Definitional issues

The initial response to the use of the term ‘women who use force’ is to ask why not use the term ‘female perpetrators’ or ‘women who use violence’, the same terms used for men but noting the gender difference. The question goes to the heart of the +SHIFT programme’s development—while women who use force are not a homogenous group, the ways in which they predominantly use violence and abuse in their relationships differ markedly from the dominant patterns of male violence towards women in ‘motivation, intent and impact’.1 Patterns of women’s use of force continue to emerge from research that is specifically focused on community-based domestic violence services.2–8

Central to recognising ‘motivation, intent and impact’ is placing the use of violence and abuse in context. It goes beyond a narrow understanding of ‘physical assault’, which has often been used to measure violence.9 Instead, the patterns of abuse and impact need to be explored. It is argued that the use of coercive control is the pattern frequently used in male perpetrated domestic violence, in which power over partners and ex-partners is established through intimidation, a wide range of tactics of abuse (financial abuse, emotional degradation, isolation, undermining the mother–child relationship, sexual assault, physical assaults) and oppression that draws on the wider cultural mores that generally privilege male dominance over women.10

In general, women who use force in their intimate heterosexual relationships are often abused by their male partners;6 11 12 motivated to use force for a range of reasons including protecting themselves and their children and asserting their dignity;11 12 and face severe relationship and societal consequences.10 13 Larance and Rousson analysed 239 unduplicated programme participants during 6 years of a group work co-facilitation, of which 90% of the women were referred by probation or the courts.5 They argue that most women were wanting power through the assertion of personal autonomy from a partner (use of force) rather than having power through the exercise of personal authority over a partner (coercive control).5 The use of force may involve primary or retaliatory aggression with the motivation of gaining short-term control over a situation, rather than an ongoing pattern of coercion and tactics of abuse which create fear and subjugation of the victim.3 Further evidence of the contrasting patterns is provided by Miller et al who observed men and women directed by the courts to single sex group work programmes for their use of violence.7 The researchers found that 95% of the 95 women were not ‘batterers’ (ie, they were not using violence to control others) but rather were involved in defensive behaviour or using force in the short term out of frustration with the abusive behaviour used against them by their partners.6 They talked openly about the incidents that led to their arrest and tended to blame themselves rather than others. By contrast, the men’s group was dominated by men who protested their innocence, and blamed others despite evidence of ongoing acts of violence.

The British Crime Survey with its module on experiences of personal violence may shed further light on this issue.14 Analysis of these self-completed questionnaires showed that approximately 25% of women aged over 16 experienced some form of non-sexual domestic abuse.14 However, the survey also showed (using the same definition) that approximately 17% of men had also experienced domestic abuse since the age of 16. While this is a significantly smaller percentage than that experienced by women, it is nevertheless considerable. Importantly, when the number of incidents in a 12-month period was examined, 81% were male violence against women, and of those who suffered four or more incidents, 89% were women.14 Men were three times less likely than women to be living in fear and three times less likely to be injured. The data suggest that women are involved in perpetrating significantly fewer incidents and these are creating little fear or injury. These data are complemented by detailed analysis of police files in northern England.15 The majority of men (83%) had at least two incidents recorded, whereas when women were named as perpetrators, in 62% of cases, only one incident was recorded.15 The data showed the severity of violence by men was also much greater.

The patterns of violence also raise serious questions about the misidentification of female perpetrators of violence by police and the courts. Programmes for women in the USA initially developed following legislation mandating arrest in cases of domestic violence.16 In this ‘gendered injustice’17 process many women became ‘caught’ in the criminal justice system when they were not the predominant aggressor.2 Recent research highlights the difficulties in identifying victims of coercive control.18 These concerns are live in the Victorian context, where women are being misidentified as perpetrators of violence, due to their violent male partners manipulating the police as an act of coercive control.19

The impact for women who use force suggests that while they may harm others in this process, their use of force will frequently result in greater adverse consequences to themselves.12 They have negligible effect on changing the behaviour of their partners who are generally not afraid of them.6 Most women who use force are themselves survivors of domestic violence either in their current or past adult relationships, or through childhood experiences in their families of origin.3 7 20

Given the complexity of the issues involved, sustainable, effective programmes for women who use force have been slow to develop with issues of feasibility needing to be tested.

Programme development

The dynamics associated with women’s use of force, and that of coercive control,21 call for the development of gender-responsive programming that addresses intersectional identities.22–24 Programme design must acknowledge women’s victimisation and trauma history, while simultaneously facilitating awareness of viable non-forceful alternatives.5 6 25 In Victoria, Australia, the Vista programme framework,3 26 which provides a contextual view of women who have used force, has been identified as the most appropriate to tailor to the Australian context as +SHIFT and test for feasibility.

The Vista programme was attractive because it actively engaged participants in its development; was identified by the antiviolence service sector as a gender responsive service;27 recognised that women who use force are often domestic violence survivors; and is currently used by a range of entities including the US Air Force’s Family Advocacy Program, where it is currently being evaluated. Additionally, the programme is grounded in an ecological approach which recognises and situates women’s use of force in its wider sociocultural context; it is trauma informed in its approach; and it is designed to heal as well as challenge. Other programmes such as Beyond Violence developed by Covington were also explored.28 However, while showing promising evaluation results, the programme is not specific to family violence, and has been implemented in Australia primarily with incarcerated women.

Because +SHIFT has been adapted from the Vista curriculum, it requires feasibility testing for implementation in the Australian context. For example, while the Vista curriculum is typically run as a continuing group, with new participants entering the group at any point in the session cycle, +SHIFT will be piloted as a 16-session group programme, closed to new participants once the programme has commenced. Additionally, +SHIFT eligibility will be more inclusive, for example, participants who have used force in the context of kinship care (eg, the children’s mother assaulting her own mother who has care of the children), in same-sex relationships, and as extended family members.29

A key feasibility testing issue lies in whether organisations (justice and human services) will refer women to the +SHIFT groups, and whether women referred and assessed as suitable will attend. Most US-based programme referrals are Child Protection and/or court mandated.4 16 This route is not as common in the Victorian context for men who use violence, and there is an assumption that the referrals for women who use force will be through the community-based organisations as well as corrections, police and the courts. In the latter case, civil protection orders may have a condition to attend, rather than the order mandated through criminal proceedings.

Study design and timeline

+SHIFT will be implemented at three community-based domestic violence intervention service sites between October 2018 and June 2019. All three regions, two in metropolitan Melbourne and one in regional Victoria, have been identified as having high levels of domestic violence. The metropolitan sites differ in that one is located in an area serviced by a ‘Support and Safety Hub’, a single coordinated entry point to services for families experiencing domestic violence or needing other supports.

The study will adopt a mixed methods approach30 to assess programme feasibility in relation to programme content and quality of delivery, participant recruitment, engagement and retention (process evaluation) and the feasibility of selected measures in providing information about the change process.

The intervention

The +SHIFT programme is a 16-session group work and case management programme, adapted from the Vista programme curriculum developed by Larance and colleagues, in the USA as a contextual response to the complex needs of women referred by the judicial system to behaviour change programmes.26 Like Vista, +SHIFT takes a therapeutic, gender-responsive, trauma-informed approach, addressing intersectional identities while facilitating participants’ engagement with viable alternatives to force and promoting healing.

The intervention includes several components:

1. Assessment and engagement. While this is the first step in serving women, it is conceptualised as an ongoing process throughout a woman’s contact with programme staff.20 The Family Violence Risk Assessment and Risk Management Framework32 will be incorporated into the assessment to identify risk factors associated with family violence. In addition, the Contextual Abusive Behavior Inventory3 will be used at intake to assist women in seeing the contexts in which their use of force has occurred, and to facilitate the beginning of therapeutic conversations.

2. Sixteen group sessions of 2 hours each. Group sessions will occur twice weekly during the first programme cycle. Later programme cycles will offer group sessions once a week for 16 weeks. Group session content addresses a range of issues, including personal identity, impact of force on self and others, defence mechanisms, personal boundaries, shame, communication strategies, protection planning and healthy relationship dynamics. Each session is presented within the ecological nested model framework,33 paying attention to sociocultural issues and how such issues impact women’s perceptions of viable alternatives to using force.

3. Individual case management offered throughout the intervention period to women participants. The groups will be free of charge, and assistance with transport and childcare will be offered. There will be limited involvement with women’s family members—assistance with referrals to appropriate services will be offered.

Inclusion and exclusion criteria

Programme inclusion criteria:

• Women aged 18 years and over who use force in their relationships with other adult family members.

Programme exclusion criteria:

• Current drug/alcohol use at levels which hinder effective programme participation.

• Mental illness that includes psychosis or delusions.

• Severe cognitive limitations.

• Medical conditions constituting the primary cause of violence (eg, acquired brain injury).

• Continued lack of commitment by a woman to increasing her safety or changing her behaviour that causes disruption to the group.

Service providers and training and programme fidelity

The programme will be provided by two agencies with established histories of domestic violence service provision. Programme staff will be required to take part in a 3-day training course provided by the third author.

As the programme will be implemented at separate sites with different facilitators, ensuring programme fidelity across programme cycles is essential. While some group facilitation nuance is built in, due to the diversity of women and their individual histories of trauma and use of force, fundamental programme principles and core content should remain consistent. To this end, a range of strategies will be employed. Two facilitators will attend each group session, and an observer trained in the +SHIFT programme model will attend selected sessions of each programme cycle. The programme manager will provide regular supervision to facilitators and the third author will lead and support reflective practice on a weekly basis.

Sampling and recruitment

+SHIFT will run three times at three separate locations over 9 months—a total of nine programme cycles. Up to 10 participants are expected to take part in each programme cycle, with a possible total of 90 participants. However, due to this new service provision area, intake numbers may be smaller.

Programme participants will be recruited through a range of services, including courts, police, child protection and domestic violence services. Flyers will be circulated to potential referring agencies and staff will also proactively provide outreach.

Evaluative data collection activities are built into the programme design. Once participants are assessed as eligible, they will be invited by programme staff to take part in the evaluation. Consent or non-consent will then be recorded. All programme participants will be asked to complete the self-administered questionnaires. Three randomly selected participants from each programme cycle will be invited to take part in the postintervention semistructured interview, and then again in a follow-up interview, to obtain detail about their programme experiences. Participants who drop out during a programme cycle, or do not attend all 16 sessions, will still be invited to complete the postintervention and follow-up questionnaires. Women who participate in the evaluation will receive a small honorarium at each point of data collection to recognise their time and costs.

Programme staff will be invited to participate in a semistructured, postintervention interview after each programme cycle. In addition, a small number of referring professionals will be identified through the facilitators and approached by the researchers to participate in semistructured interviews.

Data management

To ensure anonymity, programme participants and professionals will each be assigned a unique research ID for use throughout the study. Data will be collected in paper questionnaires and electronic spreadsheets by programme staff and transferred to a secure electronic database at the University of Melbourne. Interviews will be recorded (with consent), transcribed and similarly stored in an NVivo database.

Measures to be assessed for feasibility

When considering outcome measures to be trialled, validated psychological measures were initially reviewed with a preference for brief measures to reduce participant burden. However, this is a relatively new area of research and validated instruments that directly match our objective of providing women with viable alternatives to using force and assisting them to heal from trauma are underdeveloped. As a result, validated scales have been combined with more contextualised but unvalidated intervention tools, developed specifically for programmes using the Vista curriculum. Table 1 summarises outcomes, measures and data collection time points.

Data analysis

Qualitative and quantitative data will be collected for this project. In line with recently published mixed methods evaluations,43 44 the two forms of data will be analysed separately and then synthesised.

The quantitative data will be analysed with the assistance of SPSS version 25 software. Analysis will include descriptive analysis of frequencies and cross-tabulation.

The qualitative data will be analysed by collection source according to the thematic analysis method set out by Braun and Clarke.45 This form of thematic analysis involves an inductive coding process assisted by NVivo software, where patterns of ideas or actions are identified and brought together into meaningful groups. The qualitative data will provide detailed and in-depth data to capture the broad context of women’s experiences that are not captured in measures.

Participant and public involvement

The +SHIFT programme was adapted by staff from the organisations involved in implementing the programme, who were also consulted about the development of the research questions, methodology and tools. Neither the public nor potential participants were involved at this developmental stage, but their views will be obtained during the process evaluation to assess feasibility and improve programme delivery. Interested study participants will receive a summary of the study findings.