group of 498 residents that had an initial domestic violence incident but no FVRT treatment.

Victims in the intervention group were interviewed at six and 12 months about the effectiveness of the services received by the FVRT.

Cases of domestic violence were also examined and grouped according to the risk of recidivism. Analyses were then carried out to identify those factors that increased the risk of recidivism between the control and intervention groups.

Outcome for victims are more positive when intervention services are tailored to specific individual and family needs.

While this FVRT pilot program was designed to empower the victim, it may also unintentionally generated conditions that can precipitate further violence.

This program was implemented without the batterer’s consent, which may have precipitated a loss of control by the batterer and increased the risk of recidivism.

Variance within and across case managers and the families to whom they provided services may have contributed to a decreased ability to detect real differences with this analysis.

Alternatively, the combined police and social services intervention may have increased in the rate of reporting of domestic violence in the intervention group. This creates a situation where the intervention may have resulted in both increases in reporting and in the frequency or severity of violence.

Zweig JM & Burt MR 2007. Predicting women’s perceptions of domestic violence and sexual assault agency helpfulness—What matters to program clients? Journal of Violence Against Women 13(11).

United States

Assesses whether community agency interactions, the characteristics of services provided by staff and the combinations of services received can predict women’s perceptions of victims service helpfulness around domestic violence and sexual assault.

Involving a two-stage data collection method, this study sought to ascertain the level of interagency collaboration between federally funded victim’s support services and the effect on victim satisfaction and perceptions of agency helpfulness.

The first stage of the data collection process involved surveying 200 program directors or staff within victim’s service agencies, and the second phase involved interviewing women from a variety of communities in which services are available.

The more both victim service and legal system agencies worked together to assist women, the more positive and fewer negative behaviours the staff participated in and the higher women’s sense of control when working with the agency, the more helpful women found the domestic violence agencies work.

Women find the services of private non-profit victim service agencies more helpful based on the characteristics of staff behaviour in those agencies and the helpfulness of these services in enhanced when victim service agencies interact with the legal system and other relevant agencies in their community.

Staff behaviour also influenced domestic violence victims’ willingness to use such agencies again if the need arose.

Robinson AL & Tregidga J 2007. The perceptions of high-risk victims of domestic violence to a coordinated community response in Cardiff, Wales. Journal of Violence Against Women 13(11).


The purpose of this study was to identify the levels of re-victimisation among very high-risk victims of domestic violence within a 12 month period following a MARAC and their perceptions of this intervention.

One hundred and two women over a four month period between October 2004 and January 2005 were interviewed and then follow-up interviews were conducted 12 months after their MARAC assessment.

Interagency collaboration was seen as very beneficial by victims of domestic violence, partly because victims felt that they were receiving an enormous amount of support.

Some criticisms that arose were associated with the need to ensure that regardless of the number of agencies involved, information about case developments needed to be provided to victims as soon as practicable.

A multi-agency response needs to take into account the complex set of issues that confront women to ensure that they have the resources and support that they need when they feel able to make the difficult decision to leave and ultimately that will help them remain free from victimisation.

Salazar LF, Emshoff JG, Baker CK & Crowley T 2007. Examining the behaviour of a system: An outcome evaluation of a coordinated community response to domestic violence. Journal of Family Violence 22.

Georgia, United States

Sought to identify whether a coordinate community response to domestic violence in Georgia would be effective at increasing criminal justice system sanctions for male perpetrators.

A longitudinal interrupted time series experimental design was used, with chi-squared tests applied to the data to determine whether there was any changes in the percentage of cases that received probation, goal, or fines and to assess changes in the percentages of men that were mandated to attending batterer programs.

The data used in this study included arrest records, prosecution and conviction, and sentence outcomes.

Following the introduction of a coordinated community response to domestic violence in the two counties in Georgia, there was a significant, abrupt and sustained increase in the number of arrests, particularly of males.

Prosecutors were found to be only pursuing a conviction if it was assured, the vast majority of which are achieved through plea-bargaining.

This coordinate community response was designed with multiple elements directed at a variety of audiences using a variety of modalities over a wide range of time. Consequently, the outcomes from this evaluation may not be as promising, and should be implemented with caution optimism.

Banks D, Landsverk J & Wang K 2008. Changing policy and practice in the child welfare system through collaborative efforts to identify and respond effectively to family violence. Journal of Interpersonal Violence 23(7).

United States

Focusing on six demonstration sites, this study examines the degree to which the Greenbook recommendations have been implemented and the impact of a multi-agency coordinated response in changing organisational practice. The Greenbook is a series of guidelines for collaboration between child welfare and domestic violence agencies, particularly in cases where there is a co-occurrence or maltreatment and domestic violence.

This evaluation was undertaken with the participation of program staff, rather than an independent evaluation. A combination of quantitative data, in the form of reviews of child welfare cases and surveys of direct service works, and qualitative data was used.

Qualitative data indicated that the Greenbook initiative has served to define the activities of different agencies and as a result, they became more valuable and productive in the child welfare environment.

The majority or sites had implemented activities related to multidisciplinary case review, in support of the finding that child welfare agencies are more likely to work with domestic violence service providers to address co-occurrence.

Confidentiality and information sharing were identified barriers to more effective collaborative service delivery.

Banks D, Dutch N & Wang K 2008. Collaborative efforts to improve system response to families who are experiencing child maltreatment and domestic violence. Journal of Interpersonal Violence 23(7).

United States

Examines the effectiveness of federally funded programs through a series of stakeholder interviews.

In total, 86 individuals completed the baseline survey and 62 completed it at the follow-up stage.

The survey was designed to capture those dynamic factors that contributed to project planning, activity implementation and the level of collaboration in each location.

The survey also identified those obstacles that may be inhibiting collaborative efforts.

While increased involvement of male perpetrators in batterer programs occurred during the intervention, the sustained involvement of victims and their families was declined slightly.

Some of the barriers to collaborations between service providers were a lack of time by participants, conflicting organisational cultures and a poor understanding of the initiative.

Stakeholder responses indicated that needs assessments, relationship building and engaging the community were most critical to effectively planning services for victims.

Malik NM, Ward K & Janczewski C 2008. Coordinated community response to family violence—The role of domestic violence service organisations. Journal of Interpersonal Violence 23(7).

United States

This study involved a survey of victim service staff and upper-level managers about the level of collaborative efforts between domestic violence service organisations and other service providers that work in the domestic violence system.

In total, 167 domestic violence staff members participated across six locations, with an additional 47 participants from other service providers.

This study employed a participatory action research design, rather than the more traditional experimental model. Qualitative data was used throughout the study to complement the quantitative data.

Truly collaborative efforts may have been undermined by a perceived hierarchy within the service providers and justice agencies, with the courts having the most power.

For collaborations to function effectively, there needs to be an ‘inclusive climate’ where there exists equality of power among participants. Of equal importance is the need to ensure that each agency is aware of their role and that of other agencies in the child maltreatment and domestic violence sectors.

The remains issues around ongoing program funding, for despite enthusiasm for increasing services, there were, quite often, insufficient resources to do so.

Klevens J, Baker CK, Shelly GA & Ingram EM 2008. Exploring the links between components of coordinated community responses and their impact on contact with intimate partner violence services. Journal of Violence Against Women 14(3).

United States

This study sought to evaluate the impact of a coordinated community response on the level of intimate partner violence across 10 intervention sites. Also of importance was the need to better understand the perceptions of victims towards the effectiveness of specific intimate partner violence services.

Six hundred participants from the 10 different intervention sites completed a 16 item survey, which looked at exposure to a variety of different forms of abuse. Questions were also asked about level of contact with specific intimate partner violence services in their community.

This study focuses more on individual factors and how they impact on intimate partner violence rates in comparison communities rather than the intervention as a whole.

Overall, the evaluation found that the coordinated community response has little impact on rates of intimate partner violence across the 10 intervention sites.

However, some specific factors contributed to improved service delivery to intimate partner violence victims, these included developing goals based on community needs, selecting priorities based on the salience of the need in the community, efforts to coordinate services and disseminating information on the frequency of intimate partner violence in the community.

In this way, the success of this particular program was more around awareness raising of both this prevalence of the problem in the community and the types of services that are available to support victims. Less pronounced changes were produced in actual levels of intimate partner violence and service coverage.

Post L, Klevens J, Maxwell CD, Shelly GA & Ingram E 2009. An examination of whether coordinated community responses affect intimate partner violence. Journal of Interpersonal Violence 25(1).

United States

This study examines the impact of a coordinated community response on reducing intimate partner violence and raising awareness and changing attitudes.

A random telephone survey method was used, and 12,039 participants were involved from 10 test and 10 control sites across 23 different counties.

To establish the prevalence of knowledge, attitudes, behaviours and rates of intimate partner violence, a stratified random sample of approximately 600 adults from each of the 23 sites.

The majority of respondents reported that slapping your partner was acceptable (88% strongly) and that one or more criminal intervention should be used against intimate partner violence perpetrators (85% strongly).

The interval between the intervention and expected impact may have been too short, as developing relationships and mechanisms for coordinated response and implementing programs can take several years.

It was assumed that intimate partner violence behaviour was intrinsically linked to perceptions and attitudes about the acceptability of violence and about gender roles. However, following the implementation of the coordinate community response, the reported levels of intimate partner violence increased. This may be attributed to awareness campaigns in the text communities, serving to sensitise the population and increase the willingness to report.

Morgan M, Coombes L & McGray S 2008. Responding together: An integrated report evaluating the aims of the Waitakere Family Violence Court protocols. Wellington: Ministry of Justice

New Zealand

Evaluates New Zealand’s Waitakere Family Violence Court (WFVC) and makes recommendations for improvement.

Describes the operation of the Waitakere Family Violence Court.

Discusses the role of non-government organisations in the court.

Describes programs provides to victims and offenders.

Describes the perceptions of victims involved with the court.

The report finds that the WFVC has been successful in collaborating with the community and taking violence seriously. The authors recommend the development of closer relations with the Family Court, more extensive victim services and extended specialist services.

Hanmer J, Griffiths S & Jerwood D 1999. Arresting evidence: Domestic violence and repeat victimisation. London: Home Office

United Kingdom

Project commissioned as part of the Home Office Police Research Group program on repeat victimisation in the United Kingdom and draws on earlier research by the West Yorkshire Police.

The aim of the project was to reduce repeat victimisation through a three-tiered program of operational interventions. Actions at each of the three levels of intervention are described. An equal focus on the victimised woman as on the offending man is required in order to set up an interactive crime prevention approach that both protects the victims and demotivates the offender. The program required police officers to proactively ensure the safety of the victims and develop close interagency involvement.

Achievements included reduced repeat victimisation through early intervention, systematic identification of repeat offenders, women encouraged to ask for assistance, identified factors associated and not associated with repeat victimisation, developed recording categories for domestic violence and improved interagency communication and cooperation.

Selected Australian evaluations of criminal justice system-focused family violence interventions 2002–09
Author/year/jurisdiction Scope, sample and method Key findings

Rodwell L & Smith N 2008. An evaluation of the NSW Domestic Violence Intervention Court Model (DVICM).

Sydney: NSW Bureau of Crime Statistics and Research

Evaluation of the specialist court model used at Campbelltown and Wagga Wagga (implemented 2005)

Model comprises five initiatives:

  • the distribution of evidence collection kits to police;
  • the establishment of victims’ advocate services;
  • measures designed to speed the finalisation of prosecutions;
  • perpetrator programs; and
  • regular interagency meetings.

Evaluation consisted of:

  • Police and court data extraction and analysis, using four pre and three post-DVICM time periods, pair-wise tests between time periods to identify pre-existing trends and changes post DVICM implementation. All domestic violence incidents recorded within New South Wales between 2003–07 were used for analysis. Local areas outside Wagga Wagga and Campbelltown used for control group comparison.
  • Victim survey (50 interviews conducted, structured open and closed questions, voluntary, finalised cases only, $25 gift voucher given to participants).
  • Key stakeholder interviews (41 interviews conducted).

Found that while rates of reporting and charging did not change, victims and stakeholders were generally supportive of the model.

The evaluation did not find clear evidence of an increased number of reports of domestic violence incidents or an increase in offenders being charged following the introduction of the court model.

Guilty pleas did not increase and charges withdrawn or dismissed did not decrease following introduction of the model.

There was an increase in the use of bonds as a penalty at Campbelltown but no change in the penalties imposed at Wagga Wagga. There was some evidence of an improvement in time taken to finalise prosecutions.

Victims reported that they were largely satisfied with the response from police and victims advocates, but were often critical about court procedures and outcomes.

Stakeholders thought the pilot was successful and the initiatives should be expanded.

Western Australia. Department of Justice. Court Services Division; Western Australia. Police Service. Crime Prevention and Community Support Division 2002. Joondalup Family Violence Court: Final report

Perth: Department of Justice

Joondalup Family Violence Court (JFVC) project, pilot 1999, aimed to:

  • improve the criminal justice response to family violence;
  • make perpetrators accountable for their behaviour;
  • support victims in the criminal justice system and ensure their safety; and
  • reduce the incidence of family violence in the Joondalup district.

The JFVC utilised an interagency and alternative sentencing approach for dealing with the civil matters of Violence and Misconduct Restraining Orders, and all criminal matters related to family violence.

Evaluation tested effectiveness in achieving its stated aims and key information regarding the profile of domestic violence in Joondalup and other nominated locations.

Evaluation methodology comprised:

  • an analysis of domestic violence incidents in three police districts;
  • an evaluation of the changes brought
  • about by the intervention;
  • an evaluation of the court’s effectiveness; and
  • an evaluation of the cost-effectiveness of the process.

Used a control group for comparison. Compared outcomes of the control group, comprising individuals attending two other Courts of Petty Sessions, with the outcomes of individuals attending the JFVC.

Two risk assessment tools were used in the pilot—the Spousal Assault Risk Assessment (SARA) and the Danger Assessment Scale (DAS).

Risk assessment instruments found to be valid and appeared to identify varying degrees of risk.

Offenders in the control courts were more likely to have substantial criminal histories for a variety of offences, while offenders in the JFVC were more likely to be specific domestic violence perpetrators.

Individuals in the control courts were more likely to be referred to a range of other programmes apart from domestic violence perpetrator programs than offenders in Joondalup.

Offenders in Joondalup were more likely to breach their orders than offenders in the control court, put down to the increased supervision at JFVC and highlights the need to have ongoing communication with victims and other agencies when supervising domestic violence offenders.

The costs of the JFVC are relatively high by comparison with other courts, primarily due to dedicated staffing at the court, lost or delayed resolutions and associated programmatic interventions.

The critical success factors identified in the JFVC were:

  • the clear identification of prosecution matters related to domestic violence;
  • the use of affidavit-assisted restraining order applications;
  • the identification on listings of domestic-related restraining orders;
  • the use of a single day for the hearing of criminal matters;
  • the use of interagency information in the development of pre-sentence reports;
  • the use of case management techniques for the supervision of offenders;
  • the use of the SARA and the DAS as valid risk assessment tools; and
  • the provision of support for victims attending court for domestic violence-related matters.

From these findings, a number of recommendations were developed, providing a rationale and guidelines for a proposed future family violence court model for Western Australia.

New South Wales Attorney Generals’ Department Violence Against Women Specialist Unit 2004. The Dubbo domestic violence project: Integrated response to domestic violence by the criminal justice system and support services. Sydney: NSW Attorney Generals Department

A report on the Integrated Response to Domestic Violence (IRDV) by the Criminal Justice System and Support Agencies—Dubbo (IRDV) Project, which was funded by the Commonwealth Partnerships Against Domestic Violence and managed by the NSW Violence Against Women’s Specialist Unit. The Project was designed to develop a model for an integrated response to domestic violence by the criminal justice system and support agencies in Dubbo, New South Wales. It documents the key steps needed for the basis of an integrated response to domestic violence. Part 2 examines models in other locations in Australia and overseas through a literature review that explores concepts of ‘integration’ and ‘collaboration’. In addition to a literature review the project included:

  • focus group discussions;
  • survey of service providers;
  • provision of training; and
  • trialling new service delivery methods.

Focus groups identified gaps and problems with service delivery, particularly experienced by Aboriginal women.

Survey results were used to develop the training program.

Trialling service models identified a need for crisis-oriented service provision, which led to the implementation of an ‘on-call’ service and a police sticker referral system. This latter service is an alternative to the on-call service and allows victims who identify that they do not want to use the call-out service at the time of the incident to receive support at a later time. Police ask if they can pass on details to the court assistance service and victims sign the sticker to confirm their consent. The sticker is placed on the police interview record. The court assistance service contacts the victims the following day.

An MoU was developed and signed and seen to be a key facilitating feature of an integrated response.

Wangmann J 2003. The Tamworth domestic violence project: An evaluation of a different model of service provision to victims of domestic violence in a police setting. Sydney: NSW Attorney Generals Department

Presents the key findings, aims and objectives of the Tamworth Domestic Violence Project and discusses the methodology and results of its evaluation in terms of who and how the project assisted, the unique features of the model adopted, police engagement with the project and clients’ perceptions of overall benefit deriving from the project.

Essentially directed at enhancing follow-up services for victims of domestic violence and reducing the incidence of repeat violence, the project was based at the Tamworth Police station and ‘proactively’ approached victims in an attempt to provide a ‘middle road’ between conventional shelter/refuge services and government/criminal justice agencies.

Operated from Feb–Nov 2001.

A civilian project officer was placed in Tamworth Police Station to contact the people who experienced domestic violence and to provide support, counselling and referral. This proactive scheme approached the victims instead of merely offering the service. Around 33 percent of the potential clients were assisted, 30 percent rejected the service and 21 percent ignored the telephone calls, letters or home visits. The people who received the service found that the civil character of the project officer was less intimidating and more accessible than police officers and that the assistance made a difference to their dealing with the violence. Police were also satisfied with the outcomes of the project.

Last updated
3 November 2017