Preventing violence against women and girls - what works: effective investments summary

2. What works to make environments safe?

This section looks at interventions aimed at making specific environments safer for women and girls. It looks at physical environment interventions, bystander programmes and domestic abuse disclosure schemes.

2.1 Interventions focused on modifying unsafe physical school environments:Effective

There is strong evidence that the Shifting Boundaries programme (focused on classroom and physical environments) is effective in preventing VAWG.

According to Taylor et al (2013:64), within middle schools^[17] in the USA this curriculum for the classroom-based elements of this intervention involved six sessions that focused on:

• the laws and consequences for perpetrators of dating violence and sexual harassment
• the social construction of gender roles
• healthy relationships
• the definitions and applications of 'personal space' and boundaries

The building-based interventions within the Shifting Boundaries programme include:

• use of building-based restraining orders^[18]
• higher levels of faculty/security presence in safe/unsafe 'hot spots' mapped by students
• posters to increase dating violence and sexual harassment awareness and reporting

As Taylor et al. (2013:64) outline, the building-based element of this intervention: "aim to have students develop a greater sense of respect for personal boundaries, as with the classroom curriculum, but through different mechanisms".

High-quality evaluations of the Shifting Boundaries intervention indicate that there were reductions in perpetration and victimisation of sexual harassment, peer sexual violence, and adolescent relationship abuse (De Gue et al., 2014; Taylor et al. 2017). Through their randomised control trials (RCT) across 23 US middle schools, Taylor et al. (2017:94) report that:

for most of our examined outcomes^[19], providing the Shifting Boundaries program to the 6^th grade^[20] only in middle school does just as well in terms of peer violence and dating violence outcomes as a more saturated process of treating multiple middle school grades.

In this context, Taylor et al. (2017) conducted research where 'full saturation' involved conducting the intervention with grades 6 to 8; school ages 11 to 14.

While their results showed that providing the Shifting Boundaries programme to one grade (6^th, with children aged 11-12) did as well at preventing peer violence and adolescent relationship abuse as treating multiple grades, their results also showed that additional saturation led to sexual harassment reductions (Taylor et al. 2017). In particular, "schools that delivered Shifting Boundaries to 6^th and 7^th graders (compared to just 6^th graders) reduced sexual harassment victimization 6 months post-treatment" (Taylor et al. 2017:79).

Taylor et al. (2013) note that:

• combining classroom and building-level interventions is more effective in reducing sexual harassment and violence than classroom intervention alone
• the building-only Shifting Boundaries intervention can be implemented with very few extra costs to schools

Although Taylor et al.'s (2017) results show that combined classroom and physical environment interventions can be effective in reducing violence among children aged 11-14. They suggest that their results raise further questions about whether such programmes should "work with even younger groups to invoke a true primary prevention effort to reduce abusive behaviours in peer and dating relationships" (Taylor et al. 2017:95).

Key findings for service delivery and practice:

• The physical environment can play a relevant role in reducing the risk of violence perpetration, and the Shifting Boundaries programme offers lessons as an effective intervention.
• Some building-based changes such as use of building-based restraining orders, higher levels of faculty/security presence in safe/unsafe 'hot spots' mapped by students, and posters to increase dating violence and sexual harassment awareness and reporting can be implemented with very few extra costs.

2.2 Bystander interventions: promising

Bystander^[21] approaches aim to shift: "gender inequitable attitudes, beliefs and cultural norms which support abuse, and ultimately increasing pro-social bystander behaviour^[22] to prevent it" (Gainsbury et al. 2020:2).

The most robustly evaluated bystander interventions have been predominantly based in secondary school environments. There is strong evidence to suggest that some bystander programmes (e.g. MVP) are promising interventions to prevent VAWG, but there is variation in the evidence of programme effectiveness between different bystander intervention programmes.

Evaluations of bystander interventions predominantly focus on attitudinal change, rather than the reduction of violence as an explicit outcome, and it should be noted that "attitude change does not guarantee behaviour change" (Flood, 2006:28). Evidence, however, does suggest that attitude is linked to perpetration. Studies have found that men who hold negative gender role attitudes, alongside the belief that their peers find violence against women acceptable, are more likely to be perpetrators of VAWG (Schwartz et al. 2001). Conversely, those men who believed that their peers found such violence unacceptable were less likely to become perpetrators, even if they held those negative gender attitudes themselves (McNaughton Reyes et al. 2015). It can therefore be argued that disrupting these beliefs could contribute to violence reduction (see Education Scotland, MVP progress report 2018-19).

Berkowitz^[23] identified four stages that must be present for bystanders to act: notice the behaviour; interpret it as a problem; feel responsible for taking action; and have the skills to act. This can be a helpful model when assessing the evidence for bystander approaches.

One well established and extensively evaluated bystander intervention is Mentors in Violence Prevention (MVP). MVP programmes are most commonly undertaken within schools and university campuses, with those programmes shown to be effective having been delivered within these contexts. Katz et al. (2011:697) describe the MVP programme as "a peer-driven, prosocial bystander model that offers a forum for student exploration and discussion". Through a peer-to-peer learning approach, MVP involves training peer mentors^[24] to deliver sessions. Williams and Neville (2017:4) highlight this as a strength:

The fact that 'mentors' are in the same social group as 'mentees' (i.e. high school pupils) is designed to qualify them as representative of prototypical group norms^[25], and therefore credible messengers of information regarding how to feel and act.

Peers are encouraged to discuss and consider their responses to the scenarios presented. In doing so, this discussion informs the participants of appropriate actions, while also empowering individuals to become "proactive bystanders in the face of GBV" (Williams and Neville, 2017:4). Katz (2018) highlights an example of a scenario with participants considering how women are objectified in the media. Deliberately provocative questions were asked about whether and/or how such objectification can lead to abuse or harassment. In this context, the "correct" answers were not provided as part of the materials, but rather the MVP programme created "space for dialogue that allows people to hear and express a range of viewpoints" (Katz, 2018:1755).

MVP programme evaluations have found positive results in changing pupils' attitudes and behaviours (see Powell, 2011; and MVP evaluations here). Predominantly evaluated within a North America context, MVP programmes have been shown to encourage participants to see forms of violence as being wrong and be more likely to take actions to intervene than students not exposed to the programme (Williams and Neville, 2017).

Within a Scottish context^[26], an evaluation was undertaken in three secondary schools using a version of the original MVP playbook and programme, adapted by the Violence Reduction Unit. Notably, this is the first peer-reviewed academic evaluation of the MVP programme in Europe (Williams and Neville, 2017: 7). Qualitative evidence from the evaluation of this programme suggests that the peer-learning element of MVP was a strength of the programme as it "overcame the taboo of 'snitching' (to teachers) through provision of a network of accessible senior students" and the peer-to-peer element resulted in the reinforcement of social group norms against GBV (Williams and Neville, 2017:23). However, William and Neville's (2017) qualitative study shows that while some male mentees said that their attitudes and behaviours had changed, female mentees felt that the boys' behaviours and attitudes had not changed following this year-long programme (Williams and Neville, 2017:19). However, the authors do not reflect in detail upon why this was the case.

Since 2017, the original MVP scenarios have been modified to reflect the language and culture of Scotland (Education Scotland, 2020). Consultations with young people and practitioners have led to the identification of new, relevant, topics for additional scenarios, and mentors^[27] are encouraged to use current media stories to enhance learning. In 2019 a new scenario on 'sexual harassment' was co-created with a group of Scottish young people.

Fixen's implementation science framework has been used by Education Scotland to guide the delivery of MVP and increase programme fidelity. This has led to the requirement for two core mandatory sessions to be delivered before any scenarios to allow exploration of gender norms and the link to violence.

As Williams and Neville (2017:29) note, the adoption of MVP programmes must involve evaluating the programme on an ongoing basis to "inform and update best practice and assess long term change". Their research also highlights the need to:

• conduct a process of continual development/refinement for MVP programmes and scenarios within it
• ensure age and cultural appropriateness
• embed MVP into participating school's curricula and cultures
• enact flexible approaches to developing the programme within participating schools

Beyond the MVP programme, two others that have been evaluated are Green DotandCoaching Boys into Men.

The Green Dot programme is a theory-based bystander approach programme through which male and female participants work together in the same training groups^[28]. Participating students are trained by Rape Crisis Centre trained educators to identify active bystander behaviours – to be taken by individuals or collectively. Educators worked with high school staff to identify student leaders to undertake intensive 5 hour bystander training.

Coker et al.'s (2019) longitudinal evaluation of the Green Dot programme in Kentucky (USA) high schools focused on whether it effectively reduced dating violence and sexual violence acceptance attitudes. An RCT of over 70,000 students over four years, found that this intervention was successful in reducing these forms of violence acceptance at both a school and an individual level. This evaluation did not measure violence as an explicit outcome, but it is acknowledged that "changes in norms may precede changes in actions (bystander behaviors) and changes in effect (violence)" (Coker et al. 2019:154)

The eleven week-long Coaching Boys into Men intervention involves 60 minute training for sports coaches, and brief weekly scripted discussions of 10-15 minutes with male athletes on ending dating violence. This programme has shown modest positive outcomes in reductions of negative bystander intervention behaviours and reducing abuse perpetration (Fulu et al. 2014; Miller et al. 2012, 2013) but also highlighted issues with implementation. However, at 12 month follow up the programme did not appear to have had any lasting effect.

Fulu et al. (2014:23) suggest that, based on these results, "a brief programme with few resources, utilising coaches as key influencers, may buffer against the initiation of dating violence perpetration during a critical developmental period for youth". However, the longer term impacts of this bystander programme are not as promising as the results at 3-month post-intervention.

Kettrey and Marx (2019: 213) conducted a systematic review and meta-analysis of the effects of fifteen studies of bystander programmes in preventing sexual assault across the college years. They note that peers are important in preventing violence, and that: "victims may trust their peers to provide a valuable source of support after an assault has occurred, but just as importantly, peers have the potential to play a pivotal role in the prevention of sexual assault by intervening when they witness its warning signs". Their results show that "bystander programs have a desirable effect on bystander efficacy, intentions, and interventions" (Kettrey and Marx, 2019: 223). However, there is no direct discussion about violence reduction as a direct outcome.

An overarching moderating factor is the understanding that: "the process of achieving behaviour change is complex, encompassing multiple levels or stages and requiring time. There is limited evidence that short one-off interventions have the capacity to change behaviour" (Public Health England, 2016:6).

Gainsbury et al. (2020:2) identify the following potential facilitators for bystander programmes within community contexts:

• Longer programmes which are cumulative, sequential and delivered over time by well-trained facilitators are more effective
• A wide range of teaching pedagogies including emphasis on role-play for skills acquisition and use of socio-culturally relevant materials
• Mixed-sex groups are also appropriate for bystander programmes^[29]

Key findings for service delivery and practice:

• Most evidence on bystander interventions come from middle schools in the USA, with some evidence from secondary schools in Scotland.
• They are premised on the idea that changing attitudes and perceived norms will change behaviour and decrease violence. For bystander behaviours to enhance safety, a person must go through all four stages of: noticing a behaviour or incident, recognising it as a problem, feeling responsible for intervening, and have the skills to act.
• Peer relationships have been shown to be helpful in reinforcing positive social norms, creating networks of support that are safe from the perception of "snitching", and making the intervention relevant and socially acceptable.
• It is important for services and practitioners to continually reflect and develop their content based on feedback and evaluation.
• Short, one-off interventions have less evidence to support them than more intensive engagement sustained over longer periods.

2.3 Domestic abuse disclosure schemes: inconclusive

Domestic abuse disclosure schemes have been adopted in England and Wales (2015), Scotland (2016) and New Zealand (2015)^[30]. These schemes provide potential victim/survivors of domestic abuse with the opportunity to ask about a new or existing partners' previous convictions.

Known as the Disclosure Scheme for Domestic Abuse Scotland (DSDAS):

DSDAS aims to tackle and prevent domestic abuse by enabling the public to request disclosure from the police if they suspect their current partner may have an abusive past. Requests can also be made, on their behalf, by a concerned family member, friend or neighbour (Police Scotland website, no date).

Within NHS Health Scotland's Domestic abuse: what health workers need to know about gender based violence (2019:43) they note that:

If a disclosure is deemed necessary, lawful and proportionate, the person potentially at risk, or person best placed to safeguard that information, will receive the information.

Police Scotland are required to conclude via the 3 point test^[31] that disclosure is necessary to protect the person at risk from being the victim of crime^[32]. At all times, the power to both share and/or disclose information must be considered on a case-by-case basis. In the case of disclosure, Police Scotland work closely with other agencies in a multi-agency approach to help and support the potential victim (NHS Health Scotland, 2019).

Currently, the evidence base about the effectiveness of Domestic Violence Disclosure Schemes (DVDS) is limited to pilot evaluations (see Home Office, 2013; New South Wales Government, 2016). There are currently no evaluations of the domestic abuse disclosure scheme currently available in Scotland (Brooks-Hays, 2018). Consequently these interventions have been classified as inconclusive due to insufficient evidence. However, it is still useful to consider the findings that are available on these schemes.

Hadjimatheou and Grace (2020:1) highlight that within England and Wales: "The DVDS has fast become established as a routine tool of domestic abuse safe guarding in England and Wales, with the number of disclosures made doubling from 3410 in the year ending March 2017 (Office of National Statistics 2017) to 6583 in the year ending March 2019 (Office of National Statistics 2019)". However, they also suggest that caution should be taken not to "conflate more frequent with better use of the scheme". Their findings suggest that there "is significant divergence both in disclosures themselves, and in practitioner views about what constitutes a fair and effective disclosure". They encourage awareness that all disclosures may not be equally effective or fair, citing an example of different disclosure experiences based on different geographies. Consequently, they call for a national systematic evaluation of the DVDS scheme ensuring that feedback from specialist case workers and survivors are included.

It is worth noting that some controversy exists around Domestic Abuse Disclosure schemes. As Brooks-Hays (2018:28) highlights, victim-focused initiatives such as this scheme are controversial, "not least since they do not guarantee victim safety in domestic abuse cases (Duggan, 2012) and may even have the effect of exacerbating the situation for living with violence (Fitz-Gibbon and Walklate, 2016)".

According to the Home Office pilot evaluation (2013:4), some effective practices and approaches were identified:

• practitioners highlighted the importance of having a safety plan^[33] in place following a disclosure
• practitioners and respondents receiving a disclosure also highlighted the importance of having a support worker attend a disclosure alongside the police, in order to give a potential victim immediate support
• Practitioners felt it was essential that there was sufficient support service coverage in place if the scheme was rolled-out more widely

This Home Office (2013) research also highlighted a number of lessons and recommendations for both policy and practice:

Box 1: Lessons and recommendations from Home Office (2013) DVDS pilot

Perceived bureaucracy of police process: police officers felt certain stages of the process were bureaucratic and lengthy, particularly conducting research on an individual's offending history.

Public awareness and understanding of the scheme: practitioners felt that public awareness of the disclosure scheme was low with some confusion about what the disclosure scheme was for and how the process worked (misunderstandings were resolved once the process was explained).

Frontline police officer awareness of the scheme: practitioners suggested that not all frontline police officers knew about the existence of the scheme and it was felt that a basic knowledge for all was useful.

Overlap between disclosure processes: some practitioners identified a need for further guidance about how the DVDS overlaps with and complements other disclosure processes, such as Multi-Agency Public Protection Arrangements and the Child Sex Offender Disclosure Scheme.

Lack of understanding of the term 'pressing need to disclose': practitioners involved in decision-making forums felt that the term "pressing need‟^[34] was unclear and subjective, but reported that this had been overcome in practice^[35].

Delivery of Right to Know disclosures: Police officers felt it was difficult to practically manage the delivery of a Right to Know disclosure. Support services were concerned that this could place a potential victim at greater risk of domestic abuse if not managed carefully.

Lack of consistency in information given in disclosures: There were differences between pilot areas in the level of detail contained within a disclosure and what previous offences were disclosed, achieving some level of consistency across areas was felt to be useful.

Follow-up support for non-disclosures: There was a lack of consistency between pilot areas in the type of follow-up support given to those who were told there was no information to disclose, a set of "minimum standards" of support to provide for nondisclosures was seen as useful.

Key findings for service delivery and practice:

• Services and practitioners should be aware of domestic abuse disclosure schemes, and where appropriate may play a role in suggesting a client make a request and supporting them to make that request, as well as supporting them afterwards whether a disclosure is made or not.
• Safety planning is essential when a disclosure is made.
• Having a support worker present in addition to police at the time of disclosure was considered important. Adequate resourcing should be available to ensure staff can support victims or potential victims both during and after a disclosure.