Evaluation of the good night out campaign: a sexual violence bystander training programme for nightlife workers in England

Sexual violence is global public health, human rights and gender equality issue. Sexual violence bystander programmes for nightlife workers are emerging across a few countries and further examination of such programmes is required. This exploratory study evaluates the potential effectiveness of the Good Night Out Campaign, a sexual violence bystander programme for nightlife workers. Two hundred and seven trainees attending the 1.5 hour training programme across two cities in England were recruited opportunistically, immediately prior to training delivery. Sexual violence myth acceptance and readiness and confidence to intervene in sexual violence were measured at baseline and post-intervention. Analyses used paired-sample tests to examine differences in the three measurements pre to post-training and effect sizes were quantified using Cohen’s d. Compared to pre-training, post-training participants were significantly (p < 0.001) less likely to agree with sexual violence myths, and more likely to be confident and ready to intervene in sexual violence or incidents of vulnerability. Effect sizes were small–medium. The study adds to emergent evidence suggesting that sexual violence bystander programmes may be promising in decreasing sexual violence myths and barriers to bystander intervention, and increasing willingness to intervene amongst nightlife workers. Findings can support the emergence of sexual violence prevention activities implemented in nightlife spaces. Further programme implementation and evaluation using experimental designs is needed to explore outcomes in greater depth, considering the complexity of the nightlife environment.


Background
Sexual violence is an issue of growing concern within global public health, human rights and gender equality fields (Home Office 2021; Montesanti and Thurston 2015;World Health Organization 2010;. Sexual violence is defined as "any sexual act, attempt to obtain a sexual act, unwanted sexual comments or advances, or acts to traffic, or otherwise directed, against a person's sexuality using coercion, by any person regardless of their relationship to the victim, in any setting" (World Health Organization 2010). The impacts of sexual violence can be immense, affecting an individual's health and well-being, and placing increased burdens on local services and communities (World Health Organization 2019). Across many countries there is increasing concern with regard to nightlife-related sexual violence, that can occur within nightlife settings as well as surrounding environments (e.g. public transport, fast-food establishments and the home) (Calafat et al. 2013;Graham et al. 2010Graham et al. , 2014Gunby et al. 2020;Kavanaugh 2013;Quigg et al. 2020). Whilst nightlife-related sexual violence can be experienced by all, the existing evidence demonstrates a clear sex divide with men most likely to be perpetrators and women most likely to be victims (Gunby et al. 2020;Quigg et al. 2020). Critically, evidence suggests that sexual violence within the night-time economy has been normalised; sexualised comments, groping and other forms of sexual violence and harassment have become key features of the lived experiences within nightlife spaces for many young adults (Christmas and Seymour 2014;Gunby et al. 2020;Nicholls 2017;Quigg et al. 2020Quigg et al. , 2021. Whilst there is increasing understanding and recognition of sexual violence both globally (WHO 2019) and across the UK (Home Office 2021), recognition within the nightlife environment may be complicated by various factors characteristic to this setting. Nightlife environments are places of leisure and relaxation, away from everyday stressors, and for some, a key setting for meeting romantic or sexual partners (Fileborn 2017). Alcohol and drug use can intensify the issue, reducing inhibitions and increasing risk of vulnerability (Abbey 2011;Orchowski et al. 2020), and can reduce bystanders awareness of the issue and intentions to intervene (Hust et al. 2019). Further, research demonstrates how problematic, and often gendered, stereotypes and myths (e.g. relating to how people dress, intoxication and gender roles) can lead to tolerance and acceptance of sexual violence (Gunby et al. 2020;Nicholls 2017;Waitt et al. 2011;Tinkler et al. 2018). This can prevent victims from reporting it (Tinkler et al. 2018) or others from intervening (Graham et al. 2010). It is imperative that sexual violence in the nighttime economy is challenged and addressed in order to counter this normalisation and acceptance of such behaviours (Montesanti and Thurston 2015).
Bystander intervention programmes are increasingly being identified as an effective strategy for preventing violence, including sexual violence, particularly in education settings (Banyard et al. 2007;Fenton et al. 2016). Bystander programmes primarily aim to address societal norms that facilitate violence, and promote positive and safe bystander intervention to prevent and respond to violence, shifting responsibility from the individual to the community (Fenton et al. 2016;Fileborn 2017). Berkowitz (2009) explains bystander intervention as being a process involving four stages: noticing the event, interpreting the event as a problem, having a sense of responsibility to intervene, and possessing the necessary skills and knowledge to intervene. Hence, a level of knowledge and understanding is needed across all stages in order for bystander intervention to be successful. Embedded in social norms theory, through engaging in and witnessing other's positive bystander intervention, the acceptability of sexual violence is anticipated to reduce and bystander intervention to increase. Despite strong evidence for bystander programmes in education settings, most evidence stems from the USA, and few studies have explored the impacts of bystander programmes delivered in the wider community, including nightlife settings (Fileborn 2017;Gainsbury et al. 2020;Quigg et al. 2020).
Building on existing evidence highlighting the potential of bystander training programmes in preventing sexual violence (Banyard et al. 2007;Burn 2009;Cares et al. 2014;Fenton et al. 2016), in recent years, such programmes have started to emerge across various nightlife settings (e.g. USA [Powers and Leili 2018]; Europe [Quigg et al. 2021]; New Zealand, [RespectEd Aotearoa 2021]). Whilst programme evaluation is scarce, recent evaluations of programmes implemented in the USA and Europe have found significant reductions in sexual violence myth acceptance, and an increased willingness and confidence to intervene amongst nightlife workers (Powers and Leili 2018;Quigg et al. 2021). However, further research is needed to understand the impact that such programmes can have in addressing the normalisation of sexual violence within nightlife, across programmes, nightlife settings and different countries. This exploratory study aims to build on this limited but emerging evidence and to our knowledge examines for the first time changes in nightlife workers readiness and confidence to intervene in sexual violence, and sexual violence myth acceptance, following participation in a nightlife worker bystander training programme (The Good Night Out Campaign [GNOC]) implemented in the United Kingdom (UK).

The intervention
In 2014, the GNOC (www. goodn ighto utcam paign. org/) was established to support nightlife spaces and organisations to better understand, respond to and prevent sexual violence, through provision of specialist training, policy support and an accreditation programme. The GNOC bystander training programme aims to support staff who work in bars, nightclubs, pubs and festivals (and other settings) to better understand, respond to and prevent sexual violence. The multi-component programme consists of policy support and provision of materials (e.g. posters) for venues, and an accredited in person training programme for staff (1.5 hours). In 2018/19, a train-the-trainer programme was developed by the GNOC team to enable local stakeholders (e.g. statutory/third sector partners) to implement the training across their local community (supported by a core GNOC programme team). In 2019/20, both statutory (local government) and third sector (i.e. sexual violence support services) organisations across two UK cities were trained by the GNOC team to implement the programme across licensed premises in their respective city's nightlife. Due to the high societal prevalence of sexual violence and the sensitivity of the issue, third sector organisations were those who had knowledge and experience of supporting victims and survivors of sexual violence, who could discuss the topic sensitively and provide direct support for trainees who may have experienced or witnessed sexual violence. During training implementation, GNOC staff were available (virtually) to provide ongoing advice and reflection on local stakeholder training delivery.
Across both cities, participating venues were recruited largely via direct contact between the local partners and/or other community contacts who raised awareness of the programme. Managers within each participating venue invited their staff to attend the GNOC training as part of the venue's wider staff training programme (it is not known if training participation was mandatory). The GNOC training programme was delivered to staff of mixed sex groups within 19 on-licensed premises (often during the evening, with an average of 11 staff per training session), using a range of learning techniques, including an interactive lecture (using PowerPoint slides/handouts), group discussion and scenario based activities. Key topics covered included: what do we mean by sexual harassment and assault; UK legislation relating to sexual violence; the 'whole team' approach to handling sexual violence, considering the core concepts of the bystander programme theory; removing barriers in reporting; and de-escalation techniques and responding appropriately. Group discussion points and scenario based activities were developed by the sexual violence specialist agency and local partners, and drew on real-life examples of sexual violence occurring within nightlife settings. Training concluded with details of available support services, with support available during or after the session from the sexual violence specialist agency facilitator.

Study design, participant recruitment and sample
The study was implemented across two UK cities with vibrant nightlife settings (i.e. more than 200 on-licensed premises) and large universities. Whilst local universities played no role in programme delivery, they are active in supporting multi-agency partnerships to prevent sexual violence across the community. At the start of each training session, trainers provided all trainees with a verbal description of the study, a participant information sheet and the opportunity to ask questions. All trainees consenting to take part were administered a paper-based survey pre and immediately post-training. Both surveys were designed to take no more than five minutes to complete, due to the limited opportunity to collect data. Across the two sites, all 206 trainees voluntarily consented to participate (City 1, n = 71; City 2, n = 135). Liverpool John Moores University provided ethical approval for the study and the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Participant characteristics
Socio-demographic variables included age, sex, job role and experience of working in nightlife (Table 1). Participant's initials and partial residential postcode were collected to link pre and post-test data for analyses. In the post survey, experience of sexual violence was also measured by asking: has anyone ever touched you sexually in a way that you did not want to be touched or done something else sexual to you that you did not want them to do, either on a night out or whilst working (with the option to select one or both settings).

Instruments
Due to data collection time constraints, the study used a shortened version of a non-validated tool used in a previous evaluation of a sexual violence bystander programme for nightlife workers (piloted across European nightlife settings in 2016/17 [Quigg et al. 2021]). The original tool aimed to measure three core concepts: sexual violence myth acceptance, and readiness and confidence to intervene in sexual violence, and was developed based on existing sexual violence scales, with items altered to reflect the nightlife setting and European context. In this study, in 2018/19, the evaluation team consulted with the GNOC programme deliverers and local trainers to review and revise the tool to ensure that it was relevant to the GNOC programme, and could be completed within the anticipated time available for participants to consider study participation and complete the survey prior to training delivery (around 10-15 min). Only items discussed in the GNOC training remained in the tool. Thus, for sexual violence myth acceptance, four items were removed (leaving three items); one item was removed from readiness to intervene (leaving two items); and, four items were removed from confidence to intervene (leaving four remaining).
Sexual violence myth acceptance: was measured using three items (Table 2), with participants asked to what extent they agreed/disagreed with each item (1, strongly agree to 5, strongly disagree). The average of the three item scores was used as an overall outcome measure score (with one item score reversed prior to inclusion in the combined score: sexual violence is never the fault of the victim). Cronbach's Alpha for the three items was 0.4 at pretest. Readiness to intervene: was measured using two items (Table 2), with participants asked to what extent they agreed/ disagreed with each item. The average of the two item scores was used as an overall outcome measure score (with one item score reversed: I do not think sexual violence is a problem in nightlife). Cronbach's Alpha was 0.2 at pretest.
Confidence to intervene: in situations where patrons may be vulnerable to sexual violence and/or are promoting sexual violence was measured using four items (Table 2), with participants asked how confident they thought they would feel in each scenario (1, really not confident to 5, really confident). An overall confidence to intervene outcome measure score was developed based on the average of the four items. Cronbach's Alpha was 0.8 at pretest.

Analyses
SPSS was used to undertake all analyses. Paired sample tests (Wilcoxon signed rank) were used to examine changes in the three areas measured pre-to-post training and post-hoc tests calculated the effect sizes (r), with Cohen's categorisation of effect sizes (small, 0.10; medium 0.30; large, 0.50) used to determine the magnitude of effect.
Compared to pre-training, post-training sexual violence myth acceptance scores were significantly lower (i.e. stronger agreement) for the item sexual violence is never the fault of the victim (p < 0.01), and higher (i.e. lower agreement) for the item unwanted sexual advances are a normal part of a night out (p < 0.01) ( Table 2). Overall, post-training, participants were significantly less likely to agree with sexual violence myths (mean average score: pre, 3.91; post, 4.09; p < 0.001 [small effect size, r = 0.2]). Post-training, Table 2 Sexual violence myths, and readiness and confidence to intervene, pre and post-training Responses ranged from: 1, strongly agree, to 5, strongly disagree; and 1, really not confident to 5, really confident. * Item scores reserved prior to inclusion in combined scores readiness scores were significantly lower (i.e. stronger agreement) for the item I think I can do something about sexual violence in nightlife (p < 0.001), and higher (i.e. lower agreement) for the item I do not think sexual violence is a problem in nightlife (p < 0.001). Overall, post-training, participants were significantly more likely to be ready to intervene (mean average score: pre, 2.09; post, 1.65; p < 0.001 [medium effect size, r = 0.4]). (Table 2). Scores for confidence to intervene increased significantly for most items (p < 0.01) (i.e. greater confidence) ( Table 2). Compared to pre-training, overall the post-training mean score was significantly higher, indicating greater confidence to intervene (mean average score: pre, 4.06; post, 4.26; p < 0.001 [small effect size, r = 0.2]).

Discussion
Emergent evidence suggests that bystander training programmes can be a key strategy to prevent violence, and promote norms that protect against violence (Fenton et al. 2016). The implementation of such programmes to prevent sexual violence has developed in recent years (Fenton et al. 2016), and extended from implementation in educational establishments to other community settings where sexual violence is a key concern (Gainsbury et al. 2020). More recently, sexual violence bystander training programmes for nightlife workers have been piloted in the USA (Powers and Leili 2018), Europe (Quigg et al. 2021) and New Zealand (RespectEd Aotearoa 2021). Our exploratory study adds to this emergent literature, and to our knowledge is the first UK study to examine the effectiveness of a sexual violence bystander programme implemented in the nightlife setting. Our finding suggests that a sexual violence bystander training programme delivered in nightlife has the potential to effectively address myths that can promote sexual violence and improve nightlife worker's readiness and confidence to intervene. Over the past decade there has been an emergence of literature demonstrating nightlife users experience of sexual violence in the night-time economy (Quigg et al. 2020), and an increased focus and recognition to prevent violence against women (and men) and implement effective solutions, including across public spaces (Home Office 2021; WHO 2010; 2019). Similar to our study, examination of sexual violence bystander training programmes for nightlife workers across the USA and Europe suggest that programmes can be effective in addressing risk factors for sexual violence (Powers and Leili 2018;Quigg et al. 2021). Critically, they have been shown to remove cultural and practical barriers to positive bystander intervention, by addressing social norms that promote sexual violence and teaching practical skills to safely intervene (Powers and Leili 2018;Quigg et al. 2021).
Our study found small effect sizes for sexual violence myth acceptance and confidence to intervene, and a medium effect size for readiness to intervene, and similar effects sizes have been found in previous pilot studies (Powers and Leili 2018;Quigg et al. 2021). Given the extent and normalisation of nightlife-related sexual violence (Quigg et al. 2020), collectively our study and previous studies provide strong support for implementing, and evaluating (through experimental designs) sexual violence bystander programmes in the nightlife environment (Powers and Leili 2018;Quigg et al. 2021). Powers and Leili (2018) argue the evidence behind bystander programmes are promising and that the adoption of campus-based interventions need to be expanded into wider communities and beyond the student community. They argue that nightlife bystander training programmes should be integrated with existing training that those working in the night-time economy already undertake (e.g. responsible service of alcohol). Bystander programmes implemented in nightlife settings are typically short in duration (e.g. 1.5-2 hours), in comparison to similar programmes implemented elsewhere (e.g. college/university) (Fenton et al. 2016;Powers and Leili 2018;Quigg et al. 2021). This, in part, is due to the limited opportunity to deliver training to staff and the need to frequently deliver training due to the transient nature of nightlife workers, with many staff working in the sector for short-term employment periods. Balancing the practicalities of training delivery, in an ever evolving setting with the need to deliver an impactful programme is an important consideration. Whilst this exploratory study found significant improvements in pro-social attitudes and behavioural intentions, evaluation of such programmes across college settings suggests that longer programmes have greater effects compared to shorter programmes (Jouriles et al. 2018). Future research should explore the feasibility, acceptability and impacts of programmes of varying lengths to identity the optimal and most pragmatic programmes for delivery within this complex setting (Fileborn 2017).
Changing cultures and preventing harms in evolving settings, such as the night-time economy, is complex, and the strongest evidence suggests that community-based multi-component programmes that address underlying risk factors and promote factors that protect people from harm are most effective (Jones et al. 2011). Addressing sexual violence across all levels of the socio-ecological model within the space of the night-time economy is vital. A wealth of studies illustrate that nightlife-related sexual violence is associated with a combination of factors at an individual, relationship, and community level (Quigg et al. 2020). It is vital that venues and wider nightlife settings are working collectively to prevent and address sexual violence, support victims and survivors to approach staff and peers for support, and enable nightlife workers to identity issues and safely intervene (Fileborn 2017). Critically, the impacts of bystander programmes (and other approaches) may be undermined by the gendered and sexualised nature of nightlife (Anitha et al. 2020); Griffin et al. 2012). Studies suggest that women are more likely to intervene than men (Hust et al. 2019), and exposure to objectifying images of women is associated with negative attitudes towards women and lower intentions to intervene in sexual violence (Hust et al. 2019).
Our findings should be considered in light of several limitations. Whilst our findings are comparable to other similar studies, reliability analysis of the combined attitudes and readiness scales showed low alpha scores meaning that these scales may not be a strong reliable measure. This is likely due to our low number of scale items, compared to a previous study using the same, but a wider suite of scale items (Quigg et al. 2021: myth acceptance α = 0.7; readiness α = 0.6). Future research should explore the inclusion of additional items to measure myth acceptance and readiness, with consideration of the data collection time constraints inherent within the nightlife setting. Whilst our exploratory study was not designed to be experimental or to have a control group, this should be considered for future studies. Finally, venues volunteered to host the training for their staff and whilst not mandated, staff were expected to attend the training, meaning that staff are likely working in an environment supportive of preventing sexual violence (even prior to training delivery). Further testing of the intervention across a wider range of venues is warranted.

Conclusion
Our study supports the limited but emerging evidence suggesting that sexual violence bystander programmes can be implemented in nightlife settings and are promising in promoting factors that can prevent sexual violence. The GNOC bystander training programme, and similar programmes, can play an important role in addressing sexual violence and preventing violence against women (and men). Enhancing bystander approaches to preventing sexual violence in nightlife will likely support, and be complemented by, wider efforts to address sexual violence and systemic violence against women and girls (Home Office 2021; WHO 2019). However, further research is required to understand the feasibility and acceptability of implementation within an evolving and complex nightlife setting, and to identify if effects can be sustained over time, and if positive changes in pro-social attitudes and behavioural intentions result in reductions in sexual violence.