Children exposed to domestic violence: a discussion about research ethics and researchers’ responsibilities

Abstract

Children’s exposure to domestic violence has attracted increased interest from researchers. This greater interest necessitates discussion about the methods by which children’s exposure to and descriptions of violence are studied. This article (1) discusses ethical dilemmas in research involving interviewing children exposed to domestic violence in relation to constructions of children as competent and as vulnerable, and (2) suggests a conceptual framework to aid in the design of such studies. The ethical dilemmas discussed concern: (1) research being ethically justified, (2) consent and (3) confidentiality and unsought disclosures. We suggest that combining children’s rights to agency and protection in ethical research that involves interviewing children exposed to violence can be facilitated by using the concepts of closeness and distance.

Keywords:

• childhood studies
• children’s rights
• violence
• research ethics

Introduction

Recently, children’s exposure to domestic violence has attracted increased interest from researchers, particularly research focusing on children’s own descriptions and understanding of their experience of violence as a social problem. The necessity of such research has been established by Berman (2000) and Peled (2001), among others. The last decade has seen a significant increase of such studies (e.g. Cater and Forssell in press; McGee 2000; Mullender et al. 2002; Cater 2007; Øverlien and Hydén 2009; Øverlien 2012). This greater interest necessitates discussion about the methods by which children’s exposure to and descriptions of violence are studied. This article focuses on the ethics of such studies, and has its source in our experiences as researchers conducting research on children exposed to domestic violence.

Balancing children’s competence and vulnerability

The almost worldwide ratification of the UN Convention on the Rights of the Child (UNCRC) during the 1990s represents a shift from seeing children as passive objects subordinate to their parents, to human beings with their own legal rights. During the same decade, James and Prout’s groundbreaking anthology Constructing and Reconstructing Childhood (1997) established that by including children in research, we can study them in their own right. The childhood studies paradigm has had a significant impact on researchers’ interest in children’s lives and on the theoretical understanding of childhood. Today, describing children neither as social agents nor as competent informants is original or extraordinary. In Nordic social research, the ‘new paradigm of childhood’ has become almost mainstream.

However, the UNCRC states both that children are competent and have the right to be heard (e.g. in Article 12) and that children are vulnerable and therefore have a right to protection (e.g. in Article 19). The fact that the UNCRC embraces both these perspectives is something many child researchers must take account of, because social research on childhood has come to include a significant body of research on children’s problems and vulnerable situations or situations over which the child has little control. The perspective of ‘children as social actors’ has created a field with new ethical dilemmas and responsibilities for researchers in the social research on childhood. Christensen and Prout (2002) argue that codes of ethics, reflexivity and collective professional responsibility are all needed in order to satisfy the ethical demands that flow from these newer perspectives on children as social and competent actors.

Research ethics in social research about and with children

Ethical questions, concerning e.g. how a body of material is collected and analysed, and how results are reported and used when central cultural and social values conflict (see Alver and Öyen 1998), arise in all kinds of research. However, research involving children poses particular demands, and is at special risk of triggering such conflicts (cf. Stanley and Sieber 1992; Graue and Walsh 1998; Greig and Taylor 1999). This is partly because children have a special position in relation to adults. As minors, their welfare is the responsibility of adults. These adults may be their legal guardians as parents, other primary caregivers, or staff at an institution. This also means that children often are dependent on adults making wise and altruistic decisions.

Children also have a special position in relation to research and researchers. Mahon and others (1996, 149) argue that ‘with adults, the interviewer can be matched to the person being interviewed on a range of social characteristics such as age, gender, or ethnic origin, in order to create a relationship in which the person being interviewed feels rapport and identification with the interviewer’. They continue by stressing that this is often difficult or impossible with child informants and adult researchers, because of the obvious differences in age and, often, also in authority. Some have suggested alternative research approaches in which children, rather than being considered objects of research, are involved as co-researchers alongside adults (e.g. Fern and Kristinsdóttir 2011; Kellet 2011). However, there has been little discussion of how researchers might manage the power discrepancies related to children being subordinate to adults, an issue which is especially salient when it comes to domestic violence.

Furthermore, issues related to informed consent and the interpretation of informants’ accounts are more complex with children than adults. Research involving children in vulnerable situations comprises even more sensitive ethical dilemmas, and when violence is involved, special circumstances that influence and complicate research also apply. Researchers such as Ellsberg and Heise (2002) and Fontes (2004) have discussed domestic violence research in general, while Becker-Blease and Freyd (2006) describe research with abused children and Mudlay and Goddard (2009) discuss their own study about children exposed to domestic violence. However, the increasing interest in using qualitative interviews with children exposed to domestic violence calls for continuous discussion about the general ethics of qualitative non-clinical research with children in these life circumstances. Christensen and Prout (2002) suggest that in order to develop ethical practices for the future, dialogue is needed as a means of collectively sharing experience, both among researchers and between researchers and children participating in the ongoing research process. We hope that the discussion below will contribute to the dialogue about the ethics of research involving children exposed to domestic violence.

Aims of the article

The aims of the article are (1) to discuss ethical dilemmas in research involving interviewing children exposed to domestic violence¹ in relation to constructions of children as competent and as vulnerable, and (2) to suggest a conceptual framework to aid in the design of such studies.

The first dilemma concerns whether an individual research project is ethically justified at all. The expected gain for evaluations of social interventions and other applied research is often apparent. However, if we are not given the possibility to understand children’s thoughts concerning experiences of violence in a wider sense as well, we cannot fully understand their actions or reactions, or the processes that prevent children from suffering from it and that support their recovery. How can such information be obtained at minimal cost or risk for the participating children?

The second dilemma concerns the issue of consent. Regulations and legislation about when children’s and legal guardians’ consent is needed for a minor to participate in research vary between countries. However, because of the nature of the topic, researchers investigating children’s exposure to domestic violence must always consider the benefits and drawbacks of requiring consent of the child, the victimized parent and the perpetrating parent. Further, adults and children may understand and define situations differently and have different reasons for consenting or not. What do we as researchers need to ask ourselves when obtaining consent from the children’s legal guardians and the children themselves?

The third dilemma we discuss concerns how to handle the issue of confidentiality in relation to the risk of unsought disclosure of child maltreatment. The child has a right to speak in confidentiality and to be treated with respect as an agent in his/her own life. On the other hand, a researcher cannot dismiss information about a child’s safety being at risk. How can this be balanced? Moreover, how can the child be protected from negative consequences of over-disclosure yet at the same time be treated as a competent informant?

These issues are discussed below in the light of the theoretical paradigm of childhood studies and the (more traditional) view of children as vulnerable and in need of protection.

Ethical value of research

The primary ethical issue is to weigh the value of the possible knowledge gain of a study against its possible harm to the participants. According to the first international document on research ethics (World Medical Association Declaration of Helsinki 1964), the subject’s welfare must always take precedence over the interests of science and society (Article 5). In other words, our primary obligation as researchers should always be to the people we study, rather than to our research project or our discipline (Denzin 1989). Research can only be approved if the risks for participants are outweighed by the scientific value of the expected results.

When it comes to particularly vulnerable groups, studies can be legitimatized by the possibility that they will contribute to improving support for these groups (Alver and Öyen 1998). Therefore, research is often planned to maximize the possible general knowledge gain while limiting the possible harm to individuals. However, research focusing on children’s own understandings or descriptions of experiences of violence often entails an interpretive or qualitative approach. Such research is often not primarily directed towards practical outcomes or solving problems in the social services. The main ethical risk associated with ‘pure research’ is that the benefits of research cannot be judged beforehand, and because the risks of harm might be more obvious than the long-term benefits, an entire project may be ‘unethical’. If the aim of the study is theoretical development, it is motivated by the absence of existing theory and ‘interest’, and its applicability to social or intervening practice can only very roughly be estimated in advance. In such studies, it is difficult – if not impossible – to predict both benefits and possible short- and long-term risks (Alderson and Morrow 2004). Thus, studies intended for theoretical development that include interviews with children about their exposure to violence require clear and explicit objectives. They may be motivated by the need for more nuanced understanding of how children deal with such experiences or of their understanding of their living situation. Such knowledge might also form a foundation for the development of support practices. In the case of children exposed to domestic violence in the Nordic countries, such research has motivated politicians and practitioners in recent decades to start developing support interventions for these children. Today, however, when interventions are more widely available, studies in which children are interviewed about their exposure to violence – but that lack obvious applicability – might need other justifications to fulfil ethical demands.

Therefore, to be able to answer questions and develop practice, researchers need children. And because we can hardly find out how to help children if we cannot understand their experiences of violence, their thoughts, fears and actions, children also need research. The argument that children are experts on their own experience and that both theoretical development and practice must accordingly be developed on the basis of children’s accounts derives from the discourse of the competent child. But, at the same time, if the research in any way – even indirectly – is related to developing practice, then it also encompasses the vulnerability of children who need such support. In many of our encounters with children and adolescents exposed to violence, the children have told us that they participate in research to help improve the situation for other children who may be exposed to violence in the future. Thus, research can be an empowering experience for children. Children and adolescents can also support each other during group interviews, which may have empowering effects. However, if children’s grounds for participating in research are that they hope through contributing to improve the lives of other children, the researcher has a great responsibility. As a researcher, you are often neither responsible for, nor aware of how the findings will be received and the purposes for which they may be used. This may therefore need to be clarified to the child.

Bischofberger and others (1991) stress that research that does not have a direct positive outcome for the informant only can be justified if the child’s safety is warranted. Whether interviewing children exposed to domestic violence is safe can be related to whether the children are seen as a ‘clinical’ population. Children in studies about domestic violence are often ‘in between’. On the one hand, they are often recruited in institutional settings where they may be living because of difficulties at home. Further, experiencing domestic violence as a child has been found to be a strong risk factor for behavioural and emotional problems among children (Edleson 1999a; Morrel et al. 2003; Wolfe et al. 2003; Turner and Finkelhor 2006) and for being directly abused (Edleson 1999b; Avery, Hutchinson, and Whitaker 2002; Cox, Kotch, and Everson 2003). On the other hand, some research indicates that many children handle such experiences well (see Edleson 1999a; Stith et al. 2000; Wolfe et al. 2003) and Edleson (1999a, 866) stresses that defining the witnessing of violence as maltreatment ‘ignores the fact that large numbers of children [show] no negative development problems and some [show] evidence of strong coping abilities. [It may also] ignore battered mothers’ efforts to develop safe environments for their children and themselves’. Further, research has shown that children actively deal with or even try to stop the violence (McGee 2000; Mullender et al. 2002; Edleson et al. 2003; Broberg et al. 2011; DeBoard-Lucas and Grych 2011; Øverlien 2012). Thus, seeing these children as a ‘clinical population’ may stress their vulnerability and obscure their competent agency and expressions of resistance. On the other hand, when violence is an issue, the safety of the child and victim is a prerequisite for research to take place. It may therefore be necessary to use the clinical setting to promote their safety, though a clinical setting is no guarantee of safety.

Since the topic of children exposed to domestic violence by definition also includes other individuals – who are likely to see the problems very differently – the risk-and-benefit calculations must include them as well. Hence, the ethical considerations in research involving children exposed to domestic violence concern relationships, possible differences, or even conflicts of interest between at a minimum the children (both as individuals and as members of a social group), the victims of violence the perpetrators, and society’s need for research findings. Before returning to our suggestions for how to think about this, however, we wish to discuss another core aspect of research ethics, namely consent.

Considering consent – whose and how?

Generally, the informants’ consent is required for research to be regarded as ethical. When research involves children below 18 years of age, parental consent is generally considered crucial (Morrow and Richards 1996). However, the regulations and legislation about this vary internationally depending on how children’s competence is defined, which creates a varied field of different situations in which some circumstances allow for children to provide consent themselves and others require parental consent as well. As an example, if the respondent has reached a certain age (e.g. 15, 16 or 17 years) and understands what the research implies, it may be sufficient to inform the child about the research and obtain his or her consent to participate. On the other hand, although it is difficult in practice to pursue any research against a child’s will, children are not always informed in such a way as to make their meaningful consent possible. Thus, to some extent the competent child discourse presupposes that children are capable of telling us researchers whether or not they want to be research subjects.

Some circumstances may make the issue of consent particularly complicated in research involving children exposed to domestic violence. On the one hand, talking about experiences of violence is a sensitive matter; these children may never have spoken to others about their experiences and may feel, or have been told by adults, that they should keep it secret from outsiders. Therefore, initial contact may need to be made with the children’s legal guardians. Their consent may also help the child feel at ease in talking about the issues in an interview situation (see Øvreeide 1998). Peled (2001) suggests that ethically sound research on children’s exposure to violence demands both children’s and adults’ informed consent, and that no child should participate if a parent with whom he/she has ongoing contact is known to object or suspected of objecting to the child’s participation. On the other hand, the topic of violence is complicated because the definition of violence is socially constructed (Hearn 1998, 15). A specific act may from the victim’s perspective be (very) negative, while from the perpetrator’s it is justified (cf. Finkelhor et al. 1983). Thus, the victimized parent and child might have different interpretations about what the other parent has done. Therefore, if the perpetrator and/or victim is the child’s legal guardian(s), the child’s participation may be obstructed either by a perpetrator not acknowledging the violence as such, or by a victim or other adult wanting to protect the child from such issues. Demanding consent from only the victimized parent may put the child at risk when/if the perpetrator finds out about his/her child’s participation in a research project focusing on what he/she has done. On the other hand, a child could also be put at risk by a researcher asking for consent from the perpetrator.

Here, the notion of the socially competent child actor can be related to the child’s right to be informed in such a way that he/she can give consent and be treated with integrity, speak freely and be listened to by the researcher, without having to be concerned about his/her parents’ possible reactions.The notion of the vulnerable child can be used to highlight the child’s right to protection from parents’ restrictions on the child’s right to speak freely, and from negative consequences of participating in the research.

However, what we also see is how the vulnerable-child discourse interacts with the researchers’ organizational and legal position. Legislation in most countries is based on the assumption that it is the child’s guardians, usually the parents, who have the right and obligation to take responsibility for and to represent the child’s best interests. However, in domestic violence research, parents may be problematic gatekeepers, as they may ‘gatekeep’ their own secrets or privilege their own understanding of situations over the child’s welfare and rights. Furthermore, parents who do not see their children as having a right to participate and who do not see that the participation may have positive empowering effects may refuse to provide consent, a problem that should be regarded as an issue by every research ethics committee (Balen et al. 2006).

If adults deny a child the right to participate in research, the research may be invalidated, and the possibilities of developing appropriate help for children limited. On the other hand, when children give their own consent, they may not be able to judge how they may be affected in the longer term. Furthermore, some researchers argue that all children should have the opportunity to object (e.g. Wendler and Shah 2003). Others argue that this does not respect children’s autonomy, but instead restricts it, to the benefit of other parties in the research, for example the researcher (e.g. Halila and Lötjönen 2003). Halila and Lötjönen (2003) also argue that relying on age limits for consent reduces the personal responsibility of researchers to evaluate the developmental stage of each child participating in the research, and they claim that children can make valid decisions if they are given adequate, non-coercive information about research in language they understand. They can then weigh the pros and cons in relation to their own perspectives and experiences. But, it is also up to researchers to be sensitive to children’s attempts to communicate that they do not wish to participate in a planned activity by not showing up for a scheduled appointment, or being reluctant to speak (see for example Evang and Øverlien, in press; Mahon et al. 1996). Such sensitivity can be promoted by informing child participants that their participation is voluntary and can be terminated at any time, and by negotiating ‘ongoing consent’, that is, ‘being sensitive and responsive to any negative reaction the children might have to being observed and recorded’ (Flewitt 2005, 556) in each contact and situation on a minute-by-minute basis. Sensitivity to any negative outcomes or reactions to the research situation can also benefit from discussion with those (parents or staff) who have more knowledge about the child and his/her behaviour.

While for some groups of children participation in research may be considered challenging but intrinsically desirable (cf. Cocks 2006), we argue that for others – such as children who have been exposed to domestic violence – it is essential for consent to be truly voluntary and obtained through a process sensitive to the problematic nature of the subject. Another issue which must be dealt with throughout the research process and which is related to the information given in the consenting situation is confidentiality.

Handling confidentiality and unsought disclosures

Once consent has been given for a child to participate in research, that child has a right to speak in confidence and secrecy in the interview context, and this is also essential for research validity. On the other hand, Hermerén (1996, 231) stresses that the researcher has a responsibility to discontinue a study if the knowledge gained is not in proportion to the discomfort the study is causing the participants. Furthermore, in many countries, if the researcher receives knowledge about circumstances that require action to be taken for the child’s welfare or safety, the researcher has responsibility as a witness and/or a duty to report.

For children who have experienced violence, this dilemma is particularly complex. First, the presence of one kind of violence is significantly related to other kinds of violence and to violence directed at other victims (Edleson 1999b; Avery et al. 2002; Cox, Kotch, and Everson 2003). This means that research about one form of violence may also generate information about other forms of violence. This entails a risk that interviews focused on children’s accounts of violence directed toward the mother might uncover violence directed toward the child, which means that the child may be at risk and the researcher’s promise of confidentiality vis-à-vis the child would then have to be broken.

This also entails a risk that situations may develop in which protecting the child means harming the family as a social unit. ‘Divergent data are common in family research because of the multiple perspectives that derive from having different positions and occupying different roles in the family’ (Sands and Roer-Strier 2006, 249). Therefore, we argue that it is the researchers’ duty to protect not only informants, but also other family members, and the family itself from harm.

These different perspectives make the issues concerning unexpected disclosures, confidentiality and child safety even more problematic in child research about domestic violence. Alver and Öyen (1998) stress that each unique situation demands a unique decision, and we agree that such an approach is the only way to maximize confidentiality and safety in research with children about violence. Since researchers have a duty to report in several countries, we can never guarantee child informants complete confidentiality (cf. Williamson and Goodenough 2005) and the issue of confidentiality is thus closely intertwined with unexpected disclosures in research.

In practice, different procedures might be appropriate with regard to breaking confidentiality. As decisions about whether to break confidentiality often are very delicate, researchers may turn to colleagues and/or professionals working with child protection for advice and consultation. Such consultation can take place without compromising the child’s anonymity.

But, the questions of when to take the decision to break confidentiality and how to go about doing so are not the only aspects of how to handle this in an ethically defensible way. Children exposed to domestic violence often have extensive experience of not being listened to, of not being asked about their preferences or wishes, and of having to adjust to adult decisions taken over their heads, and we therefore consider it particularly important to inform these children about every decision regarding breaches of confidentiality. This means making it clear when the study if first explained to the child that confidentially may need to be broken if there are serious concerns about harm. The child will therefore be able to keep this possibility in mind when making his/her decision whether to participate and/or when responding to research questions. It might even be justified to remind the child about this during data collection, even though this involves a risk that the child will hold back information. When informing the child about this, it is important to stress that the reason for any possible breach of confidentiality would be to protect the child from harm. Including information in the recruiting documentation about the possibility that confidentiality might not be maintained if people are found to be at risk and stressing that this is a standard procedure, may help children feel that the researcher is looking after their best interests rather than betraying them.

Discussion

Based on our experience of research on and with children exposed to violence, and in light of the discussion above, we conclude that children can be interviewed about any topic, but not for any purpose and not under any circumstances. Children can be competent yet still vulnerable in certain situations, and they are generally not responsible for their living conditions. Therefore, relying on participatory models from disciplines that do not focus on children in vulnerable life situations limits the development of a participatory ethos in social work research (cf. Healy 1998). Because of this, we have argued that research involving children in vulnerable situations requires that special attention be given to research ethics; researchers must beware of adopting a focus on ‘obtaining ethical clearance’ or ‘child access’, and instead take full responsibility for children’s participation and protection.

We do not believe that acknowledging the fact that children’s exposure to violence places them in a vulnerable situation necessarily means that they are in a ‘state of oppression’ (cf. Kitzinger 1997). Children living away from home because of violence, such as children in shelters for abused women, are in exceptionally vulnerable situations in which they may encounter researchers interested in finding out what their experiences of violence may mean to them. We agree with Lindsay (2000) that there are concerns regarding the research process that are specific to children, and that it is a researcher’s duty to ensure that the degree of intrusion in children’s lives is minimal. As researchers, we believe it is our obligation to ensure that the research experience is not invasive or in any way exploitative. Children have a right to be protected from any research situation that could be regarded as intrusive. We argue that this is especially true when it comes to children exposed to violence.

How can we minimize the risk that, despite our best intentions, an interview with a child exposed to violence harms the child? Christensen and Prout (2002) and Komulainen (2007), among others, propose that researchers should develop a set of strategic values in their everyday practice, and that reflexivity should be one central strategy for conducting ethical research. ‘But what is reflexivity? An honest answer to this question might be that it is something easier to do than define’ (Davis, Watson, and Cunningham-Burley 2000, 201). Let us therefore suggest a specification of the content of such reflexivity. First, we believe that ensuring reflexivity means that the researcher does not only apply rules in a routine way, but (also) takes active responsibility. Second, such individual responsibility (supplemented by support from research ethics committees) demands empathic understanding of each child’s changing situation during the research process. Such empathic understanding requires closeness. To understand a child’s everyday life, we must (attempt to) step into the child’s world and treat each child as an expert on his/her own life. However, closeness takes time to develop, and if we are not prepared to take on a very long commitment, an ethical research design should also imply distance. All research eventually comes to an end and so does the researchers’ responsibility. For this reason, research also involves distance, distance for theoretical reflection that goes beyond the data and the informants’ understanding to enable the development of theory and hence the generalizability of knowledge (cf. Silverman 2000), which is the justification for research. Taking responsibility for the child’s well-being during his/her participation in research demands distancing oneself from the child’s immediate world. Furthermore, to ensure an ethically sound distance in research with children exposed to violence, the boundaries concerning one’s role as a researcher need to be clear. It is crucial to clearly delineate the boundary between research and therapy (Mahon et al. 1996). Although it may be tempting to take on the role of a counsellor or therapist when confronted by a child in need of support, especially if one has played that role in different professional settings, it may be a better solution for the researcher to refer the child to an outside therapist. In conclusion, we suggest that combining children’s rights to agency and to protection in ethical research involving interviewing children exposed to violence can be facilitated by using the concepts of closeness and distance in structuring and designing research involving interviewing children exposed to violence.

In practice, we suggest that the researcher needs to shift back and forth between understanding the children’s perspectives, which demands closeness, and taking their vulnerability seriously as a responsible adult, which demands distance. Taking such responsibility includes closely monitoring how participating children understand the aim of research so that they do not wrongly believe that they may contribute to developing services if this is not actually the case. In fact, we argue that the rationale for the research should itself be valid from the perspective of children participating in the study. The consequences of this are that the aim of a research project is only fully ethically justified if the participating children judge it so, and that if the children have not truly understood the purpose of the research, their consent has little value. Closeness to the child is fundamental to explaining the research in this way and being able to judge whether the child has understood. Furthermore, we suggest that the only way to handle the delicate task of balancing the need for confidentiality with ethical handling of unsought disclosures is to alternate between getting close to the child’s world and backing off to be able to make a professional judgment about how to ensure the child’s safety without violating the child’s integrity. Finally, we propose using available support systems for those children who need help after revealing information about violence. However, we caution against the assumption that a clinical setting is a warrant for the child’s safety.

Let us now return to the question posed in the introduction. How can children’s right to protection and their right to agency be combined to achieve ethical social work research? Sandbaek (2004) argues for an understanding of children as agents that includes their autonomy while still acknowledging their need for care. We argue that a similar perspective may apply in the case of research involving children exposed to domestic violence. On the one hand, children must not be viewed as so vulnerable and in need of protection that they cannot participate in research. We need their knowledge, and they need us to build theories and develop practice on the basis of their knowledge. At the same time, special attention must be given to ethical concerns when it comes to children and research. In line with Ellonen and Pösö (2011), we argue that practices and methods need to be developed that are sensitive to children’s rights and needs. The general methods we use in the social sciences can be used to study children, though with some modification. Children have additional rights beyond the standard ethical concerns researchers always need to take into account, and the design and conduct of such research also need to be modified depending on variables such as age, developmental stage, gender, ethnicity and social class. In fact, the child’s age plays an especially significant role in a number of issues discussed in this article. Informing or interviewing a pre-school child in an ethically sound way may require specific techniques, which are not required when informing or interviewing a teenager. Moreover, research ethics must not be reduced to a number of principles to be handled routinely. What is needed in this kind of research is ‘reflective responsibility’ on the part of all the adults involved, not least the researcher. We must ensure at every step of the way that our research is not being carried out unnecessarily and that it is based on respect for the competent yet vulnerable child in our hands. Gaining access to children as informants is not easy – and should not be. Because adults have responsibility for the children, researchers – whose relationship with the children focuses on information the children have and the researcher needs – must weigh the benefits of the parent’s close knowledge about and responsibility for the child against the risks of the parent giving priority to his/her needs or preferences over those of the child in each individual case. After all, the parents have the legal responsibility for their children and their opinion about their child’s participation must be taken seriously.

There is no recipe for how to conduct ethical research with, for and on children. The issues surrounding ethically justified research, consent, confidentiality and disclosures remain less than precise. Furthermore, children are not a homogenous group and should not be treated as such. We must handle the specific conditions of each individual child in the context in which they arise. However, as child researchers we must, at every step of the research process, reflect on our own ethical stance. We need studies about the benefits and harms research procedures might involve for child participants (cf. Runyan 2000); clearer guidance for researchers, child participants and parents about the limitations of confidentiality (Williamson and Goodenough 2005); and finally, to pursue ongoing discussions about the ethics of research involving children exposed to violence.

Notes

1. The violence we discuss may be physical, psychological and/or sexual and be directed toward the child or someone close to the child, such as a parent.