Abstract
This article seeks to understand the experiences of bystanders to domestic violence and abuse (DVA) during the COVID-19 pandemic in Wales. Globally, professionals voiced concern over the COVID-19 restrictions exacerbating conditions for DVA to occur. Yet evidence suggests this also increased opportunities for bystanders to become aware of DVA and take action against it. This mixed methods study consists of a quantitative online survey and follow-up interviews with survey respondents. Conducted in Wales, UK, during a national lockdown in 2021, this article reports on the experiences of 186 bystanders to DVA during the pandemic.
Results suggest that bystanders had increased opportunity to become aware of DVA due to the pandemic restrictions. Results support the bystander situational model whereby respondents have to become aware of the behaviour, recognise it as a problem, feel that they possess the correct skills, and have confidence in their skills, before they will take action. Having received bystander training was a significant predictor variable in bystanders taking action against DVA; this is an important finding that should be utilised to upskill general members of the community.
Key messages
- •The COVID-19 pandemic and associated restrictions had allowed people to become aware of DVA.
- •Participants’ experiences of witnessing or having concerns about DVA and intervening had a negative impact upon their wellbeing, yet most would not have done anything differently.
- •When people have knowledge and skills to intervene, most will act as prosocial bystanders when they witness DVA, therefore bystander training for DVA should link to public awareness campaigns to enable people to act safely.
Introduction
Domestic violence and abuse (DVA) is a major human rights, criminal justice and public health issue. It is a significant cause of inequality and ill-health, and has adverse psychological, economic and social impacts on individuals, families and communities (World Health Organization, 2021a). Preventing DVA can improve the health and wellbeing of individuals and communities, which can have a wider positive impact for the economy and society (NICE, 2014).
Estimates across England and Wales, suggest that 2.4 million adults experienced DVA in the year ending March 2022 (ONS, 2022). Between March 2018 and 2019, Welsh police forces recorded 80,924 DVA related incidents (ONS, 2020), yet this is likely to be a fraction of incidents, as DVA often goes unreported.
COVID-19 and DVA
During the COVID-19 pandemic, countries implemented measures to reduce the spread of the virus. In Wales, restrictions included a stay-at-home policy, self-isolation, social distancing, and the closure of most retail outlets and public spaces. While these measures were intended to keep the population safe, home was not a safe space for everyone, including victims of DVA (Campbell, 2020).
Throughout the pandemic, experts voiced concern that restrictions exacerbated conditions for DVA (WHO, 2021b). The restrictions forced victims to stay at home, for extended periods of time, with their abusers. Victims may have been unable to obtain support (both formal and informal), and had limited access to money, the internet and their phones (Kofman and Garfin, 2020; Sacco et al, 2020; Speed et al, 2020).
In Wales, as with countries around the world, helplines reported dramatic increases in the number of contacts from third parties (friends, family, neighbours and colleagues) looking for advice and support regarding someone they were concerned about (Ivandic et al, 2020; ONS, 2020). This suggests that different groups of people may have had new opportunities to notice DVA as a result of the restrictions.
Bystanders
Bystanders are ‘witnesses to negative behaviour (an emergency, crime, rule violating behaviour) who, by their presence, have the opportunity to step in to provide help, contribute to the negative behaviour, encourage it, or stand by and do nothing but observe’ (Banyard, 2015: 8). Mobilising bystanders who are willing and able to help within their communities is an effective strategy to prevent violence against women and has been a research focus for decades. These studies emanate overwhelmingly from the US, conducted predominantly in university settings, with some preliminary studies in the UK (Fenton and Mott, 2018; Bovill and White, 2020; Roberts and Marsh, 2022). Recent systematic reviews and meta-analyses indicate an upsurge in research across this field, including randomised control trials and quasi-experimental designs, indicating improvements across a range of measures (Jouriles et al, 2018; Kettrey and Marx, 2018; Mujal et al, 2019; Addis and Snowdon, 2021; Wong et al, 2021). Some studies have found complex interactions between racial identity, gender and year of study (Brown et al, 2014), however studies have primarily been carried out with white student samples, and race remains understudied both in campus and community settings (Banyard et al, 2020). It is within this context of evaluating bystander training in universities that much of the research into the experiences and behaviours of bystanders has been conducted.
Bystander programming is based on the idea that everyone in a community or peer group has the potential to help when they witness problematic behaviours and by ensuring that helping becomes the peer group or community norm will, over time, result in the social unacceptability of the non-desirable behaviour. The process to action will be influenced by many different intrapersonal, situational and social determinants, with barriers present at each step (Burn, 2009; Banyard, 2011; McMahon, 2015).
Latané and Darley’s (1968; 1969) theoretical organising framework, applied by Banyard (2011), Berkowitz (2009) and Burn (2009) to sexual assault situations, posits that for bystanders to take action, a number of cognitive and behavioural processes must occur at the individual level (Berkowitz, 2009). First, the bystander must notice the event, requiring knowledge and awareness. Second, they must interpret and recognise the event as a problem meriting intervention (Burn, 2009). Third, they must feel a sense of responsibility and motivation to act (McMahon, 2015; Bennett et al, 2017; Rothman et al, 2019). Personal attitudes and beliefs which minimise violence, such as rape myth acceptance, are likely to reduce responsibility-taking and are associated with lower likelihood of intervention (Banyard, 2011). Attitudes to victims such as victim-blaming may influence perceptions of victim worthiness of help (Pagliaro et al, 2020). Fourth, the bystander must possess the skills to act, and in the final step, take action and perform a bystander behaviour. Bystanders must have confidence in their ability and skills to intervene safely (Burn, 2009). Moving through these stages is related to increased readiness, intention and confidence to help (Banyard et al, 2014; Jouriles et al, 2018; Mujal et al, 2019).
Intrapersonal characteristics such as gender and age also influence intervention likelihood, as does relationship to the victim. Women and girls are more likely to intervene in sexual violence and more likely to help victims generally (Banyard, 2011; Rothman et al, 2019). Men are more likely to intervene when the situation is deemed an ‘emergency’ (Burn, 2009). Having a relationship with the victim is associated with helping in the general bystander literature (Levine et al, 2002), and in Burn’s (2009) findings, but not by Banyard’s (2008) study. Studies have also found that bystanders are much more likely to take prosocial action if they themselves had been a victim (Christensen and Harris, 2019) and when they perceived the behaviour to be life threatening (Fleming and Wiersma-Mosley, 2015).
Environmental factors beyond the individual-level influence an individual’s bystander decision-making (McMahon, 2015). Social influence may impede intervention, such as the bystander being unsure whether there is a problem based on others’ reactions (Latané and Darley, 1969). McMahon (2015) reports that the more norm-violating a behaviour or incident is, the more likelihood of intervention. Rothman et al (2019) found that high school students’ bystander behaviours were influenced by perceptions of how others behaved in their community, particularly if there was strong community-cohesiveness.
Another environmental factor is sense of community. In the wider bystander literature, factors including social cohesion and connection, commitment to neighbourhood and involvement in the community are connected to higher likelihood of intervening in relation to crimes (McMahon, 2015; Rothman et al, 2019). In the context of intimate partner violence (IPV), positive bystander behaviours were connected to a higher sense of collective efficacy on the part of young adults in rural communities (Edwards et al, 2014). In the general community, Banyard et al (2020) found that prosocial bystanders had a significantly higher sense of community than passive bystanders.
Outside of formal education settings there is evidence that bystanders may be in a position to help. Hamby et al (2016) found bystanders to be present at around two thirds of incidents of victimisation, while Taylor et al (2019) found up to a third of DVA incidents may be witnessed. Yet, capturing the experiences of the informal supporters is rarely explored and comparatively little is known about bystander action in the general community.
Frye et al’s (2012) concept mapping study found that neighbourhood bystanders in two US urban areas considered intervention with a range of actions geared towards the victim, perpetrator and community to be feasible. Actions focused on victims and formal and semiformal systems, were rated as most feasible and actions focused on the abuser were least feasible with community-focused actions slightly higher. Participants viewed connecting the victim with formal systems was perceived to be the most effective.
In Weitzman et al’s (2020) US study of actual experiences with a nationally representative sample, just over half had known of a victim of IPV and they were most likely (in order) to be friends, family or acquaintances, with women having higher odds than men of knowing victims. Of the overall sample, a quarter had intervened for IPV, with this rising to just over one half of respondents who had known a victim of IPV. The relationship with the victim was shown to be important, with 70 per cent lower odds of intervening for an acquaintance than a family member. No demographic differences were found for IPV intervention. Victims of DVA will often seek informal support before reaching out more formally (Ansara and Hindin, 2010), so it is important that friends and family are the people whom bystanders are most likely to help. The intervention strategies most commonly adopted in Weitzman et al’s (2020) study (in order) were offering safe haven, offering sympathy to the victim and telling the abuser to stop. Weitzman’s study did not examine how the bystander became aware of the IPV or what types of IPV were noticed and intervened upon and thus how intervention strategies might differ according to the context of abuse remains unexplored. The study did explore barriers to intervention, albeit hypothetically, finding that the most perceived barrier was fear of physical injury (almost half of respondents) with women and Black respondents having much higher odds of reporting this fear than men and White respondents. Women were also less likely to physically intervene or tell the abuser to stop and were also less likely to report perceiving IPV as a private matter as a barrier.
Taylor et al’s (2019) study with a rural US sample explored the experiences of bystanding from the perspective of IPV victims as opposed to bystanders themselves. This study explored five categories of IPV, four of which were physical, and the fifth being threats of harm. They found the highest reports of bystander helpfulness were for being ‘pushed, grabbed or shook’, yet victims reported higher rates of injury when bystanders were present for being pushed or grabbed and being hit by a partner, and higher rates of victim injury when the bystander was also harmed or threatened. The authors suggest that bystanders may have become aware and present due to the physical seriousness of the IPV incidents measured.
In Storer et al’s (2021) study, young racially minoritised adults outside of formal education or employment settings in urban communities overwhelmingly expressed disinclination to use bystander behaviours in dating and community violence, citing fear for their own safety, and norms which equated intervention as ‘snitching.’ Consistent with Weitzman et al’s and Frye et al’s findings, extreme physical dating violence was, however, deemed more intervention-worthy, as was proximity of relationship to the victim. While the study explored hypothetical not actual bystander behaviours, it is important in recognising that the situational model may operate differently for ethnic minorities.
Bystander training might be utilised to improve informal community-based responses to DVA, and there is preliminary evidence of effectiveness of bystander training in a UK context (Gainsbury et al, 2020). The potential of bystander community training programmes depends on furthering our understanding of bystander experiences in DVA. The current study sought to add to the literature exploring in detail who bystanders might be, the behaviours they witness, their responses and their motivations and barriers to intervening. Bystander and victim safety is also of paramount concern for developing training programmes. In Taylor’s (2019) study, a fifth of victims reported that bystanders were harmed or threatened, and this was also related to poorer victim outcomes. Given that we know relatively little about the impact on bystanders’ own wellbeing, the current study sought to add to the literature by exploring the impact of intervening on the bystanders.
Study aims
This study piloted a mixed-methods design using survey and interview techniques to explore experiences and behaviours of bystanders to DVA in Wales during the COVID-19 pandemic. The following research questions were posed:
- 1.What are bystanders’ experiences of DVA during the COVID-19 pandemic?
- 2.What are the motivations and barriers for bystanders taking action to prevent DVA during the COVID-19 pandemic?
- 3.What is the impact on the bystanders and what support do they need?
Methods
Participants and recruitment
The survey was conducted via the online platform Qualtrics and was open from 15 February 2021 to 8 March 2021: a 21-day window during a national lockdown period in Wales. Those aged 18 years or over and either residing or working in Wales during the pandemic were eligible to participate. Recruitment advertisements targeted individuals who had seen or become concerned about DVA, or warning signs, since the beginning of lockdown restrictions in Wales (March 2020). The term ‘bystander’ was not used in the advertisements as it was considered that this would not be widely understood among the public. All study materials were available in both Welsh and English and participants could choose to participate in either language.
The survey advertisement was disseminated via email and social media via stakeholders including the Welsh government, health boards, police forces, local authorities, specialist domestic abuse and sexual violence services, housing organisations, higher education institutes, transport organisations and care organisations. Further, an advertising company was commissioned to help disseminate the survey through paid advertisements on social media and coverage in online news.
Individuals accessing the survey were provided with an overview of the purpose and nature of the study and provided informed consent before proceeding to survey questions. After completing the survey, participants were invited to email the research team to participate in the interviews.
A total of 395 survey responses were received. For this study, data were restricted to those participants who reported having witnessed DVA during the pandemic (47%; n=186). Six survey respondents volunteered to take part in an additional interview. Of these, three were excluded from the study as two were survivors of DVA rather than bystanders and one included experience only in their professional capacity as a domestic abuse support worker. The three remaining bystanders were women who had become concerned about a friend (two participants) or a parent (one participant) during the pandemic.
Measures
The survey, designed by the research team, drew upon available literature and criminal law. The survey consisted mostly of questions from validated surveys such as the Crime Survey for England and Wales (ONS, 2020) and US Campus Climate Surveys (Cantor et al, 2020). However, at the time of delivery there were no validated surveys on bystander experiences during the pandemic; therefore, some questions were developed or adapted by the research team. The final survey was not validated but was discussed with an expert advisory group and tested with colleagues, then edited based on feedback.
The survey began with demographic questions, followed by questions on sense of community and knowledge of DVA. The sense of community question was taken from Peterson et al (2008) and asked, ‘To what extent do you agree with the following sentence: I want to help members of my community?’ Responses were on a five-point Likert scale, from ‘strongly agree’ to ‘strongly disagree.’ For analysis, responses were grouped into ‘agree’, ‘neither agree nor disagree’ and ‘disagree.’ The knowledge of DVA question asked, ‘How knowledgeable are you about domestic violence and abuse?’ Responses were again on a five-point Likert scale, from ‘extremely knowledgeable’ to ‘not knowledgeable at all’.
The survey then asked, ‘Since the pandemic began, have you noticed or become concerned about any of the following behaviours in relationships?’ These behaviours, included in the survey, were taken from AAU Campus Climate Survey (Cantor et al, 2020), and the Crime Survey for England and Wales (ONS, 2020). Survey respondents were also asked ‘Which, if any, of the following actions, however small, did you take since the pandemic began in response to the behaviour you had seen?’, followed by ‘Why did you take action?’ or ‘Why did you not take action?’
The survey also asked specific questions about the person’s status during lockdown, including ‘Since the pandemic began, which of the following applied to you?’ Response options included ‘working from home’, ‘furloughed’ or ‘continuing as normal.’ The survey also asked, ‘To what extent to you agree with the following statement: I feel more connected to my community or neighbourhood since the pandemic began?’ This was rated on the same five-point Likert scale as the previous community question. Last, respondents were asked ‘How likely is it that the circumstances of the pandemic influenced you being able to witness this behaviour?’ Responses were ranked on a five-point Likert scale, from ‘extremely likely’ to ‘extremely unlikely’.
A focus of the interviews was to capture the impact of the experience on the bystander. Written consent was obtained for these interviews through email with the participants. Due to COVID-19 restrictions, the interviews were conducted online, through Microsoft Teams or Zoom. The audio recordings were transcribed verbatim.
Data analysis
Survey data were analysed using IBM SPSS Statistics V24. Descriptive analyses used chi squared with Fisher-Freeman-Halton Exact Test used where expected counts were below five (analyses performed in SPSS V29). Multivariate analyses used binary logistic regression (enter method).
Three interviews were not considered sufficient for thorough qualitative analysis; therefore, they have been used to offer additional insight into the experiences highlighted within the survey data.
Ethical approval
Ethics approval was obtained from Health Research Authority and Health and Care Research Wales (ref. 20/HCRW/0061). The contact details for Live Fear Free, a Welsh domestic abuse helpline, was provided throughout the survey for participants needing support or advice. Similarly, if the participant felt that someone was in immediate danger, they were encouraged to call 999.
Results
Sample demographics and traits
The 186 participants were aged between 18 and 74, with the majority being women (85%) and of White British/Irish (96%) ethnicity. Most respondents worked in the following sectors: industrial work and other tertiary jobs (for example, hairdressers and postal workers) (24%); health and social care (22%); local authority, government, or other key public services (21%); and education (17%); the remaining participants (16%) were retired, unemployed or students. Three quarters of respondents (76%) had been primarily home-based during the pandemic, whether that be working from home, furloughed, retired or unemployed.
The sample had a high self-reported level of knowledge of DVA, with 64 per cent of participants reporting that they were very or extremely knowledgeable. Almost half (48%) reported having completed some form of DVA training in the past five years (the survey did not ask what type of training survey respondents had received). Eighty per cent of respondents agreed or strongly agreed with the statement ‘I want to help members of my community’, while 45 per cent agreed or strongly agreed ‘I feel more connected to my community since the pandemic began’ (Table 1).
Bystander circumstances and the proportion taking action after witnessing DVA
| Sample | Took action (Yes) | ||||||
|---|---|---|---|---|---|---|---|
| n | % | n | % | X2 | P | ||
| All | 186 | 100 | 164 | 88 | |||
| Gender of bystander | Man | 28 | 15 | 24 | 86 | ||
| Woman | 158 | 85 | 134 | 89 | 0.191 | 0.750 | |
| Age of bystander (years) | 18–34 | 69 | 37 | 56 | 81 | ||
| 35–54 | 98 | 53 | 90 | 92 | |||
| 55–74 | 19 | 10 | 18 | 95 | 4.689 | 0.087 | |
| Behaviour witnessed | Coercive control | 122 | 66 | 106 | 87 | ||
| Warning signs | 15 | 8 | 11 | 73 | |||
| Vulnerability | 6 | 3 | 6 | 100 | |||
| Threats | 13 | 7 | 13 | 100 | |||
| Physical | 23 | 13 | 23 | 92 | |||
| Sexual | 5 | 3 | 5 | 100 | 6.931 | 0.201 | |
| How did you initially come to witness/know about the behaviour? | Physically in person | 75 | 40 | 65 | 87 | ||
| Told by victim | 67 | 36 | 64 | 96 | |||
| Told by someone else | 23 | 12 | 21 | 91 | |||
| Do not want to answer | 15 | 8 | 9 | 60 | |||
| Online | 6 | 3 | 5 | 83 | 13.037 | 0.006** | |
| Relationship between victim and perpetrator | Family members | 38 | 20 | 33 | 87 | ||
| Intimate or ex-partners | 139 | 74 | 124 | 90 | |||
| Unsure | 9 | 5 | 7 | 70 | 3.565 | 0.143 | |
| Gender of victim | Man | 28 | 15 | 25 | 89 | ||
| Woman | 152 | 82 | 136 | 89 | |||
| Unsure | 6 | 3 | 3 | 50 | 6.363 | 0.035* | |
| Gender of perpetrator | Man | 145 | 78 | 131 | 90 | ||
| Woman | 35 | 19 | 31 | 89 | |||
| Unsure | 6 | 3 | 2 | 33 | 11.334 | 0.003** | |
| Did anyone else witness or know about the behaviour? | Yes | 106 | 57 | 102 | 96 | ||
| No | 39 | 21 | 34 | 87 | |||
| Unsure | 41 | 22 | 28 | 68 | 19.887 | <0.001** | |
| Have you done DVA training in the past 5 years? | Yes | 89 | 48 | 84 | 94 | ||
| No | 97 | 52 | 80 | 82 | 6.311 | 0.012* | |
| Want to help members of community | Agree | 149 | 80 | 133 | 89 | ||
| Neither agree/disagree | 33 | 18 | 27 | 82 | |||
| Disagree | 4 | 2 | 4 | 100 | 1.641 | 0.395 | |
| Feel more connected to community since pandemic | Agree | 84 | 45 | 75 | 89 | ||
| Neither agree/disagree | 44 | 24 | 39 | 89 | |||
| Disagree | 58 | 31 | 50 | 86 | 0.324 | 0.842 | |
Note: DVA = domestic violence and abuse *p<.05 **p<.01.
DVA witnessed
Participants had witnessed or become concerned about a range of DVA behaviours since the pandemic began. These included warning signs for DVA (for example, someone behaving worried and fearful all of the time), coercive control, abuse of a vulnerable person, verbal abuse of a LGBTQI+ person for their sexuality, threats of abuse, actual physical abuse and sexual abuse. Coercive control was the most witnessed DVA behaviour, reported by 90 per cent of participants, twice the amount of physical abuse (45%), followed by warning signs (71%) (Figure 1). Three quarters (77%) of respondents reported having witnessed more than one category of DVA behaviour.
Proportion of participants reporting witnessing each DVA behaviour category during the pandemic
Citation: Journal of Gender-Based Violence 8, 2; 10.1332/23986808Y2024D000000020
‘There was a definite change in my friend’s ability to be able to talk freely… We felt that there was a lot of controlling behaviours and isolation tactics really, trying to keep her away from friends, family, he had become imprinted in every aspect of her life.’ (Interview 2)
‘She was quite fearful of the pandemic, he was using that fear to keep her in the house more, to control her more.’ (Interview 1)
Survey participants were asked to provide further information on DVA behaviour they had witnessed during the pandemic, with those who had witnessed more than one form asked to select a specific behaviour to report on (see Table 1). The most common behaviour reported on was coercive control (66%), followed by physical abuse (13%), warning signs (8%), threats (7%), abuse of a vulnerable person (3%) and sexual abuse (3%). Forty per cent had witnessed the behaviour in person, while 36 per cent had been told about it by the victim. Three quarters of respondents (74%) said the incident took place within an intimate relationship. The majority of victims were women (82%) and the majority of perpetrators were men (78%). Over half (57%) of respondents said another person (besides themselves) had witnessed or knew about the behaviour.
‘I think it would have been more easily hidden or we might have been distracted from it and we might not have been as proactive or as aware and worried about it if we weren’t in a pandemic.’ (Interview 3)
However, a Chi-squared test found no difference between those who were primarily at home during the pandemic and those who were going out to work as normal in becoming aware of DVA since the pandemic began.
Actions taken
Survey participants were asked what action, if any, they had taken in response to the DVA behaviour they had witnessed. Most respondents (88%) reported some form of action, the most common relating to supporting the victim (56%), unofficially sharing concerns (for example, with family and friends) (44%), and looking for more information (41%) (Figure 2). Around 60 per cent of participants reported more than one type of action.
Proportion of participants reporting taking each action type after witnessing DVA during the pandemic
Citation: Journal of Gender-Based Violence 8, 2; 10.1332/23986808Y2024D000000020
The proportion of respondents taking action did not differ by gender, and while fewer younger respondents reported taking action this difference was not significant (Table 1). All those who reported witnessing abuse of a vulnerable person, threats of abuse or sexual abuse had taken action in response to the behaviour, 92 per cent for physical abuse, 87 per cent for coercive control and 73 per cent for warning signs. Taking action was highest among participants who reported having been told about the DVA by the victim (96%), while there was no relationship between taking action and type of relationship between perpetrator and victim (intimate or ex-partner, or family member). Bystanders were less likely to have taken action when the gender of the victim or perpetrator was unknown (Table 1), and more likely to take action when they believed someone else knew about the DVA. They were also more likely to have taken action if they had attended DVA training in the past five years (94%, versus 82% of those who had not received DVA, p=0.012) (Table 1). There was no association between taking action and responses to statements about wanting to help community members or feeling more connected to the community since the pandemic.
Significant variables were entered into a binary logistic regression (enter method) model to examine independent effects on taking action. The gender of victims and perpetrators was not included due to no differences being seen between men and women and very low numbers in the ‘unsure gender’ category. Taking action was found to be independently associated with having attended DVA training in the past five years (AOR 4.25) (Table 2). Odds of taking action were also increased in those who reported that someone else had witnessed or knew about the DVA, and those who had been told about the DVA by the victim (versus witnessing it in person). Significance levels are likely to have been affected by the low sample size (Table 2).
Adjusted odds ratios for taking action
| Bystander took action | |||||
|---|---|---|---|---|---|
| AOR | 95% CIs | p | |||
| How did you initially come to be witness/ know about the behaviour? | Physically in person | Ref | 0.117 | ||
| Told by victim | 4.53 | 0.88 | 23.42 | 0.072 | |
| Told by someone else | 0.61 | 0.10 | 3.80 | 0.596 | |
| Do not want to answer | 0.45 | 0.10 | 1.97 | 0.291 | |
| Online | 0.22 | 0.02 | 2.80 | 0.245 | |
| Did anyone else witness or know about the behaviour? | No | Ref | 0.002** | ||
| Yes | 4.52 | 0.95 | 21.59 | 0.058 | |
| Unsure | 0.33 | 0.09 | 1.18 | 0.089 | |
| Have you attended DVA training in the past 5 years? | No | Ref | |||
| Yes | 4.25 | 1.13 | 15.96 | 0.032* | |
Notes: *p<.05**p<.01; CI = Confidence Interval; Ref = Reference Category.
Motivations and barriers to taking action
Participants who took action were asked to select from a list of possible motivations. The most common motivations related to feeling responsible (75%), recognising the situation as problematic (73%), personal reasons (49%) and possessing the right skills/feeling supported (37%). There were no significant differences between male and female respondents in being motivated to take action through reasons relating to having the right skills (Table 3). However, having completed DVA training in the last five years was found to have a significant association with the bystander feeling that they possessed the correct skills to respond (p<.001, Table 3). The proportion reporting skills-based motivations was highest among those who reported on an incident of physical abuse (65%) and was higher among those who agreed that they wanted to help members of their community and felt more connected to their community since the pandemic began (Table 3).
Proportion of bystanders who took action who reported being motivated by feeling that they possessed the skills
| Sample | Skills-based motivation (Yes) | |||||
|---|---|---|---|---|---|---|
| n | n | % | X2 | p | ||
| All | 163 | 60 | 37 | |||
| Gender of bystander | Man | 24 | 6 | 25 | ||
| Woman | 139 | 54 | 39 | 1.688 | 0.254 | |
| Age of bystander (years) | 18–34 | 55 | 24 | 44 | ||
| 35–54 | 90 | 30 | 33 | |||
| 55–74 | 18 | 6 | 33 | 1.656 | 0.447 | |
| Behaviour witnessed | Coercive control | 105 | 33 | 31 | ||
| Warning signs | 11 | 3 | 27 | |||
| Vulnerability | 6 | 3 | 50 | |||
| Threats | 13 | 5 | 38 | |||
| Physical | 23 | 15 | 65 | |||
| Sexual | 5 | 1 | 20 | 10.346 | 0.054 | |
| Have you done DVA training in the past 5 years? | Yes | 84 | 43 | 51 | ||
| No | 79 | 17 | 22 | 15.409 | <0.001** | |
| Want to help members of community | Agree | 132 | 55 | 42 | ||
| Neither agree/disagree | 27 | 4 | 15 | |||
| Disagree | 4 | 1 | 25 | 7.465 | 0.018* | |
| Feel more connected to community since pandemic | Agree | 75 | 37 | 49 | ||
| Neither agree/disagree | 39 | 13 | 33 | |||
| Disagree | 49 | 10 | 20 | 10.994 | 0.004** | |
Notes: DVA = domestic violence and abuse; *p<.05 **p<.01.
‘I feel incredibly impotent… unless I report it to the police… there’s nothing I can really do.’ (Interview 1)
‘If you say the wrong thing to them, it can have the adverse effect to what you’re trying to do so you have got to bite your tongue and be so careful with what you say and do.’ (Interview 2)
Impact
‘It has played on my mind a lot, second guessing myself, did I say the right thing? Did I push enough? Should I have pushed more?… It has been a lot of questioning myself.’ (Interview 3)
Three quarters of survey respondents (121 out of 161 who answered this question) indicated that they felt that having some form of DVA bystander training would be helpful.
Discussion
This study sought to explore the experiences and behaviours of actual bystanders to DVA in the general public during the COVID-19 pandemic using a survey and interviews. While implemented on a small scale, this study was the first of its kind, and provides new insights into bystanders’ experiences during a global pandemic. The actions of bystanders in the general public have thus far been understudied and we sought to add to the literature by conducting a study which examines actual bystander behaviours as opposed to hypotheticals, explores intervention from the bystander’s perspective rather than the victim’s, measures a larger range of DVA behaviours, and includes action taken, barriers, motivations and the outcomes for the bystander themselves. The discussion explores the learning from the study and considers its relevance to DVA prevention in general, and in COVID-19 recovery and future public health emergencies.
Recruitment
One of the intrinsic difficulties encountered in this study was how to recruit participants to a study about ‘domestic abuse’ without using the words ‘domestic abuse’, in recognition of the fact that many people may be unable to identify behaviours that they witness as being ‘domestic abuse’ – and thus would not take part. Given the multitude of behaviours making up ‘domestic abuse’, it was not feasible to advertise the study based on descriptors of these and so ultimately the words ‘domestic abuse’ were used. This may offer explanation as to why the majority of survey respondents self-reported having a high knowledge of DVA, as only those with the knowledge would be able to recognise the behaviours witnessed and subsequently know that this survey was aimed at them. Thus, it is likely that we were unable to capture the behaviours witnessed by people who did not identify or categorise such behaviour(s) as domestic abuse. Future research is required to examine the optimal methods for advertising and recruiting the general public who witness behaviours which would constitute domestic abuse.
Noticing DVA
The study reveals that participants had passed through the first stage of the situational model, namely having the knowledge and awareness to notice behaviour. This study found that the pandemic-enforced health protection measures increased people’s opportunity to become aware of concerning behaviours. In the literature, having opportunity is crucial: evaluations of bystander training programmes have often struggled to capture interventions made post-training within survey follow-up periods because in real life timely opportunities to intervene must present themselves, and so survey responses must be screened for opportunity (Banyard et al, 2020). While campus bystander training has often focused on peer leaders, future research in communities might usefully consider how to ‘replicate’ or capitalise on lockdown conditions: those who are more likely to stay at, or work from, home could be usefully targeted for training.
This study provides useful information on the types of behaviours witnessed when general opportunity is present. Consistent with expert concerns that the pandemic may have allowed perpetrators to fully control the social lives and means of correspondence of victims/survivors (Bradbury-Jones and Isham, 2020), the behaviours about which participants had most commonly become concerned were warning signs of DVA and coercive control. By developing a more expansive range of behaviours in our survey design than previously used in other studies, focusing on behaviours and not situations, and extending them beyond physical violence – which is more immediately perceivable as ‘high risk’ (Taylor et al, 2019; Weitzman et al, 2020; McInnes, 2022) – it is an interesting addition to the literature that the most commonly noticed behaviours are those which are more nuanced forms of DVA. This may simply speak to our self-selecting knowledgeable sample, but it might suggest that bystanders are noticing a wide range of behaviours and further research should explore this with representative samples. The positive recognition of coercive control, criminalised only relatively recently in England and Wales (under s.76 of the Serious Crime Act of 2015), may indicate that dissemination and messaging about the offence/behaviour has been received by some members of the public.
That most respondents were women fits well with Burn’s (2009) hypothesis that women may have heightened awareness of risk because it is more salient to them as women and other gendered characteristics which situate them as more relationally focused. It is also consistent with the literature that suggests that women are more likely to know victims (Weitzman et al, 2020).
Although just under half of respondents reported having domestic abuse training in the past five years, the ability to notice behaviours in our sample does not appear to be based on prior training as there was no difference in the behaviours noticed between those who had training and those who had not. Perhaps participants were able to recognise the behaviours due to their self-reported, good knowledge of domestic abuse, which is consistent with other studies. It remains unclear where participants obtained this knowledge and future surveys should examine this further.
Taking action
The majority of participants had taken action in response to the behaviour(s) they had become concerned about, suggesting that people responding to the survey had progressed through the next stages of the situational model – sense of responsibility and recognising it as a problem, through to possessing skills and ultimately taking action. This supports the applicability of the bystander situational model to DVA in the general community. For coercive control and warning signs of DVA there was a high likelihood of taking action, but not as high as the likelihood of taking action after witnessing sexual abuse or the abuse of a vulnerable person, which was almost a certainty. This may be because these latter situations are deemed less ambiguous or more clearly ‘high-risk’ whereas warning sign behaviours and coercive control may carry more potential for uncertainty and interpretation. This is consistent with the literature on high risk or emergency situations (Fleming and Wiersma-Mosley, 2015; Storer et al, 2021).
The actions taken by our respondents are consistent with the literature on community bystanders’ behaviours (Frye et al, 2012; Weitzman et al, 2020). The majority of bystanders offered support to the victim, consistent with the literature that women are more likely to offer support to victims (Banyard, 2011).
While the study indicates that for the majority, barriers to intervention had been overcome, we know little about the experiences and barriers faced by those who had noticed but did not take action. Of those 12 per cent who did not take action very few explained why. Further research should explore this.
This study sheds further light on the relationships between awareness, noticing and action. Becoming aware of DVA ‘in person’ was significantly associated with the bystander taking action. Being told by the victim was also a strong predictor of the bystander taking action, whereas those who became aware of DVA online were least likely to take action. This adds to the evidence that knowing the victim and connection to the victim is related to a heightened sense of responsibility and helping behaviours (Levine et al, 2002; Burn, 2009). Becoming aware of DVA online may diminish that sense of responsibility as the bystander is not in close proximity to the victim at the time of having concerns (Coyne et al, 2019). It is also possible that being directly told by the victim or becoming aware of it ‘in person’ reduces the operation of social determinants such as the ability to diffuse responsibility to others.
The applicability of the situational model, and in particular the importance of having skills, and confidence in those skills to intervene (Berkowitz, 2009) is again confirmed by this study: for 37 per cent, feeling that they possessed the correct skills was a motivator to taking action. Some form of training is likely to be important in moving people through the situational model, as having had DVA training increased action-taking and was a strong predictor of offering support to the victim as well as being associated with belief in possessing the correct skills. Further, for the very few who reported why they had not taken action, not feeling that they possessed the skills was a key barrier. Unlike other studies (Weitzman, 2020; Storer et al, 2021; McInnes, 2022), fear of physical safety was not a key barrier, and this is likely because those previous findings were in relation to more ‘high-risk’ incidents of physical abuse and emergency situations. It is an interesting finding that skills possession was the highest motivator in action for physical abuse and sense of community connectedness. Perhaps perceiving that one has the correct skills obviates fear for personal safety.
Impact
It is an important finding and addition to the literature that over half of the participants indicated that intervening had a negative impact on them, yet despite this, most would not have done anything differently. Perhaps this is because they were motivated to intervene and did so but had no other skills or strategies at their disposal. Three-quarters of those indicated the utility of bystander training to guide them in how to take appropriate prosocial action. This suggests bystander training in a multiplicity of intervention strategies and concurrent bystander behaviour modelling campaigns might be important in ameliorating this impact by providing not only the skills to intervene safely and appropriately but also with the confidence in the skills which may overcome self-doubt. This suggests that there is a need to develop bystander training, which goes beyond awareness-raising and is accessible to the public for all ages as an important tool in preventing DVA.
Limitations
There were limitations to this pilot study that need to be addressed when considering the results. First, the survey was solely available online, in Welsh and English. This limited responses to only those with internet access, who could understand English and/or Welsh.
Second, this study struggled to recruit participants to interview. The small number of interview participants (n=3) limited the amount of analysis that could be conducted on their experiences. The recruitment method for the interviews should be improved and simplified in future iterations of the study to optimise the number of people consenting to take part in the interviews.
Last, this study did not aim for a representative sample, instead aiming to elucidate the experiences of those who volunteered to participate in the survey and interview. A large proportion of the sample had a high self-reported knowledge of DVA, the majority were women, and White British. No one over the age of 75 participated in the survey. Further research should look to engage a larger, more representative sample, including more men, people from racial and ethnic minority groups and older people to ensure a broader representation of bystander experiences. Further research should also aim to collect data on what type of DVA training had been received by participants. A larger sample would also increase the validity of the logistic regression model.
Recommendations
Practice
This study suggests that while public health restrictions implemented during the pandemic exacerbated DVA, they also increased the opportunity for bystanders to witness DVA behaviours, and the opportunity for bystanders to intervene. Bystanders in the study reported that possessing the correct skills and confidence to act were a significant motivator in taking prosocial action, and that bystander training which mitigates negative impact on bystanders would be helpful in developing and providing confidence in using these skills. As such, this study provides a case for the development of bystander training programmes and public awareness campaigns as an important element of DVA prevention in future pandemics. These bystander training programmes must be evidence based and theoretically informed (Fenton and Mott, 2017).
Research
The data highlights how the participants’ experiences had a negative impact upon their psychological, social, financial and physical wellbeing. Future research should explore how this negative impact could be mitigated. This may be through bystander training to encourage confidence in actions taken or support services for bystanders.
Conclusion
Findings from this study suggest the COVID-19 pandemic had allowed people to become aware of DVA. Further, when people have knowledge and skills to intervene, most will act as prosocial bystanders when they witness DVA. Providing bystander training for DVA should be linked to public awareness programmes so that people are aware of how they can act safely when they witness DVA.
Funding
This study was funded by Public Health Wales.
Acknowledgements
The authors would like to thank the advisory group for their input in the study, and all research participants, without whom this study would not have been possible.
Conflict of interest
The authors declare that there is no conflict of interest.
References
Addis, S. and Snowdon, L. (2021) What Works to Prevent Violence Against Women and Girls Domestic Abuse and Sexual Violence (VAWDASV)? Systematic Evidence Assessment, Cardiff: Wales Violence Prevention Unit.
Ansara, D.L. and Hindin, M.J. (2010) Formal and informal help-seeking associated with women’s and men’s experiences of intimate partner violence in Canada, Social Science & Medicine (1982), 70(7): 1011–18. doi: 10.1016/j.socscimed.2009.12.009
Banyard, V.L. (2008) Measurement and correlates of prosocial bystander behaviour: the case of interpersonal violence, Violence and Victims, 23(1): 83–97. doi: 10.1891/0886-6708.23.1.83
Banyard, V.L. (2011) Who will help prevent sexual violence: creating an ecological model of bystander intervention, Psychology of Violence, 1(3): 216–29. doi: 10.1037/a0023739
Banyard, V.L. (2015) Toward the Next Generation of Bystander Prevention of Sexual and Relationship Violence: Action Coils to Engage Communities, Cham: Springer.
Banyard, V.L., Moynihan, M.M., Cares, A.C. and Warner, R. (2014) How do we know if it works? Measuring outcomes in bystander-focused abuse prevention on campuses, American Psychological Association, 4(1): 101–15.
Banyard, V.L., Rizzo, A.J. and Edwards, K.M. (2020) Community actionists: understanding adult bystanders to sexual and domestic violence prevention in communities, Psychology of Violence, 10(5): 531–41. doi: 10.1037/vio0000281
Bennett, S., Banyard, V.L. and Edwards, K.M. (2017) The impact of the bystander’s relationship with the victim and the perpetrator on intent to help in situations involving sexual violence, Journal of Interpersonal Violence, 32(5): 682–702. doi: 10.1177/0886260515586373
Berkowitz, A. (2009). Response Ability: A Complete Guide to Bystander Intervention, Chicago, IL: Beck.
Bovill, H. and White, P. (2020) Ignorance is not bliss: a UK study of sexual and domestic abuse awareness on campus, and correlations with confidence and positive action in a bystander program, Journal of Interpersonal Violence, 37: 5–6. doi: 10.1177/0886260520916267
Bradbury-Jones, C. and Isham, L. (2020) The pandemic paradox: the consequences of COVID-19 on domestic abuse, Journal of Clinical Nursing, 29(13–14): 2047–9. doi: 10.1111/jocn.15296
Brown, A.L., Banyard, V.L. and Moynihan, M.M. (2014) College students as helpful bystanders against sexual violence: gender, race, and year in college moderate the impact of perceived peer norms, Psychology of Women Quarterly, 38(3): 350–62. doi: 10.1177/0361684314526855
Burn, S.M. (2009) A situational model of sexual assault prevention through bystander intervention, Sex Roles, 60(11–12): 779–92. doi: 10.1007/s11199-008-9581-5
Campbell, A.M. (2020) An increasing risk of family violence during the Covid-19 pandemic: strengthening community collaborations to save lives, Forensic Science International: Reports, (2): 3. doi: 10.1016/j.fsir.2020.100089
Cantor, D., Fisher, B., Chibnall, S., Harps, S., Townsend, R. et al. (2020) Report on the AAU Campus Climate Survey on Sexual Assault and Misconduct, Revised Aggregate Report and Appendices, 1-7_(01-16-2020_FINAL).pdf (aau.edu).
Christensen, M.C. and Harris, R.J. (2019) Correlates of bystander readiness to help among a diverse college student population: an intersectional perspective, Research in Higher Education, 60(8): 1195–226. doi: 10.1007/s11162-018-09544-6
Coyne, I., Gopaul, A., Campbell, M., Pankász, A., Garland, R. and Cousans, F. (2019) Bystander responses to bullying at work: the role of mode, type and relationship to target, Journal of Business Ethics, 157: 813–27. doi: 10.1007/s10551-017-3692-2
Edwards, K.M., Mattingly, M.J., Dixon, K.J. and Banyard, V.L. (2014) Community matters: intimate partner violence among rural young adults, American Journal of Community Psychology, 53(1–2): 198–207. doi: 10.1007/s10464-014-9633-7
Fenton, R. and Mott, H.L. (2017) The bystander approach to violence prevention: considerations for implementation in Europe, Psychology of Violence, 7(3): 450–8.
Fenton, R. and Mott, H.L. (2018) Evaluation of the Intervention Initiative: a bystander intervention program to prevent violence against women in universities, Violence and Victims, 33(4): 645–62.
Fleming, M. and Wiersma-Mosley, J.D. (2015) The role of alcohol consumption patterns and pro-social bystander interventions in contexts of gender violence, Violence Against Women, 21(10): 1259–83. doi: 10.1177/1077801215592721
Frye, V., Paul, M.M., Todd, M.J., Lewis, V., Cupid, M., Coleman, J.M., Salmon, C. and O’Campo, P. (2012) Informal social control of intimate partner violence against women: results from a concept mapping study of urban neighbourhoods, Journal of Community Psychology, 40(7): 828–44.
Gainsbury, A.N., Fenton, R.A. and Jones, C.A. (2020) From campus to communities: evaluation of the first UK-based bystander programme for the prevention of domestic violence and abuse in general communities, BMC Public Health, 20: 674. doi: 10.1186/s12889-020-08519-6
Hamby, S., Weber, M.C., Grych, J. and Banyard, V. (2016) What difference do bystanders make? The association of bystander involvement with victim outcomes in a community sample, Psychology of Violence, 6(1): 91–102. doi: 10.1037/a0039073
Ivandic, R., Kirchmaier, T. and Linton, B. (2020) Changing patterns of domestic abuse during COVID-19 lockdown, Social Science Research Network (SSRN), doi: 10.2139/ssrn.3686873.
Jouriles, E.N., Krauss, A., Vu, N.L., Banyard, V.L. and McDonald, R. (2018) Bystander programs addressing sexual violence on college campuses: a systematic review and meta-analysis of program outcomes and delivery methods, Journal of American College Health, 66(6): 457–66. doi: 10.1080/07448481.2018.1431906
Kettrey, H.H. and Marx, R.A. (2018) The effects of bystander programs on the prevention of sexual assault across the college years: a systematic review and meta-analysis, Journal of Youth and Adolescence, 48: 212–27. doi: 10.1007/s10964-018-0927-1
Kofman, Y.B. and Garfin, D.R. (2020) Home is not always a haven: the domestic violence crisis amid the COVID-19 pandemic, Psychological Trauma: Theory, Research, Practice and Policy, 12(81): 5199–201. doi: 10.1037/tra0000866
Latané, B. and Darley, J.M. (1968) Group inhibition of bystander intervention in emergencies, Journal of Personality and Social Psychology, 10(3): 215–21. doi: 10.1037/h0026570
Latané, B. and Darley, J.M. (1969) Bystander apathy, American Scientist, 57(2): 244–68.
Levine, M., Cassidy, C., Brazier, G. and Reicher, S. (2002) Self-categorisation and bystander non-intervention, Journal of Applied Social Psychology, 32: 1452–63. doi: 10.1111/j.1559-1816.2002.tb01446.x
McInnes, E. (2022) Bystander attitudes to hearing family violence: an Australian survey, International Journal of Criminology and Sociology, 11: 1452–63. doi: 10.6000/1929-4409.2022.11.06
McMahon, S. (2015) Call for research on bystander intervention to prevent sexual violence: the role of campus environments, American Journal of Community Psychology, 55: 472–89. doi: 10.1007/s10464-015-9724-0
Mujal, G.N., Taylor, M.E., Fry, J.L., Gochez-Kerr, T.H. and Weaver, N.L. (2019) A systematic review of bystander interventions for the prevention of sexual violence, Trauma, Violence and Abuse, 22(2): 381–96. doi: 10.1177/1524838019849587
NICE (National Institute for Health and Care Excellence) (2014) Domestic Violence and Abuse: Multiagency Working, Introduction and Guidance, London: NICE.
ONS (Office for National Statistics) (2020) Domestic Abuse Prevalence and Trends, England and Wales: Year ending March 2020, Newport: ONS, ons.gov.uk.
ONS (Office for National Statistics) (2021) Population and Household Estimates, Census 2021, Newport: ONS, www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/bulletins/populationandhouseholdestimateswales/census2021.
ONS (Office for National Statistics) (2022) Domestic Abuse in England and Wales Overview: November, Newport: ONS, ons.gov.uk.
Pagliaro, S., Pacilli, M.G. and Baldry, A.C. (2020) Bystanders’ reactions to intimate partner violence: An experimental approach, European Review of Social Psychology, 31(1): 149–82. doi: 10.1080/10463283.2020.1776031
Peterson, N.A., Speer, P.W. and McMillan, D. (2008) Validation of a brief sense of community scale: confirmation of the principal theory of sense of community, Journal of Community Psychology, 36: 61–73. doi: 10.1002/jcop.20217
Roberts, N. and Marsh, H. (2022) A qualitative evaluation of bystander training: what works? The Howard Journal of Crime and Justice, 61(4): 427–43. doi: 10.1111/hojo.12487
Rothman, E.F., Edwards, K.M., Rizzo, A.J., Kearns, M. and Banyard, V.L. (2019) Perceptions of community norms and youths’ reactive and proactive dating and sexual violence bystander action, American Journal of Community Psychology, 63: 122–34. doi: 10.1002/ajcp.12312
Sacco, M.A., Caputo, F., Ricci, P., Sicilia, F. et al. (2020) The impact of the COVID-19 pandemic on domestic violence: the dark side of home isolation during quarantine, Medico-Legal Journal, 88(2): 71–3. doi: 10.1177/0025817220930553
Speed, A., Thomson, C. and Richardson, K. (2020) Stay home, stay safe, save lives? An analysis of the impact of COVID-19 on the ability of victims of gender-based violence to access justice, The Journal of Criminal Law, 84(6): 539–72. doi: 10.1177/0022018320948280
Storer, H., MCleary, J.S. and Hamby, S. (2021) When it’s safer to walk away: urban, low opportunity emerging adults’ willingness to use bystander behaviors in response to community and dating violence, Children and Youth Services Review, 121: 105833. doi: 10.1016/j.childyouth.2020.105833
Taylor, E., Banyard, V., Grych, J. and Hamby, S. (2019) Not all behind closed doors: examining bystander involvement in intimate partner violence, Journal of Interpersonal Violence, 34(18): 3915–35. doi: 10.1177/0886260516673629
Weitzman, A., Cowan, S. and Walsh, K. (2020) Bystander interventions of behalf of sexual assault and intimate partner violence victims, Journal of Interpersonal Violence, 35(7): 1694–718. doi: 10.1177/0886260517696873
WHO (World Health Organization) (2021a) Violence Against Women, Geneva: WHO, https://www.who.int/news-room/fact-sheets/detail/violence-against-women.
WHO (World Health Organization) (2021b) WHO Global Plan of Action to Strengthen the Role of the Health System Within a National Multisectoral Response to Address Interpersonal Violence, in Particular Violence Against Women and Girls, and Against Children, Geneva: WHO.
Wong, J.S., Bouchard, J. and Lee, C. (2021) The Effectiveness of college dating violence prevention programs: a meta-analysis, Trauma, Violence & Abuse, 24(2): 684–701. doi: 10.1177/15248380211036058
