Involving children and young people in research on domestic violence and housing: re-visited

ABSTRACT

Children’s and young people’s opinions and experiences are important to listen to, as they offer perspectives that adults might not be aware of otherwise. Yet children are often viewed as a vulnerable group in need of protection, with adults talking for them instead of letting them speak for themselves. Sometimes this might be the correct decision. However, it is also important to let children and young people participate in research on their own terms, to identify, for example, what kind of support they might need in relation to problems they have experienced. Based on previous research, we revisit this topic and discuss it’s relevance today, once again asking the question of how best to involve children in research in order to hear their views and opinions on matters that concern them. We base our experiences on a research project examining what sheltered housing means for children living there.

KEYWORDS:

• Children
• young people
• gatekeepers
• violence
• domestic violence shelters

Introduction

Almost 20 years ago, Helen Baker (2005) published an article on how to involve children and young people in research on domestic violence and housing. She argued that there is a lack of research involving children, and that this has resulted in a lack of support services for children and young people. Throughout the paper, she discussed these difficulties in relation to a project she was conducting to demonstrate the importance of involving children and young people. Although almost two decades have passed since the article was published, and the amount of research that involves children has increased, there are still difficulties surrounding the involvement of children and young people in research, especially when it comes to research about different forms of violence.

In the present article, we will revisit the themes presented by Baker (2005) and discuss their relevance today, and once again ask the question of how to involve children in research in order to hear their opinions on matters that concern them, especially in relation to sensitive topics. This will be examined in relation to a research project we are conducting about what sheltered housing means for children living there, and in our discussion, we reflect on Swedish legislation and the United Nations (UN) Convention on the Rights of the Child (1989), UNCRC.

Policy and legislative changes since 2005

Only 60 countries in the world have legislation that prohibits physical punishment of children, or in other words violence against children (UNICEF 2017). However, around the world efforts have been made to improve the rights of children and to end violence against children (Wessells and Kostelny 2019). This includes a variety of initiatives, including local, governmental, and transnational initiatives. For example, in Uruguay primary school teachers are now being trained to detect, report and support children who are exposed to violence, in Bosnia and Herzegovina the first network of Non-Governmental Organisations (NGOs) regarding violence against children has been created, and in Vietnam the Government has introduced a National Plan on Child Protection to address violence against children (UNICEF 2014). Another example is the UN Sustainable Development Goals, as a collective effort for achieving human wellbeing, equality, and a sustainable future for all people. Still, research shows that despite positive changes in regulations, legislation and support regarding children who are exposed to violence, significant challenges still exist and that the rights of children experiencing violence often remain unprotected (UNICEF 2021; Wessells and Kostelny 2019).

In Sweden, where this study is based, two major legislative changes in the last two decades have strengthened the rights of children and their ability to make their opinions heard. First, the United Nations Convention on the Rights of the Child (UNCRC) has been incorporated into Swedish law. Sweden ratified the UNCRC in 1990, but despite it having become legally binding, it has been criticised for not being sufficient to secure children’s rights. Therefore its incorporation into Swedish law on 1 January 2020 was intended to contribute to making the rights of the child even more visible, as a way of creating a basis for a more child rights based approach in all public services and authorities. Secondly, on 1 July 2021, a new definition, Violation of a Child’s Integrity, was enacted in law, according to which not only children who are subjected to violence, but also children who witness intimate partner violence (IPV), are identified as victims in their own right. With the new law, children are given increased legal protection and receive the status of victims of crime, with all the rights that accompany that classification.

In addition to these new laws in Sweden, chapter 6 § 11 in the Social Services Act (2001:453) also states that social services have a responsibility to ensure that victims of crime receive the help and support they need to process their experiences. The paragraph was first introduced in 2001 with the new Social Services Act, but has been revised several times. For example, in 2005 a clarification was made that children who witness violence in their families have a right to support, and that social services especially should recognise this group and their needs. In 2007, the paragraph was revised to read that social services shall make sure that children receive support after victimisation. This again is signal from the government that these children’s experiences and need of support require a special focus, which also means that society needs more scientific knowledge about how children experience their situation and how to best support them.

Children’s experiences of violence and domestic violence shelters

International research shows that one in four children under the age of five live with a mother who is a victim of IPV (UNICEF 2017). In Sweden approximately 10–15% of all children have witnessed physical violence between parents and 30% verbal conflicts (Cater et al. 2015, Jernbro and Janson 2016). Still, research on children’s experiences of IPV is sparse, in Sweden and around the world, especially in relation to what the children themselves express. The same also applies when it comes to children’s experiences of living in domestic violence shelters. However, the knowledge about both these areas has increased, and it continues to do so.

Children’s experiences of IPV can vary and to be a witness of violence can create as much harm for the child as being the direct victim of a parent’s violence. Witnessing violence between parents or being the direct victim of a parent’s use of violence can have both short term and long-term negative effects (McTavish et al. 2016, Vu et al. 2016), for example depression, posttraumatic stress syndrome (PTSD), problems in school and social relations, and externalising as well as internalising behaviour and violence (Gilbert et al. 2009, McTavish et al. 2016, Vahl et al. 2016). Children who experience IPV are also at greater risk of being victims of other forms of violence (Turner et al. 2016) and to be aggressive or use violence themselves (Holt et al. 2008). The violence can also be life threatening (Gilbert et al. 2009).

For children who must flee with their mothers to domestic violence shelters, the situation is most often life threatening, and still research is scarce. The research that does exist shows that children usually experience the safety and freedom from violence at the shelter as something positive (e.g. Ornduff and Monahan 1999, Vass and Haj-Yahia 2020, Øverlien, 2011b). Upon arriving at the shelter, children exposed to violence often exhibit behavioural problems that can initially increase and then later decrease, a phenomenon that is said to be related to the shelter being a safe place where the children can externalise their emotions about the violence (Fredland et al. 2014). It is also known that a mother’s concern for her children’s health and safety and the presence of children in a relationship can affect her decision about whether to stay or leave a violent relationship (Zink et al. 2003, Rhodes et al. 2010). Arguments for leaving include the mother not wanting to put her children through the experience of living with the violence and enduring the consequences it can entail. The children’s safety is also a factor that mothers rate highly when approaching domestic violence shelters for support and protection (Jonker et al. 2014, McFarlane et al. 2016).

A stay in a shelter also affects the lives of children greatly. For example, research show that children’s schooling is adversely affected by living at a shelter, and their absence from school can sometimes last for a long time (e.g. Øverlien 2011b, 2012, Selvik et al. 2017, Bracewell et al. 2020, Vass and Haj-Yahia 2020). The children’s social relationships are also affected, because for safety reasons some cannot be in contact with friends and family who could support them (e.g. Ornduff and Monahan 1999, Øverlien 2011a, Vass and Haj-Yahia 2020). Children are also sometimes separated from their sibling due to shelter rules or the family situation (Øverlien 2011b). It can also be hard to build a new social network while at the shelter, as children might not be allowed to tell people outside the shelter where they are living, and the people they meet at the shelter might move and thus might not be able to maintain contact. In summary, experiencing violence and moving to a shelter are both hard for children, but children usually appreciate the safe haven that the shelter can be. However, the difficulties surrounding the involvement of children and young people in research in order to get their own perspective on, for example, violence and the experiences of living at domestic violence shelter still need to be addressed today.

Article 19: what sheltered housing means for abused children

The project sets out to investigate children’s and young people’s (0–17 years of age) experiences of living at a domestic violence shelter, or their experiences of being denied a placement when fleeing violence. Interviews were conducted with children aged 7–17, and for ethical reasons related to age, with the mothers of children aged 0–6. Ethical approval was obtained from the Swedish Ethical Review Authority (Dnr. 2020:04561) before the project started and has been updated on two occasions regarding informal consent from the father (Dnr. 2021–03242) and creative methods (Dnr. 2021–06928–02) discussed below. The cut-off at age six was not something we as researchers decided, but instead something that the Ethical Review Authority demanded for approval, arguing that the information given by younger children would not be sufficient in relation to possible harm their participation might cause. However, children mature at different speeds, and therefore have different abilities to express their opinions and experiences, and when involving children, it is always a balance in relation to the individual child’s situation. Also, we do not capture all children’s opinions directly. Instead, some experiences are gathered by proxy, using their mothers’ opinions of the children’s experiences on behalf of the children.

The recruitment of participants was done through domestic violence shelters, and shelter staff were sent information about the project. If staff agreed to help with the recruitment, they were asked to provide (oral and/or written) information about the project to mothers and children. The reason for going through the shelter staff was that the well-being of the mothers and children was a priority. The staff had greater knowledge about the families staying at the shelter, and they were able, for example, to determine if someone’s participation could potentially endanger their security or coping processes. Information about the project was also distributed through press releases, on the university’s website, and in interviews with local and national media outlets.

At the time of writing this article, a little more than one and a half years have passed in the data collection phase, and the project is still ongoing. So far only 20 interviews have been conducted, 14 with mothers and six with children. More children have expressed an interest in participating, but they either changed their minds or moved to a new address, or it was not possible to conduct the interview because during the first eight months of the data collection consent was required from both legal guardians. This illustrates some of the difficulties that surround projects involving children.

Involving children in research – revisiting Baker

Children and young people who have experienced domestic violence, and either have or have not lived at a domestic violence shelter, can be considered a vulnerable group. Research involving sensitive topics or vulnerable groups is often closely scrutinised due to the potential risks for participants. This is done even though in many cases it can be beneficial for the participants to participate, as a way for them to process their experiences (e.g. Kavanaugh and Ayres 1998, Decker et al. 2011, Alexander et al. 2018, Thunberg 2022). Nevertheless, several ethical aspects need to be considered to ensure that the participants are not exposed to unnecessary risks. As mentioned, we will re-visit the themes presented by Baker (2005) and discuss their relevance today, and once again ask the question of how to best involve children in (sensitive) research. This will be examined in relation to our experiences from our research project on what sheltered housing means for children living there.

Researching sensitive topics with children and young people

Research involving children and young people is often considered sensitive due to the participants’ age, and some research topics (for example violence) may be considered even more sensitive, as most people want to protect children and young people from harm. However, this does not mean that sensitive research should never include children and young people (e.g. Morrow and Richards 1996, Cater and Øverlien 2014, Källström and Andersson Bruck 2017), and excluding them from participation can perhaps be due to a misguided desire to protect children but which overlooks their right to be heard (Hyden 2021). McCafferty (2017) also argues that the dominant view of children as vulnerable and in need of protection contributes to adults and professionals making decisions for them, rather than with them, which weakens children’s rights.

Research also shows that children participating in research most often experience it to be beneficial, both as a way of processing their experiences and as a way to help others by contributing to research (Kassam-Adams and Newman 2005). This however does not mean that participating is easy, as it can bring up emotions and memories that are buried within the child. For this reason, the researcher has a responsibility to read the child’s verbal and non-verbal signs of distress (Lewis 2010), and the researcher needs to reflect on their way of interviewing and how to pick up signs of distress to protect the participants from potential harm (Källström and Andersson Bruck 2017, Thunberg 2022). This can be done by easing into a story, building a rapport with the participants before asking the more demanding interview-questions. Research also shows that it is important for children to feel safe in the interview situation, to trust the interviewer, to know that there will be no negative consequences for the child or others, and that the theme of the interview is clear to the child from the start (Jensen 2012).

One of the main reasons for involving children and young people in research is to investigate their experiences and listen to their opinions. However, children’s and young people’s experiences are often expressed by adults, such as their parents or involved professionals. This means that the children’s and young people’s experiences are filtered through other people (cf. Baker 2005). This could be beneficial in cases where participating in research may harm the child or young person; however filtering experiences through another person risks changing the message that the child or young person wants to convey, because it is interpreted by another person before it reaches the researcher. For this reason, when possible, children and young people need to be able to participate in research and express their experiences in their own words (cf. Morrow and Richards 1996, Källström and Andersson Bruck 2017). This is also in line with the UNCRC (1989), which stresses the importance of children being able to participate in all matters that involve them, and having their opinions heard and respected.

Suitable methodology

As Baker (2005) highlights, an important aspect of conducting research with children and young people is the need to develop methods that enable and facilitate children’s participation and the various ways that children express themselves. Examples mentioned by Baker are interviews, focus groups, vignettes, and drawings. Focus groups, for example, might even out power imbalances in relation to the adult researcher, as the group can act as peer support for the child (James et al. 1998 cited by Baker 2005). In our study, we (like Baker) used semi-structured interviews. During the interview situation we also gave the children the opportunity to draw or play with toys or their teddy bear, as exemplified below by a drawing made by one of the children in the project. The purpose was to let the children do something they might enjoy while at the same time talking about difficult subjects. The drawings also acted as a way of shifting between the difficult subject being discussed, and what they were drawing, making them feel comfortable in the interview setting. For example, the picture below was something the child had just learned to do in school, that he wanted to show the interviewer (Figure 1).

Figure 1. Picture drawn by one of the interviewed children.

Along the way, however, we have expanded the semi-structured interviews to include the possibility of using what we call creative methods, such as writing or recording the child or young person’s narratives. This can be done alone or with the help of the researcher. The children have also been able to choose to have their mother or another ‘safe person’ (e.g. a sibling or shelter personnel) accompany them during the interviews or help them with their creative methods/narratives. The use of the creative methods became part of what Baker (2005) calls an organic process, that is, part of an innovative methodological process developed with children’s needs in mind to elicit their stories. Creative methods can thus be understood as a way to enable children to tell their narratives in their own way, as children (like adults) express their opinions and experiences in different ways; and at the same time it gives the child the possibility to assess the situation and have control over how and what to express (cf. Jensen 2012). Involving children in research is thus a way of both gaining new knowledge through their perspectives and narratives and safeguarding the children’s right to share their perspectives on difficulties concerning them and be listened to (UNCRC 1989).

Another aspect that Baker (2005) mentions is the question of time. We find both time and place to be of relevance, and therefore we have tried to use an organic approach in this respect as well. The time available is often limited, and as researchers we have to keep this in mind and be flexible. The children might go to school or have leisure activities, or the shelter where they are living might only be able to assist at certain times during the day. The location itself is also of relevance, as it is important to meet in a place where the child feels safe (cf., Jensen 2012). Therefore in our study the children are able to choose the location of the interviews (with the help of their mothers), for example at home, at the shelter, at the university, or via digital media (like Zoom). The use of digital technology and digital interviewing can be valuable when researching sensitive topics, and it has been shown that vulnerable groups have positive experiences of online settings (Thunberg and Arnell 2021). The younger population may also be more used to handling digital technology and solving problems that might arise when using digital methods.

The organic methodological process highlighted by Baker almost two decades ago is thus still important when planning and carrying out research with children. As researchers we need to be flexible and use various methods that enable children to participate and tell their stories. However, digital technologies and tools have developed considerably in the last two decades, and we can use them to facilitate children’s participation. Another aspect is that the collective experiences of conducting research about and with children can facilitate further development of this methodological field. When working with children who have experienced domestic violence, this can include the researcher’s reflexivity, which refers to the researcher taking active responsibility for the children’s well-being during the interviews, and also being flexible in relation to the needs of each individual child (Cater and Øverlien 2014). This is to be done without taking on the role of a counsellor.

Consent

An important aspect of involving children in research that Baker highlighted (2005, Øverlien, 2012) is the question of consent. In the Swedish context, children aged 15 and above can decide for themselves whether they want to participate in research, as long as they understand what it entails. Below that age, consent is required from their legal guardians, usually the parents. This means that children under the age of 15 might not be able to participate in research even if they want to and understand what it entails, if a parent (or legal guardian) does not want them to. As Baker (2005) also points out, parents thus become gatekeepers who might refuse to allow their children to participate out of concern for potential harm, or perhaps simply because they do not want their children to participate. As a result, there is a risk of children’s opinions not being heard, even though children have the right to be heard in matters that concern them (UNCRC 1989).

In our project informed consent was obtained from the mothers before any interviews were conducted with the children. The children received information adapted to their age, so that they too could understand and choose whether they wanted to participate (if their mother had consented) (cf. Cater and Øverlien 2014). Although it is only children under the age of 15 who must have the consent of a legal guardian, we sought consent from the mothers of all participating children, because the interview might trigger reactions related to trauma they may have gone through, and because of the potential security risk that participating might pose. Before the interview began the children were also asked if they wanted to participate, however, as consent from the legal guardian is needed, the children did not need to sign a consent in the same way as the mothers did. If they had drawn any pictures, we also asked if we could use them as, for example, illustrations in publications.

The question of informed consent from the legal guardian(s) can, as mentioned, be a hindrance to involving children and young people in research, and in our project, we had to adapt to each child and mother’s situation and to the possible security risk of contacting the father (often the perpetrator). The general rule in Sweden is that consent must be obtained from both (if there are two) legal guardians, but there are exceptions. It is therefore up to the Swedish Ethical Review Authority to weigh the researchers’ arguments concerning risks and benefits and decide whether to approve a different procedure. In our case, this resulted in the ethical approval being revised so that consent was required from only the mother in cases where informing the other legal guardian about the child’s participation would put the mother and child at risk. Still, it raises questions regarding what might happen if the father (often the other legal guardian) received information about the research participation afterwards. This could result in potential problems, as he was unable to use his rights as a legal guardian, nor even been informed of the child’s participation, which can affect, for example, the life situation of the child as well as trust in researchers and public authorities. However, we believe that this is a minimal problem, as in most cases the fathers have little to no contact with their children, and we have received clearance for this method from the Ethical Review Authority. But it is something to keep in mind when planning what kind of research methods are to be used and from whom consent should be collected. Not obtaining consent from both legal guardians should be used restrictively.

The two new laws – the incorporation of the UNCRC into Swedish law and the enactment of the Violation of a Child’s Integrity law (presented previously in the paper) – can be said to affect both which children are legally seen as victims of crime and the possibility of involving children in research. As a practical consequence, this has resulted in, or at least facilitated, our project receiving ethical approval to obtain consent only from the mother without having to inform the abusive partner (usually the father) or ask for his consent, even if he is a legal guardian. Legally, the rights of the child have thus been strengthened during the last two decades, with the declaration of children’s right to participation and to have their opinions heard. It is still too early, however, to say whether it favours children’s opportunities to be involved in research.

Access through proxies and gatekeepers

When the project began, we expected our method of recruitment to give us access to children, especially because so many of the domestic violence shelters we contacted were positive about the project and thought it could generate important knowledge. However, the reality was different, and the difficulty of involving children proved to be considerable. Like Baker (2005), we have only received access to a few participants, so far mostly mothers of the youngest children in our study (0–6 years). Baker (2005) highlights both unwillingness on the part of gatekeepers and tight timetables with regard to research funding as potential obstacles to interviewing children. We agree with Baker but wish to discuss a few more aspects of this in relation to our project.

Firstly, as Baker (2005) also mentions, gatekeepers (e.g. parents and professionals) may be worried about potential harm and therefore choose not to inform children about the chance to participate. Thus, not only mothers/legal guardians can be understood as gatekeepers in our study, but also shelter staff, in so far as they decide which mothers receive information about the project and which do not. Secondly, most children living in domestic violence shelters are younger children (Chanmugam 2011), and thus the possibilities for involving adolescents are small. Thirdly, the Covid pandemic reduced the number of women and children staying in shelters, which normally house around 6,000 adults and 6,200 children per year (National Board of Health and Welfare 2020) and forced many shelters to close their activity centres for children.

Involving children in research thus requires going via a gatekeeper, and consequently adults make decisions before the invitation to participate reaches the child. On the one hand, these adults interpose themselves out of concern for the children’s well-being. They wish to make sure the research does not harm the children in any way, such as by bringing up painful memories or enabling perpetrators to find out where they are hiding. On the other hand, children have a right to have their opinions heard and respected (UNCRC 1989). The most important aspect to remember is that the children need to decide for themselves whether they want to participate. However, if they do not receive the invitation in the first place, there is a risk of their opinions about domestic violence shelters being missed or filtered through adults’ perceptions of their experiences.

It is also important to highlight that children with experiences of violence and of staying in domestic violence shelters in Sweden express that they want to tell their own stories and be listened to by adults (BRIS 2020). This means that we need to find new ways to make sure that children are given opportunities to tell their stories, while at the same time they are protected from harm. Research institutions, welfare organisations, policymakers, domestic violence shelters and organisations for children who have experienced domestic violence all need to be involved in this and discuss what issues should be in focus to strike a balance between participation and protection (cf. Cater and Øverlien 2014). This could result in ethical guidelines that, for example, would enable researchers to be more flexible in relation to the individual children who want to be involved (cf. Coyne 2010).

Confidentiality/disclosure/anonymity

Researching violence, and including children who have fled violence and sought shelter elsewhere, requires us as researchers to take responsibility for the situation and ensure that those who participate (in this case mothers and children) are not at risk of having their identities, or any other information that could reveal who or where they are, disclosed. Therefore, confidentiality is of the utmost importance. In our study, the participants are, or have been, living with protected identities, and they have resided in domestic violence shelters and/or other places that are kept secret from the perpetrator (often the father). Maintaining the anonymity of the children by changing names, places, and other recognisable characteristics or situations is thus vital, and as Baker (2005) points out, it is important to explain this to the children so that they know their information is protected.

During the last two decades, changes have also been made in relevant rules and regulations. One example is the General Data Protection Regulation, which regulates how personal data can be used within the European Union. This new regulation may not have a major impact on children’s participation, but it does affect how we collect and store data containing personal information when using digital tools/digital technology. Another example is the replacement on 1 January 2019 of the regional ethical review boards by The Swedish Ethics Review Authority. The purpose of the new authority is to achieve increased uniformity and efficiency in decision-making, and, as before, to ensure that research is conducted ethically. This means that research may only be approved if it can be carried out with respect for human dignity and human rights. These two examples do not change the involvement of children per se, but they do affect researchers and our research projects, because rules, regulations and controls regarding confidentiality and anonymity have been made stricter to ensure that research participants are not harmed or put in unnecessary danger.

One point highlighted by Baker (2005) is the question of confidentiality and whether it is possible to achieve complete anonymity. The Ethics Review Authority in Sweden notes that the use of terms like anonymity is problematic, and in our research we have tried to clarify that we cannot guarantee anonymity because the children’s mothers, shelter staff, and anyone the children themselves might tell will know about their participation. However, the universities in Sweden have updated their systems, for example regarding secure servers and computers, and have modified the way we report our projects, travels, and costs within a project to make ethical difficulties and participants’ safety high priorities. Although we can never guarantee anonymity, we can promise to do everything in our power to make it more difficult to identify participants.

Dissemination

There are several ways of disseminating research, and Baker (2005, pp. 286–287) mentions the importance of including the perspectives of participating children. This may seem obvious, but it nevertheless bears repeating. It can also be argued that not using narratives or other material collected with children is ethically problematic (Morrow and Richards 1996), but this does not mean that children can participate in research under all circumstances (Cater and Øverlien 2014). As our research is still ongoing, and the children involved are few in number, we can unfortunately only discuss our plans for dissemination. Baker (2005) notes another vital but even more difficult aspect of projects involving children, namely to report back to the children so that they know what has happened to their narratives/participation. The circumstances surrounding children who have been exposed to violence and have experiences of living in domestic violence shelters make this difficult. In our project we have asked the participating children and their mothers if we may contact them again later, and most have agreed. However, it is not really that simple; the children and their mothers move often, sometimes to a new home, sometimes to a new shelter, and in some cases back to the home they fled from. Also, contacting these families can be risky. It may reveal their whereabouts and, in the worst case, lead to further violence. We have therefore primarily chosen to keep in touch with as many accommodation facilities as possible, to update them on how things are going within the project and ask them to forward the information to the families that are interested.

We have also chosen to speak about the project and its result both publicly (e.g. by radio and newspaper) and with wider groups of professionals (e.g. social workers, children’s therapists, lawyers, police, etc.). Today, most funders also require that the knowledge produced must reach the target groups it affects and society at large, which in this case includes public authorities and social organisations that work in corresponding fields. Involving children thus implies a responsibility to disseminate the knowledge based on their stories – to let their opinions be heard – but also to facilitate the wider reach of their stories, or in other words, to make sure that the knowledge produced is used to improve the conditions of children in similar situations.

Limitations

The involvement of children in research on violence is still rare, and children’s perspectives are often conveyed through adults, as has also been the case in our study so far. This may be due to the sensitive topic of the research, but such difficulties extend to other, less sensitive, subjects of research as well. Baker (2005) also discusses the difficulty of reaching ethnic minority communities. She relates this to specific difficulties that ethnic minorities can face when seeking help. We also would like to add the possibility of language being a barrier to holding interviews, and the ethics of using an interpreter when researching sensitive subjects and housing/shelters. To clarify, each additional person (e.g. an interpreter) who is included becomes a potential safety risk, both in relation to trust/rapport and the increased risk of information being spread. Lastly, we wish to return to the question of time in relation to external funding, as highlighted by Baker (2005). The limited research time that is currently available to us affects how our projects are structured and the time available for data collection, more specifically the opportunity to be present on site for a longer period in order to build trust/rapport with the children as well as with mothers and other gatekeepers whose participation we would like to have.

Conclusion

In relation to the UNCRC (1989), there needs to be a balance between the child’s protection and participation. Adults, usually the parents, are responsible for making sure that the child is protected from potential harm. Still, children also have a right to participate in matters that concern them and to express their opinions about their situation. The question, then, is how can these rights be combined, when involving children in research on experiences of violence and staying at a domestic violence shelter?

Based on studies of sensitive research with adults and young people, it does not appear that such research harms participants, at least not in the long term (e.g. Kavanaugh and Ayres 1998, Decker et al. 2011, Alexander et al. 2018, Thunberg 2022). Instead, the results seem to lean more in the other direction; people express that participating in research helped them with their recovery process, as it gave them a way of talking about their experiences. This would suggest that, on a general level, fewer problems are associated with participating in research on sensitive topics than one might expect. It can still be difficult for participants to talk about their experiences but doing so does not seem to have any long-term negative effects. However, the question still remains: How can we involve children and young people in research projects in such a way that both the parents and children can express their opinions? As Morrow and Richards (1996) state: ‘[…] to avoid asking the questions because they are ethically difficult, thereby excluding children from research, is an ethical position in itself ’ (p. 103). Children and young people need to be able to participate in research, and researchers need to be able to involve them. Without their opinions, knowledge that is important for research and society will be missed.

Acknowledgments

We are grateful to all the children and mothers for sharing their experiences with us and thank the children for sharing and letting us use their paintings.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The present study was funded by Stiftelsen Allmänna Barnhuset (Dnr. 2020-285). The funder was not involved in the analysis, or the conclusions drawn in the study, the text written is that of the authors.