Using social work theory to engage with gatekeepers in researching the sensitive topic of intra-familial homicide

Introduction

This article developed from a research project I undertook in Perth, Western Australia. It describes the method used to engage with gatekeepers in researching, intimate partner homicide, intimate partner homicide followed by perpetrator suicide, and familicide suicide (Johnson, 2008). I am an experienced social worker with a background of working in men’s prisons, child protection and the Family Court, and undertook the research for my PhD. Social workers are trained to work with sensitive issues, and are well equipped, with their unique suite of skills and knowledge, to undertake research with vulnerable populations in a range of areas such as, domestic violence (Chung, 2007), mental health (Tew, 2008), disability (Ahmed et al., 2011), and child protection (Thørnblad and Holtan, 2013).

Lee and Renzetti (1993: 5) propose a definition of sensitivity which is based on its potential for threat, which should it emerge, may cause data collection, storage and/or dissemination to be problematic to researchers and/or those being researched. They identify several characteristics of research which may lead to it being deemed sensitive. These include research involving deeply personal lived experience, research examining deviance and/or social control, research impinging on the vested interests of powerful individuals, and research on sacred issues (Lee and Renzetti, 1993: 6). Ideally sensitive issues should be considered in the research planning phase. However, some issues may be unpredictable, emerging only through the process of the research, or developing in response to external forces. Thus, as Lee and Renzetti (1993: 5–7) postulate, sensitivity in this project was not static and did not simply apply to one aspect of the research. Rather, there were layers of interrelated sensitivity requiring consideration, including the ethics associated with interviewing participants, the process of engaging with gatekeepers, and how the findings would be published, and these were complicated by the shifting socio-political context.

Both ethics and politics are about the manifold of interests and feelings-one’s own and those of others-that must be recognized, understood, and taken into consideration to achieve optimally good results. (Sieber, 1993: 14)

Roesch-Marsh et al. (2012) refer to the increasing difficulty researchers encounter in accessing data in the social sciences. These difficulties are likely to be compounded when researching sensitive topics such as homicide. My project was particularly sensitive because, in addition to homicide, it addressed several taboo subjects; child murder, female perpetrated homicide, and suicide.

By definition, intra-familial homicide occurs within the context of close and intimate relationships, forming a substantial number of the total homicides committed in many Western countries; for example, 22.4% of murders in the United States (Cooper and Smith, 2011), 33.6% of murders in Canada (Beattie and Cotter, 2010), and 39% of murders in Australia (Bryant and Cussen, 2015). It is therefore a significant social problem. Homicide usually implies a conflict between perpetrator and victim, but in homicide between intimates, police offence reports seldom indicate whether the conflict was a one off disagreement, part of an ongoing dispute, or the final incident in a relationship characterised by ongoing violence. As the core business of many social workers is working with families experiencing conflict, domestic violence and/or child abuse, (where homicide and suicide are always issues for consideration) social workers are well equipped to undertake intra-familial homicide research.

Much of the research on homicide is undertaken by criminologists and sociologists and tends to rely heavily on quantitative methods (Brookman, 2005; Mouzos, 2000; Polk, 1994; Smith and Zahn, 1998, 1999) but as intra-familial homicide occurs within the context of a close, and/or intimate relationship, qualitative or mixed methods have the potential to better inform us about long-term antecedents to these offences. This is critical to increasing our understanding of intra-familial homicide and how to prevent it. Social workers have a natural affinity for qualitative methods because a basic tenet of social work practice is to, ‘start where the client is’ (Marsh, 2002: 34). When social workers engage with clients, they begin with the client’s meaning making of their lived experience. Qualitative researchers, and those using mixed methods, prioritise the understanding of lived experience as essential to understanding issues affecting individuals, groups, and communities (van Manen, 1990). Therefore, it is not surprising that many social work researchers choose a qualitative approach, because it is consistent with social work theory and practice.

Research on spousal homicide and familicide has shown that psycho-social antecedents to homicide, obtained directly from perpetrators, and secondary victims, point to the existence of child abuse and other forms of intergenerational violence in the families of both perpetrators and victims (Johnson, 2005, 2008; Johnson and Sachmann, 2014; Sachmann and Johnson, 2014). Again, social work is an appropriate profession to be involved in researching intra-familial homicide, because social workers always view the individual within their psychosocial context (Miley et al., 2004) taking into account inter-generational patterns of behaviour and using in-depth interviews as an important information gathering tool.

However, the data pertaining to intra-familial homicide are not located in any one place, but are spread across many sites, in addition to being held by perpetrators and secondary victims. Accordingly, I collected data from, ‘Factiva’ electronic newspaper archives (DowJones, 2008), the Western Australian Police Service, the Corrective Services Division of the Justice Department, the Coroner’s Court, and the Family Court of Western Australia. I also conducted in-depth interviews with perpetrators and secondary victims. This meant five ethics approvals needed to be obtained to facilitate data collection. Social workers are trained to conceptualise systemically and to work across complex systems, using their skills to achieve ethical professional outcomes for individuals, groups, and the community (Forder, 1976; Payne, 2014). In research, they are able to use their knowledge to engage with systems and negotiate with gatekeepers to access data.

Sensitivity and vulnerability

The sensitivity of my topic was a major consideration for decision makers in the bureaucracies holding the data. Any of the taboo areas included in the research might have given rise to concerns, but as it covered several taboo areas there was an accumulation of sensitivities. Findings clearly had the potential to impact on individuals, human service agencies, and on the wider community. Some potential participants had been the first on the scene after the offence and had discovered their loved one’s body. Some were present when their child or children were killed, and some were threatened, or sustained injury at the time of the offence. These factors contributed to the sensibilities decision makers had about their participation. If participants were perpetrators of either intimate partner homicide-suicide, or familicide by definition, they had survived a suicide attempt at the time of the offence and may remain at risk of self-harm following incarceration.

Engaging with gatekeepers

When researching homicide, as with any sensitive topic, access to data is bound to be restricted by gatekeepers. The role of the gatekeeper in an organisation is to protect its function. In qualitative research, where there is an emphasis on accessing the lived experience of participants, the role of the gatekeeper may be much more complex than when dealing with written records. Researchers have found accessing sensitive data can be a difficult and drawn out process (Clark, 2011; Hayes, 2005).

Interestingly, Wanat (2008) found that the number of obstacles researchers encounter in organisations is directly related to the number of levels and gatekeepers within it. This explains the many frustrations I experienced in data collection, because I had to not only negotiate my way through five large government departments but many divisions, sections and sub-sections within them. Sometimes, my initial attempts to access information were rejected. Sometimes, I was fobbed off, or passed from one person to another, then had to wait weeks for the next appointment. Inevitably, the timeframe for the project became drawn out and threatened its completion.

Bureaucracies dealing with homicide

Bureaucracies that deal with the aftermath of homicide, have multiple sensitivities, for example, fear of: violence, criticism, perpetrator suicide, and media attention, and the ethics of privacy for both perpetrators and secondary victims. Sensitivities and the organisational vulnerability they engender can cause bureaucracies to form a shield around potential participants that is difficult to penetrate. Attempts to research the area may cause anxiety in the system, and if this is not addressed, the research application may be refused.

Sensitivities may be long or short term and exist in one part of the organisation, permeate the whole organisation, or the whole of government. At such times a bureaucratic system, just like any other system encountering threat, becomes absorbed with its own survival, making it less concerned about matters that do not have a direct benefit to it, such as research. These somewhat cyclical events can perhaps be considered normal in the long-term functioning of bureaucracies. But for the researcher who finds themselves in the middle of such a phase, they can and do present significant challenges. The threat to organisational stability at such times may cause insurmountable delays, which can at best affect timeframes for ethics approvals, and at worst seriously jeopardise the survival of the project.

Whilst my data collection took place within five different contexts, I found prisons to be the most challenging. Prisons are total institutions, ‘where a large number of like-situated individuals, cut off from the wider society for an appreciable period of time, together lead an enclosed, formally administered round of life’ (Goffman, 1961: 11).

In the project’s initial stage serious concerns had arisen in the community about security in prisons following two incidents, the escape of a murderer who remained at large, and a prison officer being sexually assaulted by a prisoner. Each incident attracted great media scrutiny, placing the department and the state government under great pressure. The Department of Justice was threatened with a major restructure and the whole organisation became inwardly focused and less attuned to the needs of outsiders. This led to a long delay in ethics approvals and a lengthy ban on researchers entering prisons.

Later, during data collection I was kept waiting in the prison one morning for two hours because the muster ¹ was found to be incorrect. A missing prisoner demands an immediate response, because an escape is likely to be a highly politicised event. So, no movement of visitors or prisoners was allowed, whilst a recount took place. By the time the prisoner was accounted for, the time allowed to conduct interviews had expired, so I had to leave the prison without interviewing a single participant. This was a major disappointment, but I chose not to complain, because I had learnt that the researcher must be as sensitive to the needs and emotional climate of the context, as they are to individual participants, if they are to survive within the organisation. In short, I had become part of the total institution and had internalised its values (Foucault, 1977: 202–203).

Appreciating the role of the gatekeeper

Researchers are dependent upon gatekeepers because they mediate the interface between the research team organisations (McAreavey and Das, 2013). Therefore, it is important for them to have an appreciation of the gatekeeper’s role, which is to maintain the interests of the organisation and avoid disturbing its equilibrium. This may sometimes lead gatekeepers to restrict access to data (Broadhead and Rist, 1976). The need for this appreciation applies throughout the project from ethics application to completion. Researchers need to remain alert and attuned to the changing interface between researcher and organisational context, as the socio-political context shifts, so that they can negotiate in the field and on the run when necessary, to ensure the research outcomes (McAreavey and Das, 2013).

Most research projects have a finite timeframe for data collection. When frustration arises due to restrictions imposed by gatekeepers, researchers may find it difficult not to let this show. However, as Mandel (2003) found it may not always be wise to let this surface, because it can negatively affect research outcomes. Twelve months into my project, after scaling many obstacles, with only two of the five ethics approvals, and faced with a dwindling amount of time, I realised that my process of engagement needed revision or it might never be completed.

The bridge from the problem to a useful methodology (Perth to Milan)

By putting my frustration aside, and beginning to analyse the problems in accessing the data, I saw that the characteristics of the organisations I was dealing with were not dissimilar to those of many of the troubled families I had worked with in the past. They were hierarchical, in some ways rigid, fraught with complex internal dynamics, and typically closed to outsiders. I wondered whether the strategies that helped me in the past might help me now, in engaging with the bureaucracies impeding my progress. Similarly to when I encountered difficulties in casework, I turned to theory to guide me. Due to my experience as a family therapist, working in prisons, I turned to a theoretical model I felt familiar with, post-Milan Systemic Family Therapy.

Family therapy grew from the work of an anthropologist, Gregory Bateson (Bateson, 1972) who linked first-order cybernetic theory to social science, using knowledge of systems and their self-correcting feedback loops to better understand patterns of behaviour in families. It led to a focus on the interactional patterns in families rather than the symptomatic behaviour of the individual. Early models of family therapy were based on the therapist being seen as objectively observing the family’s patterns of interaction from outside the system (Goldenberg and Goldenberg, 2008).

Some family therapists developed this further into a second-order cybernetics approach, where the therapist is seen as part of a system, wherein multiple realities co-exist. By 1971, the Milan Centre for the Study of the Family was formed consisting of Mara Selvini-Palazzoli, Luigi Boscolo, Gianfranco Cecchin and Guiliana Prata. These four therapists developed the ‘classic’ Milan approach, based on strategic techniques, but incorporating systemic thinking which proved successful in dealing with anorexia and schizophrenia (Palazzoli et al., 1978). The original Milan approach to family therapy stresses the importance of engaging with each interface sequentially, starting at the outside edge, and gradually moving inwards, only as trust is established, without disturbing its homeostatic tendency. Not following this procedure would risk confronting the system, possibly causing it to close up and deny access to outsiders.

Post-Milan systemic family therapy

Eventually, Boscolo and Cecchin split from the Milan team, further developing their systemic model into what was became known as the post-Milan approach, where the therapist is seen to co-exist with the client family as part of the same system. They create, ‘a hypothesis that will explain the fit of the problem in a context that has co-evolved with the problem through time’ (Hoffman, 1982). Using the concept of the time cable or cosmic sausage, Hoffman depicts three concentric rings, represent firstly, the family’s interface with the outside community including the referring system, secondly, the interface between the family and the therapists and finally, the (inner) ring represents the internal family dynamics. The concepts of circularity, neutrality, and hypothesising are critical to this approach as the therapist works to co-create a different way of the system perceiving itself, which may or may not lead to therapeutic change depending on the family structure (Goldenberg and Goldenberg, 2008). As this approach has been found to work well with so-called closed family systems, I believed it had the potential to work well with hierarchical and relatively closed bureaucracies such as the Police Department, Corrective Services, the Department of Justice-Corrective Services Division, the Coroner’s Court and the Family Court.

Using a systems approach to engage with gatekeepers

Drawing on my family therapy experience, I decided that a second-order cybernetics approach as used by the post-Milan therapists, Boscolo and Cecchin, with its emphasis on understanding the system as evolving through time, and the therapist as being not an ‘outsider’ but an integral part of the system (Boscolo et al., 1987; Goldenberg and Goldenberg, 2008) had potential. Such an approach demanded that I begin by conceptualising each system, as I would a family, viewing each within their individual context as a series of three concentric circles.

The first ring I conceptualised as representing the organisation’s primary gatekeepers, its executive decision makers, and their interface with its community context, including its historical interface with researchers and other outsiders. These gatekeepers typically decide whether research is an appropriate endeavour within the organisation at any particular time, and if so decide on a process of ethical approval including appointing secondary gatekeepers, usually in the form of an ethics committee. The second ring I conceptualised as representing the secondary gatekeepers (the ethics committees and senior prison managers). The third (inner) ring, I conceptualised as including the interface between myself and tertiary gatekeepers, the people who helped me to access data, and facilitated my interviewing perpetrators and secondary victims. These people included police officers; administrative staff; counsellors in Victim Support Services; staff in the Coroner’s Court and prison officers and programme staff, who often went out of their way to assist, by answering my many questions, checking information, introducing me to potential participants, and smoothing the way through the maze. An example of my conceptualisation with regard to the Department of Justice-Corrective Services Division is demonstrated in Figure 1. OPEN IN VIEWER

In practice, I found approaching each individual system to access data not dissimilar to approaching a complex family system, in an effort to gain information about its history and current level of functioning. I found I needed a unique approach to each agency, just as a client family requires a unique approach in the field. Beginning with the outer ring, (each department’s executive) each interface had to be carefully negotiated on its own terms, recognising the idiosyncratic needs of the organisation, whilst never losing sight of my research goals. The concept of neutrality (Cecchin, 1987; Goldenberg and Goldenberg, 2008) or in post-Milan family therapy multi-positionality (Burck et al., 1998) was relevant here as I needed to be perceived as in tune with each of the systems and their subsystems rather than on the ‘outside’. Yet at the same time, it was important to be perceived as not on any side, and not to get caught up in any internal conflicts which might hamper the research. In much the same way when working with a family, it is important to be seen by each member as empathic and understanding of their position, whilst at the same time, remaining neutral to any internal conflicts between members.

Engaging with primary gatekeepers

My first task was to explore each system’s interface with outsiders including researchers, to determine what it needed in terms of information, submissions, assurances and feedback from the research. I did this by engaging with executive staff to identify what data might be available and in what form. I found it helpful to listen empathically to their anxieties about the potentially negative impact of the research. Not only did this aid my understanding of the system, but it helped me to appreciate the efforts gatekeepers were making, in supporting the research against a backdrop of organisational change. When a negative experience of outsiders was mentioned, I would explore the effect of this on the system, demonstrate understanding of its impact and determine what I might need to do to ensure I did not repeat it. This is similar to the way I would have proceeded with a family who might have had a negative experience of helping agencies in the past that affected its ability to engage with a social worker. Once I had explored this interface, and only when I felt accepted by the organisation, would I then move on in a respectful way, towards the next interface, building trusting and non-confrontational relationships, as I progressed through the system, to achieve access to the data.

Engaging with secondary gatekeepers

My second task was to engage with the ethics committees and senior managers, who would decide whether I would be allowed to access the data, and if so, under what conditions and circumstances. Again the position I adopted was one of multi-positionality (Burck et al., 1998). As issues of concern were raised in relation to the proposed research, I would acknowledge them empathically and attempt to convey my understanding of how the individual felt about the potential harm that might be caused to the organisation. I would then engage with them in a process to examine how their concerns might be alleviated so that the research could go ahead. For example, one issue of concern raised by the Corrective Services ethics committee was that interviewing perpetrators about their offence might render them emotionally vulnerable and precipitate a suicide attempt. After listening carefully to their concerns, I was able to negotiate a safety plan with the ethics committee, where I gave an undertaking that I would advise the prisoner prior to the interview that if he became distressed I would notify relevant operational staff, so that appropriate care could be provided. This satisfied the committee and removed this potential barrier. In a similar way, I might listen to all members of a troubled family to ascertain their punctuation of the problem, and hypothesise about the ways in which their patterns of interaction indicated their attempts deal with it.

Engaging with tertiary gatekeepers

Finally, my task was to relate respectfully to each of the tertiary gatekeepers – the workers, in the many sites from which I intended to access the secondary data. Each site had rules that set the parameters for what I was allowed to access and how and when this could be done. I had to familiarise myself with these to ensure compliance and avoid jeopardising the research. I found multi-positionality (Burck et al., 1998) was again useful. It helped me to negotiate the maze of conflicting interests at this interface. For example, the need of staff to ensure safety, in the face of a client’s desire to participate as part of the healing process, whilst at the same time risking reactivation of trauma.

There were times when data supplied to me was incomplete or incorrect, or records I was advised were available to peruse had been removed. At these times, I avoided becoming oppositional or confrontational with staff, but adopted the role of the ‘naïve enquirer’ (Seligmann, 2000) to try to establish what had gone wrong and what options there might be to remedy it, whilst being careful to acknowledge that provision of the data was seen by staff as an additional task above and beyond what they might normally be expected to perform (Munro, 2005). When working with families experiencing difficulty, the role of the ‘naïve enquirer’ is helpful in eliciting each member’s perception of the problem and assisting them to consider alternative solutions.

Previously when working in prisons, I had always endeavoured to be polite and deferential to uniformed staff, to never question authority, unless the issue was one of extreme importance or security, I never ‘dobbed’ ² on staff who bent the rules. In this way, I had established a level of credibility with staff that helped me when I returned years later to undertake the research. As a researcher and visitor to prisons, I complied with and never questioned any rules related to safety and security. I made myself aware of the muster times in each prison and complied with requirements in terms of prisoner movement at these times. I attempted to anticipate security concerns about my presence and my safety in interviewing prisoners alone, by making it known that I had previously worked in prisons, was familiar with the prison environment, its language and routines, and that I understood the need for security and strict procedures for moving prisoners and visitors, such as myself, between the various areas within the institution. In writing about the total institution Foucault observed:

He who is subjected to a field of visibility, and who knows it, assumes responsibility for the constraints of power, he makes them play spontaneously upon himself; he inscribes in himself the power relationship in which he simultaneously plays both roles, he becomes the principle of his own subjugation. (Foucault, 1977: 202–203)

Consistent with this premise, I never questioned the different practices across prisons, or between different staff members in the same institution. My goal was to access participants for the study and I kept this as paramount, even on occasions when I was spoken to brusquely, or when arrangements painstakingly made in advance, dissembled before me, because of a change in personnel or a failure by someone to pass on instructions. In short, I stayed focussed on the outcomes and went with the flow of prison culture. In post-Milan terms, I simply adopted the homeostatic tendency of the system (De Shazer, 1984; Goldenberg and Goldenberg, 2008).

Accessing secondary victims

Accessing secondary victims is a very delicate process, and there are certainly issues of trust involved in this work, but the difficulties I found to be surmountable. Secondary victims by definition have experienced trauma in relation to the offence and subsequently have frequently had negative experiences of publicity. They are typically wary of discussing their experience with strangers. I also found them to be very much aware that they comprise a small and unique group in society and to be attuned to the need for more research of homicide whilst being very clear about whether or not they wished to participate in research.

Wanat (2008) building on the work of Patton (2002) sought support from ‘known sponsors’ to gain entry to schools for the purpose of data collection. As others have found in researching sensitive topics (Mandel 2003), I also found the principle of the, ‘warm referral’ (LegalAid, 2012) sometimes termed a, ‘warm handoff’ (Horevitz et al., 2015) extremely helpful in making contact with secondary victims. This is a technique often used in social work practice to refer a client to another service when that client is particularly vulnerable. It uses direct involvement by the practitioner making the referral in linking the client to another practitioner or service. Prior to and during the interview, it was necessary to maintain awareness of trauma and the complex dynamics associated with family homicide. There is often blame attributed to members of the extended family, and in-law relationships are often fractured. Publicity of the offence can also be problematic for families. I made sure before commencing the interview that there was appropriate professional support available should it be needed afterwards.

Accessing perpetrators in prison

Prisons are, by definition, closed systems and outsiders who request entry can be a nuisance to staff because of the safety and security risks they pose. Initially, I needed to convince the prison superintendents that I would not become a security risk in their prison. I used the ‘warm referral’ method again by requesting a prison superintendent, who had been involved in the ethics approval process, to introduce me to other superintendents, and encourage participation. Superintendents then put me in touch with Prisoner Programme personnel, usually psychologists or social workers, who spoke directly to perpetrators and explained the research on my behalf. They then facilitated my contact with those who wanted to participate.

When it came to contacting perpetrators, I was quickly reminded that the prison culture is one characterised by a lack of trust. Usually trust in this closed environment is earned by the individual over a long period of time by demonstrating honesty and loyalty, (in Australian prison vernacular this is ‘to be straight’ and ‘back your mates’). There is a reticence to pry into the emotional life of others, (‘do your own time’) or to disclose any of their own emotional pain and/or relationship difficulties (‘don’t bleed’). Such a culture presents an obvious challenge to the researcher, because the purpose of the inquiry is antithetical to prison culture, added to this the window of access to prisoners is time limited, so there is little time to build relationship. Furthermore, the space allowed for contact with prisoners, to interview them, is usually neither comfortable nor private and is hardly conducive to discussing a sensitive topic such as homicide. These difficulties may be overcome by being flexible and using knowledge of the prison system to work within the culture. In sensitive research there is an, ‘intricate range of relationships between research partners’ (Enosh and Ben-Ari, 2010). Nowhere is this more evident than in prisons. The traditional and lengthy way of establishing trust in jail can be circumvented if an individual is introduced by a trusted person (similar to a warm referral). This was the method I employed, again with a positive outcome. I first built relationships with programme staff, who had existing relationships with the prisoners I wanted to interview. I met with each of them individually to, provide information about the research and liaised with them about access to prisoners. Programme staff and prison officers mediate in these circumstances, between prisoners and outsiders. This crucial role of other stakeholders has previously been recognised (Enosh and Ben-Ari, 2010). Other stakeholders have the ability to influence prisoners’ participation in research, as prisoners will sometimes seek advice from staff they trust, as to whether or not it is in their best interests to participate.

Enosh and Ben-Ari (2010) propose a continuum between cooperation and conflict as existing between research partners, dependant on the degree to which they share the same moral values, social commitments and intellectual curiosity. I lost a potential participant in this way when a prisoner, who had begun the interview enthusiastically, discussed the project with an officer during the lunch break, and when he returned, withdrew, saying he had been advised not to participate in case it jeopardised his emotional wellbeing. This man had been very open in his preliminary discussion with me, and I felt sure his contribution would have been most valuable. But I had no choice but to accept his decision, recognising that the prison officer had acted on his perception of the prisoner’s needs, with no understanding of the lengths I had gone to in ensuring that risk to participants was minimised. There was no opportunity for me to discuss this further with either the prisoner or the officer, and I was reminded that acceptance, which is so interwoven with respect in a second order cybernetic approach, is an essential quality for a homicide researcher.

The following is another example of the type of problem I encountered in accessing perpetrators, which exemplifies the prison context and the challenges for an outsider. Like any total institution, the efficient running of a prison is ensured by discipline and adherence to rules (Foucault, 1977). When these are not adhered to, this may generate disequilibrium just as in a family system. As an outsider, I found that in addition to the differences between institutions and individual officers, rules could be changed between shifts. For example, on some occasions, I was provided with a personal alarm to wear, on others, this was not deemed necessary. On some occasions, I was required to sign in at the gate control desk, but not on every visit.

In one maximum security campus-style prison over six visits, I had always been allowed to walk without a two way radio, within a ‘line of sight’, ³ from the officer at the control gate, through to the block where I was to interview prisoners. Then, on one occasion, I was surprised to be confronted by a very angry senior officer, who severely reprimanded me for being without a radio, for communication with the control gate, whilst I walked across to the block (a couple of hundred metres). I was instructed to remain stationary, until I could be picked up by a vehicle and escorted back to the gate control area.

When challenged by the officer about the radio, I was initially very concerned that I may have jeopardised my good standing in the prison and that all my previous efforts to blend in and cause no problems were about to backfire. I was worried that I might be reported, and banned from future entry to the prison, which would have been disastrous for the project. To try to explain that I had been there many times before, and a radio had not been required, would have been unwise. Firstly, it would minimise safety concerns and demonstrate beyond doubt my naiveté (in prisons naiveté translates to security risk) and secondly it would ‘dob’ the officers who had allowed me to walk through without a radio. (The prison culture of not ‘dobbing’ also extends to staff). As the visitor never really knows what all the rules are, and because they may change in certain critical or emergency situations, I assessed that it was best just to maintain the homeostatic tendency (Goldenberg and Goldenberg, 2008) of the system, by not challenging the rules and by complying with whatever I was told, to keep quiet and make as few waves as possible. Many inmates of institutions adopt this behaviour as a survival strategy (Foucault, 1977; Goffman, 1961) and I too found it effective in achieving my objective.

Conclusion

By sharing the experiences gained in this project, I intend to encourage other social work researchers to recognise the importance of social work as a profession being involved in intra-familial homicide research, an area which has traditionally been the domain of sociologists and criminologists, and dominated by quantitative studies, and not to be perturbed or dissuaded from this work either by its sensitivity, or by the inherent difficulties in engaging with gatekeepers.

I hope to convey the value in using theory, so important in guiding social work practice, to guide us in achieving social work research goals. Although I chose to use a theoretical approach that I was familiar with, from my background in family therapy, there are doubtless other theoretical models that may be employed just as satisfactorily. In research, it is likely to be important for a social worker to use the theoretical base they feel most comfortable with, and which will sustain them through the inevitable ups and downs of the research process.

Finally, I hope this article engenders understanding and acceptance of the reasons why this type of research cannot be done quickly, and of the need for project time frames to be flexible to accommodate the sensibilities that characterise intra-familial homicide research and avoid them foundering. If social workers become more involved in intra-familial homicide research, using the theories that guide their practice to engage effectively with perpetrators, secondary victims, and the systems that manage the aftermath, there is a great potential for us to better understand causal factors and ultimately to prevent intra-familial homicide. The benefits that may flow from this to the wider society are incalculable.