Establishing the effectiveness of domestic and sexual violence perpetrator intervention programs is challenging due to a wide range of factors associated with the research methodology, the implementation of intervention programs, and the broader social context.
Apart from whether an experimental or quasi-experimental design was used, other methodological differences among evaluation studies of domestic and sexual violence perpetrator intervention programs include:
the definition of domestic violence or sexual abuse
measurement of recidivism (e.g., based on official records of reconviction or self-reported re-offending by victims)
number and type of outcome measures (e.g., official records of recidivism only or recidivism plus other psychological and behavioural measures)
type of statistical analyses
measures of perpetrator characteristics and inclusion of these measures in data analysis
measures of program characteristics and inclusion of these measures in data analysis
inclusion of a control group or the characteristics of the comparison (no treatment or program dropout) group
involvement of researchers in program implementation
sample size or response rate.
In addition, a number of factors external to the evaluation study can influence the capacity of the evaluation to establish the effectiveness of intervention programs. These factors relate to program integrity and the program context.
Program integrity (e.g., Day et al., 2009; Schrock & Padavic, 2007) – differences in program effectiveness have increasingly been attributed to varying levels of program integrity (i.e., the extent to which the program is delivered as intended). Program integrity is a critical factor to reaching the intended program outcomes. However, achieving program integrity is challenging as program facilitators differ considerably in experience, training and qualifications. Limited organisation resources (e.g., staff shortage) can also compromise program integrity (e.g., ability to deliver group sessions with two facilitators rather than just one facilitator). Schrock and Padavic (2007) provided evidence that ineffective implementation rather than the curriculum of the program led to poorer outcomes for program participants.
A program may be either less effective or ineffective altogether because of inadequate funding, a more challenging client group, and limited capacity among program facilitators to engage or motivate program participants and address comorbid conditions or other external risk factors for program participants (Prescott & Levenson, 2009).
Differences in program effectiveness have been attributed to differences in the context of program implementation (Carson, Chung & Day, 2009; Gondolf, 2004; 2009), for example, the extent to which the program is part of a coordinated and multi-level intervention, as advocated by the Duluth Model. The effectiveness of the program can also be enhanced or offset by the broader societal or community norms about the gender identities and power hierarchy. In recognition of the importance of context in program outcomes, multi-site evaluations are increasingly favoured as a means of assessing contextual impacts on program outcomes (e.g., Day et al., 2009; Gondolf, 2004; 2009).
There is a continuing debate among researchers and practitioners about the extent to which program standards should be based on evidence, and how different types of evidence should be used in the development of standards.
While there are currently no national standards for domestic violence perpetrator intervention responses in Australia, some jurisdictions have developed practice guidelines or standards for service design and delivery. Victoria, Queensland and Western Australia, in particular, have relatively comprehensive practice guidelines or standards for the delivery of domestic violence perpetrator intervention responses.
Current standards document or guidelines include:
Men’s behaviour change group work: Minimum standards and quality practice (No To Violence, Victoria)
Professional Practice Standards: Working with men who perpetrate domestic and family violence (Department of Communities, Queensland)
Guidelines for responding to family and domestic violence (Department of Health, Western Australia, 2007)
Competency standards for intervention workers: Working with men who perpetrate domestic abuse and violence (Department of Human Services, South Australia, 2001)
NSW has recently produced an Issues Paper on its Minimum Standards for Domestic Violence Behaviour Change Programs (Department of Justice and Attorney General, NSW).
While the South Australian standards focus on workforce requirements, the Victorian, Queensland and Western Australian standards cover a number of elements. These elements include definitions, the conceptualisation of men’s violence, safety, client characteristics, workforce requirements, program intake and delivery practices, safety and risk assessment, partnerships and referrals and program review and evaluation. The Queensland standards identify unacceptable and essential practices, and articulate the standard required and the expected outcomes for each element. The Victorian standards include minimum standards as well as practice guidelines.
In the area of sexual assault, there appears to be one key standards document developed by the National Association of Services Against Sexual Violence (NASASV, Carmody, Evans, Krogh, Flood, Heenan & Ovenden, 2009) – Framing Best Practice: National standard for the primary prevention of sexual assault through education.
In a review of standards for domestic violence perpetrator treatment in the USA (Maiuro & Eberle, 2008), only five states did not have standards or regulations for treating convicted or non-convicted domestic violence perpetrators (as at 2007). The primary treatment philosophy in most state standards is derived from a feminist perspective that conceptualise domestic violence in terms of power and control (95%). The majority of standards combine the notions of power and control with social psychological approaches (e.g., focusing on attitudes toward violence and sexism, skill deficits and poor role models in the family) (68%). Many states explicitly forbid the primary use of treatments based on psychopathology or disease models, psychodynamic theory, and models focused on family systems, impulse control disorders, co-dependency or addiction (35%) because they are perceived to reduce the perpetrator’s sense of responsibility or compromise victim safety.
Almost all (98%) of the standards emphasise group therapy as the preferred and primary mode of intervention. It is noteworthy that despite the lack of evidence on the relative effectiveness of treatment modalities, a small number of states prohibit the use of individual treatment (5%), and the majority of states (68%) prohibit any form of couples sessions during the course of the primary intervention programs.
In terms of the intake protocol, the majority of states (75%) require intake assessment or the participation in other related programs such as substance use treatment prior to program participation. Most states (91%), however, enforce a uniform course of treatment (i.e., a one-size-fits-all approach) for all perpetrators regardless of the findings of the intake assessment, with only three sates allowing for different treatment programs based on whether a perpetrator is a first time or repeat offender, and the perpetrator’s risk level.
For 93% of the standards, some type of victim contact is required. Training requirements for facilitators vary, with 40% requiring an undergraduate degree in a human services field and a specific length of experience or training in the domestic violence field. A small number of states (15%) required a postgraduate qualification for facilitators (usually at a Masters level). Approximately one-third of the states do not have specific educational requirements but require training and experience in the domestic violence sector.
The majority of standards (75%) do not include references or evidence for the requirements specified. For those standards that include references, the majority cited dated references or only a small number of references. Approximately one-third of the standards mention the need for program valuation and research, however, few (18%) require actual data collection to determine program effectiveness.
Based on this review, Maiuro and Eberle (2008, p. 147) identified a number of positive trends in standards development:
increasing multi-faceted nature of content of treatment protocols (despite the treatment theoretical basis)
increasing number of standards requiring intake assessment before treatment
increasing recognition of the need for program evaluation and research
requirement to collect standardised data sets for some standards
increasing minimum level of education required for treatment providers.
Recommendations made by Maiuro and Eberle (2008, p. 149) to improve standards include:
expanding one-size-fits-all treatments to allow client-centred, evidence-based multivariate content and case management
enhancing screening and referral for comorbid conditions
improving risk assessment through standardised checklists
including researchers familiar with current literature on domestic violence on standards committees
developing systematic program evaluation protocols
developing modified standards for female perpetrators, military, gay and lesbian and minority populations.
A set of guidelines to develop standards for male domestic violence perpetrator programs was developed by Work with Perpetrators of Domestic Violence in Europe (2008). In particular, these guidelines relate to the pre-conditions for working with male perpetrators and key principles for the work with male perpetrators.
These pre-conditions for working with male perpetrators include:
understanding and explicitly articulating the goal for working with male perpetrators (i.e., increase the safety of the victims of violence)
collaboration with victim support services and intervention systems – perpetrator programs should be integrated as part of the broader intervention system
theoretical understanding and explicit articulation of key concepts, including gender theory, definition of domestic violence and types of abuse, origins of violence, and theory of intervention or change
Focus on relevant dimensions for the use of violence, including socio-cultural, relationships and individual factors (including cognitive, emotional and behavioural factors).
The principles for working with male perpetrators include:
partner contact and support – it is important for partners to be informed about the program content, goals and limitations, and partner input should be obtained to inform perpetrator risk assessment. Partner contact should be voluntary and the risk for partners (by contacting them) should be minimised. a child protection policy should be included in the program
approaches and attitudes in direct work with perpetrators – for example, holding the perpetrators accountable for their actions, treating perpetrators with respect, violence as a choice, and developing victim empathy.
risk management should be undertaken systematically
staff qualification – facilitators should have a commitment to violence free relationships and gender equality, violence awareness training, reflect on their own relationships, attitudes and behaviours, group work skills, comprehensive understanding of the dynamics of violent relationships and have access to continuing education and supervision
quality assurance, documentation and evaluation should be integral to each program and strategies should be developed and implemented to monitor program activities and outcomes
Given that much is still unknown about what program components are effective in reducing recidivism, there is limited capacity to analyse existing standards against current empirical evidence.