Introduction

The World Health Organization (WHO, 2021) estimates that around 1 in 3 women worldwide have been subjected to intimate partner violence in their lifetime, with intimate partners being responsible for as many as 38% of all murders of women globally. Within the national context of the authors, there were 910,980 police-recorded domestic violence and abuse [DVA] crimes in England and Wales in 2022, representing an increase of 7.7% on the previous year (Office for National Statistics, 2022). Since the beginning of COVID-19 pandemic, reports of all forms of gender-based violence including DVA, have intensified internationally (Sharma & Borah, 2022), leading to UN Women labelling it the Shadow Pandemic (UN Women, 2021).

In an attempt to explore the social and economic costs in anticipation, as consequence and in response to DVA, the Home Office (2019) undertook an analysis of the Crime Survey for England and Wales data relating to DVA, using the Quality Adjusted Life Year (QALY) method. It found that in the year ending 31 March 2017, DVA was estimated to have cost over £66 billion in England and Wales alone. The costs included physical and emotional harms, the cost of time off work, costs to the health service, criminal justice system, housing, and victims’ services. Further, there is clear recognition of the emotional, psychological, and relational impact on children who witness DVA (Moylan et al., 2010).

However, there is widespread recognition of the underreporting of DVA which acknowledges that the actual number of DVA cases and incidents are far greater than official statistics reflect. Gracia (2004) reflects on the ‘tip of the iceberg’ phenomena in official DVA statistics suggesting that victim blaming, and societal attitudes and inhibition contribute to an environment in which reporting DVA is deeply challenging and problematic. Reardon and Trevillion (2016) suggest that it is therefore impossible to know the true scale of DVA, but note that police in the UK reported receiving one call a minute related to DVA. It can be argued that as a result of underreporting, the true impact and harm associated with experiences of DVA is also likely to be underestimated. In addition to the use of the term domestic violence and abuse (DVA), this paper also uses the term ‘harm’ or ‘harms,’ this is to reflect the broad range of behaviours and experiences that are understood to be DVA.

Whilst there has been a large amount of research into DVA it continues to be a global harm experienced by many. This indicates that there remains a need for innovative research which goes beyond official sources to explore the realities of DVA, and to offer creative and collaborative approaches to research. This paper aims to support researchers in critical engagement on a range of ethical and methodological issues relating to DVA and to encourage them to enter into similar engagement within their own research. It will synthesise existing models and guidance around domestic abuse research, with the aim of centralising best practice. To address current gaps it will draw on research models from wider disciplines, with a specific focus on related work within health, in addition to the authors combined front-line and research experience. Particular attention will be paid to areas where there is a paucity of literature, including considerations around participation and remuneration within domestic abuse research, and the unique complexities of domestic abuse perpetrator involvement. A set of recommendations will then be provided for how to conduct domestic abuse research, including guidance around remuneration.

Power and Participation

Power manifests in multiple relational ways during experiences of domestic violence and abuse [DVA] and also within DVA research. Power disparities exist at many key points of the research process, such as when designing the research aims, questions and data collection methods, when deciding which studies get funded or receive ethical approval, and ultimately the knowledge that is produced (Scantelbury, 2005).

The context in which domestic violence research [DVR] occurs, is in-part shaped by the funding landscape. DVR can be funded by multiple sources, for example research councils, international funds, government departments, and charitable trusts. Funders assess a variety of aspects of the proposed project including the composition of the research team, ultimately holding the power to decide who is considered an expert worth commissioning. The agenda of each funding body varies, as will decision-makers’ understandings about DVA. The power and success of feminist activism and research on violence against women has revolutionised public thinking about DVA which in turn has shaped narratives about risk and urgency. Considering Minnich’s (1991) framework of conceptual errors, this reveals the potential for powerful ideas and groups to dominate the DVR agenda. Bent-Goodley argues that this is evidenced in the samples of qualitative DVR, which have “largely focused on White and poor women, despite the fact that domestic violence crosses race, ethnicity, socioeconomic status, religion, and sexual orientation” (2005: 197). There have been attempts to address these sampling biases in qualitative DVR (see for example Donovan & Barnes, 2020), however there is more work to be done here.

To a large extent, the power to make methodological decisions lies with the researcher(s), resulting in asymmetrical power relations between the researcher and the participants (Florczack, 2016; Scantelbury, 2005). As a result, research-related power dynamics vary across and within research paradigms and are dependent on the methodologies employed by researchers (Karnieli-Miller et al., 2009).

The traditionally unequal structures of power associated with the relationship between the researcher and the researched can be considered ‘symbolic violence’ (Bourdieu, 1999). This may be particularly the case when researching with those who have experience of domestic violence and abuse, whose experiences may make them more susceptible to entering unequal relationships (Morgan & Björkert, 2006). Malpass et al. (2016) argue that whilst ‘symbolic violence’ is inherent in any research encounter, careful consideration of the impact of these relationships, and the societal structures which underpin them, can minimise any potential harm derived from them and encourage truly informed and ongoing consent, for example by emphasising participants’ agency in decision-making.

A means of facilitating agency is in clarity of language and terminology when developing and advertising DVA studies, increasing the likelihood of fully informed consent and more inclusive research studies. For example, McGarry and Ali (2016) argue that the language used to conceptualise DVA and the experiences reflected in recruitment materials will directly influence who might participate in a study, and give some guide to how the researchers might respond to the participants and their experiences. Therefore, we argue researchers must consider the concepts and definitions operationalised by any given study since they convey to potential participants who and which experiences are relevant to the research.

This is especially important because, whilst it is widely accepted that women and girls are disproportionately affected by male violence and abuse (UN Women, 2022), those whose experiences do not fit within the White, able bodied, heteronormative, and cisgendered public story of DVA (Donovan & Hester, 2014) are less likely to recognise themselves in materials which portray DVA in this way. Arguably this silences them, since they are less likely to participate in research which does not appear to define DVA in a way which is inclusive of their experiences or of them as individuals.

The silencing of these individuals can be considered symbolic violence, which is often shaped by and reflected in the intersectional systems of oppression within wider society. Often, it is the most marginalised survivors whose voices are not heard, including those who cannot access services because they have no recourse to public funds. This may be in part, as argued above, that they do not recognise themselves within the recruitment literature, or are not known to gatekeepers or offered the opportunity to participate (Vearey et al., 2017). Some survivors, such as migrants and LGBTQIA + persons and people of colour, experience multi-faceted marginalisation (Malpass et al., 2016) and are often not represented in research samples (Bent-Goodley, 2005), except where studies seek explicitly to explore their experiences (Montoya & Rolandsen Agustin, 2013). They may encounter specific barriers to participation, including not speaking the language of the country of residence (O’Brien Green, 2018), fear of repercussions and adding to negative stereotypes about their communities (Donovan & Barnes, 2020), and cultural expectations (Mergaert et al., 2015).

In addition, some survivors may be acutely aware of the potential for further symbolic, structural, or physical violence as a result of participation (Downes et al., 2014), consequently they may actively and legitimately choose not to participate. Fisher (2012) argues that this is agency in operation, counteracting the paternalistic assumption that victim/survivors are inherently vulnerable, and consequently incapable of risk assessing the research encounter for themselves. Without consideration and address, these barriers can block survivors from taking part in research, which in turn perpetuates the silencing of specific individuals and groups (Leye et al., 2014), and ultimately, limits the knowledge produced. O’Brien Green (2018) argues that to minimise the potential for symbolic violence, researchers should seek to redistribute power across the research process, ideally through co-design with experts-by-experience. She advocates that as a minimum, consultation with experts-by-experience and other community leaders should be considered. This may require careful ethical consideration when conducting research with perpetrators of violence and harm, as discussed below.

A further complexity related to recruitment is payment, and consequently inequalities in DVA research participation can be further reinforced. Researchers should consider the questions of ‘who gets to participate and therefore who gets paid?.’ Methods of recruitment to studies of abuse frequently include gatekeepers (Aluwihare-Samaranayake, 2012), who are often key stakeholders in research. They may be involved in the design of a project, and often agree to share information about the project with their clients. However there is potential for community and service-based gatekeepers to exert their relational power intentionally or unintentionally over their clients, creating a sample bias (O’Brien Green, 2018). Whilst this policy might be appropriately protective of individuals who have experienced harm it also raises important ethical questions. Which gatekeepers do researchers approach? Which participants do the gatekeepers advertise studies to? Furthermore, some clients may feel a sense of obligation to give back to an organisation that has supported them (Taylor and Clarence, 2021), whereas others may be encouraged to take part because they will provide a positive review of the gatekeeping organisation (O’Brien Green, 2018). For perpetrators of harm participation in research may be seen as an opportunity, authentic or otherwise, to demonstrate a change in their views and/or behaviour. Furthermore, there may be an additional layer to the inducement presented by the offer of payment.

There are a number of services which have survivor groups and indeed some organisations that have a survivor network who have indicated their willingness to take part in research studies (for example Survivors Voices). There is real value in these networks for survivor support and advocacy and they can help researchers really understand lived experience. These networks can be significantly helpful in recruiting to studies, but it is also important to reflect the views of victim/survivors who are not connected to such networks and may therefore be excluded from participation in research studies. This is important for a number of reasons. Those who experience abuse may often have experienced silencing as part of their harm and the inability to participate in research studies on their area of experience may further increase this silencing. There is sometimes a reflection that certain individuals commonly are asked to participate either in research or advocacy work where others are not. It is of course essential that autonomy and choice about research participation and informed consent are adhered to at all times. However, there is a real challenge in this area about how to ensure equal access to participation. Where participation is paid this may be argued to be even more pertinent and important to consider. The experience of abuse for many individuals may impact their income and therefore opportunities for paid participation should where possible be more open than they currently are. This may also produce more diversity in participation and in turn richer and more detailed and representative research data. However, the challenges of limited or capped payments for participation by institutions (Head, 2009) and maximum sample sizes make equitable recruitment and payment extremely challenging for qualitative researchers, as a result there may be limitations in researcher capacity to truly redistribute power across the research process through payment.

Significant efforts, predominantly in health research (e.g. Boivin et al., 2018) and in feminist participatory research (e.g. Weber & Thomas, 2021), have been made to redistribute power in research on domestic violence and abuse. These efforts allow for an understanding of power as fluid, whereby the currents of power flow between the traditionally powerful researcher and traditionally vulnerable participant. The co-design of research enables key decisions, for example of the language used in recruitment materials, to be designed with those whose experiences the research is exploring. Furthermore, Houghton (2015: 138) argues that “participatory ethics are intrinsic to involvement that focuses on empowerment and impact” for survivors. An additional challenge emerges here when researching those who may have abused their power.

Gabriel et al. (2017: 160) argue that ‘all deep qualitative exploration and analytical work transforms researchers into co-participants and elicits a co-created project’, in other words the impact of both researchers’ and participants’ biographies and interests inform the research process and outcomes in a multitude of ways. Many researchers are drawn to research topics they have experience of. It follows that DVA researchers may also have experience of DVA. Since 1 in 4 women and 1 in 6 men are thought to experience some form of domestic and/or sexual violence in their lifetime (Office for National Statistics, 2022), this is likely to be a common occurrence. Indeed, the authors of this paper have conducted research on topics we have experience(s) of. By being cognisant of the positionality of researchers, the degree of proximity between the researcher and the researched (McGarry, 2010), and the potential for the research process to impact on researchers and participants, space can be created to consider how meaning is made in relation to decision making, conducting research, knowledge production and dissemination. What is clear is that power is inherent in the qualitative research process and a detailed examination and consideration is required in designing DVA research studies to mitigate the many risks of further marginalisation or silencing of those who have experienced DVA.

Payment

Participation in research requires that participants have the capacity and resources to contribute to the study. According to the charity Surviving Economic Abuse (2023), 95% of victim/survivors of DVA experience economic abuse which directly impacts the economic resources available to them, and often leaves them with excessive debt and poor credit ratings (Braaf & Barrett Meyering, 2010). This is compounded by the wider context of gendered economic insecurity (Postmus et al., 2020: 262). Therefore, compensation for participation is vital.

One method employed to seek to ensure that those who participate in research are recognised and compensated is through payment. Participant payment is now an accepted practice within qualitative research and is argued to be an increasingly common part of research recruitment (Head, 2009). Russell et al. (2000) suggest that an emphasis on participant time and effort in research studies should be considered an integral aspect of the ethical principle of respect for persons. Indeed, there is an argument that non remuneration especially among potentially vulnerable populations could be both unethical and exploitative (Gordon et al., 2018). When it comes to victim-survivors being asked to share their experiences, doing so without payment has been argued to be exploitative, and there is a clear call for all ‘experts by experience’ to be offered payment for sharing their experiences (Taylor and Clarence, 2021).

As with considerations of power, the perception of participation payment is both complex and diverse. Participant payment can be conceptualised as gratitude for the time given during participation (Rowlingson & Mckay, 1998). In qualitative research it is often seen as part of the broader drive, previously discussed in this article, to redress the balance of power in research studies (Thompson, 1996). There is an argument that ethically designed research including participant payment can provide opportunity for participants to contribute understanding and awareness and in the process may create in them satisfaction and self-respect (Emmanuel et al, 2000), and autonomy (Dickert & Grady 1999). In this context payment can be seen as a solution rather than a problem (Largent et al., 2019).

In the NHS reimbursement and payment of patient and public voices (PVV) is common practice. There are four levels of engagement and subsequent payment and reimbursement listed, these being, (1) no payment or reimbursement; (2) and (3) expenses only; and (4) payment via payroll—subject to deductions. The guidelines (NHS, 2021) suggest payment can remove financial barriers to participation and reasonable expenses can include travel, accommodation and subsistence and carer support (where appropriate). Pre-booking and timely reimbursement are deemed to be ideal. However, the guidelines also note that the payment of expenses could compromise those on benefits who may be deemed to demonstrate ‘work readiness’ by participation. Further PPV partners who participate while off work sick could breach policy. Additional guidance is provided to take extra care to ‘proactively cover travel and subsistence costs’ for children and young people who do not have their own funds. These guidelines may be important to consider for those experiencing domestic abuse which may include financial abuse and control. Researchers should take all these factors into consideration when designing studies in the area of domestic abuse and be particularly mindful of possible impacts on benefits and payments.

Further consideration should be given to the type of payment offered. Vouchers are commonly employed as a means of payment, this can ensure the anonymity of the participant as there are no expense forms to complete. It can also be a quick method to provide timely payment. However, if the argument is to remunerate ‘experts by experience’ then only offering vouchers and not monetary payment could be argued to reduce the value of participation or the choice of what to spend money on. It must be noted that if expense claims are to be made these usually need to go through institutional processes, which can be time consuming and may result in delayed payments. Cash payments may resolve some of the issues above but could place researchers at risk who need to carry money with them when conducting research (Head, 2009). Further, it could be argued that cash payments in an increasingly cashless society could become increasingly irrelevant. However, Chronister et al. (2004) argue against the use of cheques for participants in domestic violence studies as they note they may not control their own accounts. Although cheques are now less frequently used, this argument could still be pertinent for research with those who experience domestic abuse where cash payments may be required as they would be untraceable and would not leave electronic evidence of participation.

As stated, research payments are often intended to motivate participation but also to adequately recompense for expenses and time. There is little research conducted on victim/survivor perceptions of payment in Domestic Violence Research [DVR]. In one of the few published pieces female participants were clear that payment for participation was important. This payment reflected not only the time provided in participation but also the value of the information that participants had to share (Logan et al., 2008). They also noted that participants were often intrinsically motivated by the hope that they may improve situations for other women (Logan et al., 2008).

Similarly, the young survivors of familial DVA in Houghton’s (2015) research were explicit that participation was not perceived as worth the actual or potential costs (e.g. emotional impact), if real world change in policy and practice was not realised. Participation, with or without financial remuneration, that did not lead to change was described as tokenistic and voyeuristic for them. We therefore argue that understanding participants are motivated by seeking to change or improve practice, requires researchers to reflect upon the outcomes of their proposed research study. They should be clear and realistic about outcomes with potential participants. Further, researchers should give careful thought to dissemination using clear and accessible formats. Planning careful dissemination can act as an additional and meaningful positive consequence of participation, enabling wide distribution of findings and facilitating change. Examples of outputs from different organisations include freely available PDFs, films, podcasts, webinar, roundtables, and other digital resources (e.g. Daw, 2021, Meechem et al., 2023; SafeLives, 2021; Thirtyone: eight, 2021).

It should be noted that some conceptualisations of participant payment in research studies are negative with payment being seen as a ‘necessary evil’ (Largent et al., 2019). This perception of payment Largent et al. (2019) argue is surprising in a general societal context where payment for goods and services is a commonly accepted principle. A widely reported concern in writings on participant payment is the reflection that payment can be conceived as coercive (Macklin, 1981), even inducing individuals to participate in research they would not normally choose to be part of (Bentley and Thacker, (2004). Goodman et al., (2004) suggest the line between payment which incentivises and payment which is coercive is not clear. Bentley and Thacker (2004) concur that a distinction between inducement and coercion is important, suggesting that inducements are linked to motivation rather than to threats which are more commonly associated with coercion. This is of particular import to reflect on in research conducted around DVA, where coercion is a recognised element of this form of harm (Home Office, 2012). Millum and Garnett (2019) reflect that in some circumstances participant payment could even result in participants minimising their own purposes to support the researcher and/or gatekeeper’s agenda. This questions whether payment addresses the power imbalance between research and participants in the manner which qualitative researchers hope it might. What is clear from the discussions is that a considered transparent and reflective thinking process about participant payment should be an integral part of any qualitative research study around DVR.

There are further complexities related to determining how much payment should be made. There has been caution raised regarding high payments which might therefore cross the border between inducement and coercion. Goodman et al., (2004) suggest that this is especially pertinent for those on low incomes who could feel coerced into participation if the payment is particularly high. This again may be a pertinent point for DVR where financial abuse may be part of an individual's experience and this can result in low income or financial challenges. In this context, high payments could be critically considered as coercive. However, others have argued that substantial financial payments should not automatically be equated to coercion. An argument proposed is that payment of a substantial salary is not seen to undermine individual autonomy in work contexts (Wilkinson & Moore, 1999) and therefore high participant payment does not automatically undermine autonomy in research participation decisions. In some ways it could be argued that higher payment in DVR may provide finance for additional support for the impact of participation, such as child care or therapeutic support, as well as the time to participate. A further perspective is that low participant payment might equate in participants' minds to low risk or low impact research studies (Cryder et al., 2010). Again, this is of particular pertinence to DVR where tokenistic payments or low value payments could be contextualised by participants as low risk activity. Re-traumatisation in research studies is not uncommon and therefore, consideration should be given to whether there is any relationship between participant expectations of the level of impact and distress that may be experienced and low or no participant payments.

Dickert et al. (2002) proposed models for ethical payment of research participants. These include the ‘market model’ where payment is used as an incentive and additional payment can be made on completion. A ‘reimbursement model’ that reimburses expenses and a ‘wage payment model’ reimbursing equivalent to the working wage. Whilst the ‘fair share model’ viewing participants as partners is proposed by Saunders and Sugar (1999). These models provide principles for remuneration. However, even where agreed principles for remuneration have been developed there is an argument that they need to be culturally and context specific taking account of the burden of participation, reasonable remuneration and compensation for time (Gordon et al., 2018). Thus, there is currently no agreed model of payment.

Jackson et al, (2020) reflect that with PPI meaningful engagement requires proper funding and the same could be argued for DVR. However, challenges remain in research design and recruitment, even if an agreed model is created, since the challenge to obtain funding to adequately pay participants is clear. Head (2009) reflects that the amount of maximum payment given to participants may be limited by the research budget for a specific study or the maximum amount an institution allows for participant payment. Therefore, there remains the risk that even with commitment to appropriate participant payment, the approved budget may restrict adequate remuneration and therefore appear tokenistic at best. For DVR it would seem especially important to be transparent about payment with participants and the limitations to this payment where they exist. However, we urge cautious reflection on payment. It could be suggested where there is limited or no payment made for research participation, any payment could be seen as a measure of value and positively experienced. What is clear is that further DVA research is needed to understand participants' perceptions and the implications of payment more fully.

Interestingly, there is often a lack of discussion of participant payment in published research articles on DVA. For example, Watlington and Murphy’s (2006) study on the roles of religion and spirituality among African American survivors of domestic violence states that the female participants were paid to participate, however there is no discussion of the ethics of payment or how payment was determined. This is not a critique of Watlington and Murphy but rather a reflection that ethical considerations about payment and clarity on decision making processes are not routinely part of the peer review process or the development of articles for publication. Head (2009) recommends more reflexivity around participant payment. She argues the necessity to increase discussions in published work about the range of possible impacts of participant payments.

Interestingly, although participant payment is seen and reported as an ethical issue, ethical protocols related to participant payment are rare (Polacsek et al., 2016). Where these protocols do exist, there is little consistency about calculation and administration of payments (Polacsek et al., 2016). For research focused upon DVA and other forms of harm, the issue of administration is especially pertinent. Where participants may wish to maintain their anonymity, especially from institutions, this can be particularly complex in terms of creating a system which maintains anonymity but ensures appropriate payment. This often results in a voucher system being implemented where the researcher can forward to the participant a voucher payment which does not require sharing of identifying details with the institution. This may also enable those in receipt of state financial support to participate without penalty for changes to their income via the loss of benefits (NIHR, 2022). However, trauma informed qualitative research should not only consider participant payment but also payment for supporters should a participant wish to bring a supporter to a research interview (NIHR, 2022) and consider funding follow-on support sessions should a participant wish to engage with this post-participation. This model is currently being employed in a funded study by one of the authors of this paper, focused on abuse within religious contexts and was used in DVR on teen relationship abuse (Meechem et al., 2023),. However, although we see this as good practice, we are suggesting that researchers still need to be cognisant of the possible compromising of anonymity through participant payment and support.

A review of the literature on participant payment and reflections on this demonstrate several articles written by researchers but a paucity of partnership and collaborative work with participants on this topic. Published research illustrates the complexities of participant payment. Including, how payments may be perceived differently by different individuals, the issue of inducement versus coercion including for those who use harmful and abusive behaviours, the necessity for a carefully thought through process of determining the amount and form of payment, the challenges of equitable recruitment and the potential impact of participation. It is because of this complexity that we are arguing for a collaborative co-design process for DVR wherever this is possible. Working with victim/survivors at the design stage will allow for discussion of all these issues ahead of decisions about participant payment. This may necessitate researchers applying for small pots of initial funding to pay victim/survivors at the co-design stage. We suggest that this could become part of the protocol for DVR.

Platform

Further complexities arise in DVR which includes those who harm or ‘perpetrators’ as participants. Whilst there is a growing body of domestic abuse research including ‘perpetrators’ particularly studies evaluating the efficacy of behaviour change programmes (Langhinrichsen-Rohling & Turner, 2012; McGinn et al., 2021), the inclusion of those who perpetrate abuse is still a contested topic, and there remain comparatively fewer studies involving those who harm as participants.

Perceptions and practices relating to the involvement of those with lived experience in domestic abuse research and education have evolved over time. Jury and Boxall (2018) describe how both those commonly perceived as morally bereft, such as perpetrators of abuse, and those commonly perceived as having low levels of agency, such as victims of abuse, were historically left out of service-user involvement in academic processes such as social work education. Over recent years, there has been increased recognition of the need to centre the voices of victim-survivors and engage them in both education and research (see Women’s Aid, 2020), moving away from the perception of them as lacking autonomy to viewing them as active agents and stakeholders (Downes et al., 2014). However, while there has been such evolution around victim-survivor participation, men who use intimate partner violence are argued to remain the most excluded group in social work education and across research. This is in part due to the lasting perceptions of them as morally bereft resulting in their voice being positioned as unreliable and deviant (Jury, 2022; Jury & Boxall, 2018).

For researchers who are considering including those who harm within their studies, the ethical principle of minimising harm is perhaps even more complex. While it is standard practice for researchers to consider possible harm to primary participants, literature on the participation of perpetrators of abuse emphasises the need to recognise that the potential for harm extends beyond the participant themselves to the other members of their family or household (Hearn et al., 2007). Downes et al., (2014, p.6) suggest that abuse may be exacerbated as a result of perpetrators participating in research if they consider that being able to participate equates to researchers ‘supporting their rationalisations or justifications of violence’. This is seen as particularly likely in longitudinal studies where there are often multiple contacts and perpetrators may be invited for subsequent interviews after they have discussed their use of harm in initial interviews, leading to the perception of acceptance or endorsement of this behaviour (Gondolf, 2000). As a result, researchers conducting research with perpetrators of abuse must ensure their risk analysis goes beyond primary participants to consider the interests of any other stakeholders (BPS, 2021). Women’s Aid (2020) provide an example of such an approach to risk in their Research Integrity Framework, where they suggest that perpetrators should always be interviewed first in studies where data collection is happening with multiple members of the family or household, in order to minimise the likelihood of information that may increase risk being accidentally shared with the perpetrator.

As part of expanding considerations of the potential for harm beyond the participant, domestic abuse researchers must also include the potential of harm to themselves. There is growing recognition within qualitative research of the impact of undertaking such work on the researcher, particularly when exploring ‘sensitive’ topics such as domestic abuse (see Fenge et al., 2019). Texts like ‘Danger in the field’ (Lee-Treweek & Linkogle, 2000) have opened this conversation with consideration of broader risks, and the authors themselves are part of a group in the early stages of establishing a network focused on the emotional experience of researching. As a result, guidance relating to researcher safety now commonly considers both physical and emotional safety (see AQR, 2022; SRA, 2023), moving away from a sole focus on the former, however, this is usually centred on immediate emotional impacts.

When applied to DVR, publications discussing the emotional impact on the researcher most commonly focus on hearing disclosures of abuse within interviews, and the potential emotional distress caused by this during or immediately following the interview. Though this remains an important element to consider, literature on vicarious trauma demonstrates the need to also take potential long-term impacts into account. These impacts are highlighted in Beckerman and Wozniak’s (2018) study of domestic violence counsellors, which describes a shift in worldview with many practitioners feeling their work changed how they perceived the world around them. Though domestic abuse researchers may not be working on the frontline, they experience a similar proximity to stories of abuse and harm that may well lead to a similar shift in worldview and experiences of vicarious trauma.

Beyond this, domestic abuse researchers must also consider how the principle of minimising harm applies with perpetrators of abuse. While avoiding and minimising psychological harm should be a consideration across all research (BPS, 2021), research deemed as sensitive, such as DVR, is argued to require extra consideration around ‘potential psychic costs such as guilt, shame and embarrassment’ (Downes et al., 2014, p.3). This raises some important questions when conducting research with those who harm, particularly when we are asking them to talk about the abusive behaviour they have used. Is it ethical or helpful for us to try and create an environment in which perpetrators can discuss their abuse without feeling any level of guilt, shame, or embarrassment? Hearn et al. (2007) discuss the challenge faced by researchers carrying out research with those who use violence, in avoiding inciting distress without being seen to or feeling that they are colluding. This is a challenge familiar to frontline practitioners working with perpetrators of abuse, with some training for such practitioners describing the process of building a relationship with those who harm as walking a tightrope between judgement and collusion. Considering this challenge, Hearn et al. (2007, p.8–9) argue that participant distress should not be viewed as a ‘neat, coherent ‘whole’, but an area of contestation and negotiation, that needs to be critically interrogated and evaluated in the specifics of each research project’.

Despite the challenges and understandable controversy of including those who harm within DVR, there are also a range of arguments for why their participation could be both necessary and valuable. Perhaps most importantly, perpetrators of abuse may have valuable insight that can be used to advance efforts to end domestic abuse. Research conducted with those who harm has concluded that this group can be engaged in constructive discussions about their experiences of accessing services, the barriers to doing so, and what works in effecting long-term change (Jury, 2022). When it comes to perpetrators who are resistant or avoidant of support and behaviour change work, or may be considered hard to reach, there is evidence that engaging them in research and education enables the identification of the professional skills and program elements needed to engage them in support and change their behaviour (Jury, 2022). Previous research has also highlighted how the perceptions of perpetrators accessing support can differ widely from those of practitioners (Jury & Boxall, 2018), suggesting that we cannot know what works and what does not without asking those whose behaviour we are trying to change.

It is also important to acknowledge that while we must mitigate potential harm and consider possible negative impacts of engagement in research, we should not ignore the positive impact involvement in research can have (Downes et al., 2014). This positive impact has been evident in an ongoing study conducted by one of the authors, in which young people (aged 11–25) who felt they had used abusive behaviour in relationships took part in interviews. Feedback on the experience of being interviewed was captured, with every participant strongly agreeing that participation had been a positive experience and one young person describing the process as healing, while another described how it had led them to rethink their behaviour in previous relationships and how they now hoped to improve their current relationship. While participation in research is not, and should not be viewed as, any kind of support or intervention, this feedback suggests that talking about their harmful behaviour may lead some participants to reflect. In addition to having possible positive impacts on participants, previous studies also suggest that involving those who harm in research may create unique learning opportunities for researchers. When reflecting on their involvement in learning groups with women who had experienced imprisonment, social work students described having their stigmatised views challenged and developing an increased awareness of the issues facing this group (Jury & Boxall, 2018). Though there remains a very important need to avoid possible collusion when conducting research with those who harm, it could be argued that there is a need to reflect on a default position which excludes those who harm from research participation. Whilst it may be an appropriate decision for some research, in others it may hinder learning which could inform the development of effective practice and intervention, and place all the responsibility for this learning on those who experience DVA. Therefore, there is a critical need to consider this topic further, including where there is bidirectional and/or situational couple violence (Johnson, 2008).

The absence of perpetrators of abuse within research reflects the absence of these voices within practice. In Roskill’s (2011) file audit of cases involving domestically violent men, the father was neither seen directly nor contacted by phone in 32% of the core assessments studied. The lack of expectation for perpetrator parents to engage with these processes and address the risk they present, results in the burden of ‘duty to protect’ being placed on victim parents, most often mothers, who are then held responsible for the safety of themselves and their children. To work effectively and appropriately with families experiencing domestic abuse, the responsibility must be shifted back to those who are actually causing the harm. When it comes to DVR it could be argued that solely relying on the voices of victim-survivors furthers their responsibilization and the narrative that perpetrators are not expected to be part of addressing the issue of abuse. In this way, including those who harm as participants in research promotes the expectation that they must be something other than ‘silent about causes of IPV and their responsibilities in desisting from violence’ (Jury, 2022, p.2).

For researchers who decide that the benefits of perpetrator participation outweigh the costs, there are then a series of ethical considerations around their involvement, including the issue of incentives and compensation. Whilst, as previously argued, there are many considerations for fair and ethical payment of victim/survivors, there are additional considerations around participant payment when the people being asked to share their experiences are those using, or who have used, abusive behaviour. Existing frameworks for ethical domestic abuse and gender-based violence research, such as the World Health Organization’s Ethical and safety recommendations for intervention research on violence against women (WHO, 2016) or Women’s Aid’s (2020) Research Integrity Framework either neglect to mention compensation, or solely discuss compensation in relation to organisations being partnered with, and the only mention of perpetrator participation is the guidance outlined earlier around arranging the order of interviews to avoid risk escalation. Hanson et al. (2012, p.1391) explored the issue of compensation when conducting research with offenders, a group representing similar challenges and conflicts of opinion to those who harm, and concluded that there are ‘no ethical principles that would justify categorically denying incentives.’ Surmiak (2020) also describes the use of payment as important for both increasing recruitment and compensating participants for their time. This article has outlined some of the arguments for including those who harm within DVR, therefore providing some justification for the need to increase recruitment. Dutton et al. (2003) studied recruitment and retention in intimate partner research and stated that violent men report participating in studies for financial compensation. They, like others, argue for the need to set the rates of compensation so that they are high enough to encourage those using harm to take part in research, but not so high as to be seen as coercive or rewarding.

While existing literature should be taken into account by domestic abuse researchers considering the inclusion of those who harm, their approach to research must also be informed by the views of victim-survivors. Though not reflected in academic literature, there are concerns from survivor groups that the inclusion of perpetrators in research provides them with a platform to further their narrative, and that compensating them for their participation is effectively rewarding them for their use of harm, particularly when the focus of the research is on their use of violence and abuse rather than their experience of behaviour change programmes.

This issue of whether to compensate perpetrators is further complicated by its reliance on the binary labels of victim and perpetrator, which have been argued to be reductionist and are often not as binary in practice as in theory. In reflecting on Corvo and Johnson (2003) and Reich (2002), Jury and Boxall (2018, p.512) discuss how ‘these overarching labels condense the discussion around those who experience violence and those who use it to conceptualisations of violence that are restrictive in nature; batterers premeditate, progressively increase their violence, are terroristic and inherently bad while victims are powerless, gentle, passive, reactive rather than proactive and, if they are a ‘real’ victim, inherently good’. Johnson’s (2008) typologies of abuse also suggest that a binary application of compensating victims and not perpetrators would not work in cases of situational couple violence, where abuse is often bidirectional and both partners can experience victimisation and instigation of harm. Research conducted by one of the authors, discussed earlier, also speaks to the nuance of these labels. While the study aimed to speak to young people who felt they had used harmful behaviour in their dating/romantic relationships, the majority of those who responded (predominantly women and one non-binary/genderqueer young person) described experiences of victimisation either in the home or their earliest dating/romantic relationships. Though some of the harmful behaviour they went on to describe using themselves could be conceptualised under the banner of violent resistance (Johnson, 2008), many described a trajectory from victim to instigator, whereby they went on to use harmful behaviours in subsequent relationships. If a label is needed to determine whether these participants should receive compensation, which label should be applied to the young people in this study?

Though there appears to be no simple solutions, domestic abuse researchers considering the inclusion of perpetrators as participants must take time to explore why participation is necessary and valuable, considering how to balance ethical approaches to research which advocate for compensating participants, alongside the centring of voices of victim-survivors who may not agree with this practice when it comes to perpetrators of abuse.

Concluding Thoughts

This paper has presented some of the complex and challenging methodological and ethical issues in conducting DVR. In response to a paucity of literature particularly around participation and remuneration within domestic abuse research, and the unique complexities of domestic abuse perpetrator involvement, it has critically explored the multifaceted topics of power, payment, participation, and platform. In doing so, we have foregrounded the importance of articulating and managing power dynamics in DVR, and suggest measures to ensure such power dynamics are mitigated successfully to ensure participation is accessible to all who want to share their experiences with researchers, that participants are transparently and appropriately paid for their participation, and where appropriately risk assessed, those who harm are included.

Drawing on best practice from health research, the paper reflects the continued potential marginalisation and silencing of some with lived experience of DVA if existing models and methods of recruitment remain unchallenged and unchanged. We call upon DVA researchers to co-design research with those with lived experience and to engage with this process at the earliest opportunity.

Domestic violence and abuse researchers much engage in some tricky ethical and methodological decision making if they are to develop more diverse and inclusive studies exploring experiences of DVA. Resultantly we call upon those conducting DVR to be more open and transparent in these decision making processes, and to publish an audit trail of their thinking and reflections particularly around the four areas examined here: power, payment, participation, and platform. We argue that part of the peer-review process for DVR should include thorough consideration and feedback on the description of participant recruitment strategies (including sampling and payment) in papers submitted for publication. We suggest this would allow those reading to be more cognisant of how and importantly why the studies employed the strategies detailed. The paper also provides a challenge to develop ethics protocols and models for participant engagement, recruitment, and payment.

Through writing this paper we have developed a series of recommendations for those engaged in DVR to consider, which we hope will inform future research design and development:

Recommendations

  1. 1.

    Engage those with lived experience of DVA in co-design at the earliest possible stage of the research design process.

  2. 2.

    Critically consider the potential outcomes of the research study and dissemination activities and the potential impact of these, be transparent about these with potential participants.

  3. 3.

    Critically consider research recruitment strategies, preferably in consultation with co-designers with lived experience. Specifically consider if proposed strategies could constitute silencing or symbolic violence and how strategies promote the inclusion of a diverse range of participants.

  4. 4.

    Give careful consideration to participant payment including the amount given and the format, victim/survivor perceptions of payment, potential tax or policy breach implications, researcher safety and prepayment. Importantly, document the decision making process so there is a clear audit trail for inclusion in publications and dissemination.

  5. 5.

    Include the decision-making process around participant payment within method sections of qualitative articles and peer review process.

  6. 6.

    Critically consider the value of including participants who are perpetrators in research studies. Include a risk analysis process, incorporating a review of legal implications, in these considerations which takes account of risk to all.

  7. 7.

    At the design stage, include methods to ensure the findings are publicly available and accessible to a range of audiences, consider how to make dissemination strategies inclusive.