Arriving at the shelter – mothers’ narratives of their children’s experiences

Each year a large number of children are exposed to violence in their homes, for example, by witnessing one parent using violence against the other, or experiencing violence directed at them personally. As a result, together with a parent, often the mother, the children might need to flee from the violence to a domestic violence shelter. The present study is set in a Swedish context and aims to analyse mothers’ narratives of the initial time at a domestic violence shelter for mothers and their young children (aged 0–6), with a focus on children’s right to protection and participation. The study consists of interviews with thirteen mothers, which are analysed using thematic analysis. The findings show that the children rarely know why they are at the shelter, and that it is an unfamiliar place for them, as they have little knowledge of what a shelter is. The children also seem to be afraid of losing their mothers, which is expressed in how they monitor their mothers. Based on the findings, the children need to build trustful relationships, a process that takes time but can already begin on the first day at the shelter. The first day signifies the start of something different, which means that the mothers and children need time to settle into their new lives.


Introduction
Many children are exposed to violence within the family, whether perpetrated by one parent against the other parent or a sibling, directed at the child personally, or a combination of both (Gilbert et al. 2009). 1 Among children arriving at domestic violence shelters with their mothers, research shows that 80-90% have witnessed violence, and about 50% have been subjected to violence themselves (Fernández-González et al. 2018).In Sweden, 10-15% of children have witnessed physical violence between their parents, and 30% have witnessed verbal conflicts (Annerbäck et al. 2010;Cater et al. 2015;Jernbro and Janson 2016).In addition, 95% of the children who have experienced violence against a parent were at home on at least one occasion when the mother was abused (Almqvist and Broberg 2004).Being exposed to violence is a serious risk factor for children developing physical, mental and social problems (Annerbäck et al. 2012;Cater et al. 2015;Gilbert et al. 2009;Holt, Buckley, and Whelan 2008;McTavish et al. 2016).This includes the development of symptoms of post-traumatic stress, anxiety and non-suicidal self-harm, symptoms that can persist into adulthood (Cater et al. 2015;Cater, Andershed, and Andershed 2014).
The ability to protect and support children is among the factors that women value most highly about shelters (Jonker et al. 2014).The child-centred support offered directly to the children, the provision of health care, and legal support are all important (Sullivan and Virden 2017; see also Mullender et al. 1998).The process of escaping violence and the arrival at a safe place in the form of a shelter is seldom discussed in research; however, we do know that children with experience of living at domestic violence shelters often describe the escape from the home and the move to the shelter as a sudden event, and something that was not discussed with them in advance, resulting in uncertainty about what was happening (Øverlien, Jacobsen, and Evang 2009).
In addition, problematic behaviour and regression have been shown to increase among children when they first arrive at the shelter, only to decrease later, which according to Fredland et al. (2014) might be related to the shelter being a safe place where the children can externalize their emotions related to the violence (see also Copping 1996;Troensegaard, 2014).Furthermore, although most mother-child relationships are functional upon arrival at the shelter, every fourth relationship is problematic (Fernández-González et al. 2018).Children's behaviour and health have also been shown to reflect positive changes in the mother's health once the mother has acknowledged the violence and sought help (Fredland et al. 2014).This could suggest that shelter staff need to support the mother-child dyad to help it develop.Similarly, during the first time at the shelter, children also often seek explanations of the violence and of why they must move to a shelter (Vass and Haj-Yahia 2020) and might try to attribute the blame for it to someone.In most cases, the children blame the situation on their fathers, but they may also blame their mother, both parents (Vass and Haj-Yahia 2020) or themselves (Stephens, McDonald, and Jouriles 2000).
In Sweden there are approximately 280 domestic violence shelters, a figure that has increased in recent years (National Board of Health and Welfare 2020).These homes can be run by non-profit organizations (54%), private companies (37%), or municipalities (9%), which means that the condition and organization of shelters can vary greatly (National Board of Health and Welfare 2020).In addition, the sheltered accommodation varies from shared accommodation with other people in similar situations to living alone in a secure apartment.There are no exact figures on how many children spend time living in shelters each year, but the Swedish National Board of Health and Welfare (2020) estimates that approximately 6,500 adults and 6,200 children spend at least one night at a shelter each year, which is a 38% increase compared with previous estimates in 2012.
The situation of children in domestic violence shelters has received increased attention in Sweden in recent years (e.g.Bris 2020; Eriksson 2010; National Board of Health and Welfare 2020).For example, the National Board of Health andWelfare (2013, 2020) shows that children receive more information from staff about life at the shelter, and that children are given the possibility to receive support; however, the support given still varies between shelters.Nevertheless, research on children's situation in domestic violence shelters in Sweden is scanty, and we know little about whether findings from other contexts and countries, such as Canada, Norway and the USA, are transferable to the Swedish context (e.g.Bennett, Dawe, and Power 1999;Chanmugam and Hall 2012;Øverlien 2011b).
Against this background, the present study aims to analyse mothers' narratives of the initial time at a domestic violence shelter for mothers and their young children (aged 0-6), with a focus on children's right to protection and participation.More specifically, the research question is as follows: What is important for mothers and children when they arrive at the shelter and during their first day there?

Children's right to protection and participation
The United Nations Convention on the Rights of the Child (UNCRC) was ratified in 1989, with Sweden signing the document in 1990.In 2020, the UNCRC was incorporated into Swedish law after years of discussion and criticism about the advantages and disadvantages of doing so, especially due to potential conflicts with parental rights (see Quennerstedt 2009).The UNCRC stresses the importance of protecting children from harm while at the same time making sure they can participate in decisions that involve them, both of which should be done with the best interest of the child in mind (Article 3).The right to participate by expressing opinions, and to have those opinions be respected, can be found in Article 12, while the right to protection is found, for example, in Article 19 in relation to violence.Article 19 says that all signatory states are responsible for ensuring that children are protected from physical and psychological violence, for example.In the same article, the right to support is also stressed.Article 6 states that every child has a right to life, and it is each signatory state's responsibility to secure children's survival and development.This means that Article 6 also places emphasis on the protection of children.
Balancing protection and participation can be difficult, as every child is unique.It is therefore important to make sure that children have an opportunity to exercise their agency and to be heard on issues that affect their lives (see Sundhall 2012;Thunberg 2020Thunberg , 2022)).Involving children by asking for their viewpoints can help them feel included and that their opinions are valued (Thunberg 2022).Still, all involvement must consider the best interests of the child, and the situation must be adapted in relation to the individual child, for example, in terms of how much information they can and should receive about moving to a domestic violence shelter.In sum, in this study the aspects of protection and participation are of particular interest, as moving to a domestic violence shelter can affect both aspects.This is especially the case because, for various reasons, the focus is often on the mothers' needs and not the children's.This, in turn, might affect the children's perceived agency and inclusion in what happens when they arrive at a domestic violence shelter.

Method
This article is based on data from the larger project -Article 19: What sheltered housing means for abused children.The three-year project (2020-2023) focuses on understanding (1) the significance of the domestic violence shelter for children, and (2) children's life situation if they are denied a place at a domestic violence shelter for any reason.Both parts include aspects such as experiences of violence, social relationships, health, and general life situation regarding, for example, preschool and leisure activities.The data material in the project consists of interviews with children aged 7-17 and, because of ethical aspects of interviewing younger children who have fled from violence, interviews with mothers of children aged 0-6.The present article focuses on the youngest children, thereby using a subset of the interviews, specifically semi-structured interviews (Brinkman 2014) with thirteen mothers who have experience of living at a domestic violence shelter together with their child/children (aged 0-6) after leaving a violent partner (in most cases the father).In this article, we have chosen to focus on the mothers' narratives to gain knowledge about what things are of importance for mothers and children when arriving at a shelter and during their first day there.Interviewing mothers about their children yields valuable data material, but it is not the same as interviewing the children themselves about their experiences.This is a limitation of proxy interviews.However, in terms of the ethics of interviewing young children who have fled violence, proxy interviews -interviewing one person about another -are of great value in gaining a better understanding of the children's experiences.

Procedure
The interviews with mothers were conducted between February 2021 and November 2022.They were invited to participate if they had at least one child aged 0-6 at the time of their stay at the shelter.The average age of the children living with their mothers at the shelter was 4.7 years, and the sample included a total of 21 children.The larger project and the interviews were structured around Rösare's (2015) five themes or phases that women and children go through when fleeing violence and living at a domestic violence shelter.These are: (1) the arrival and first day at the shelter, (2) the first week at the shelter, (3) the continued time at the shelter, (4) the final time at the shelter, and (5) the time after leaving the shelter.The present article focuses only on the first phase, more specifically the arrival and first day at the shelter.The arrival at the shelter can be a critical time for mothers and children (Øverlien, Jacobsen, and Evang 2009;Rösare 2015), and it needs to be analysed separately from the periods that follow, because over time the mothers and children become more accustomed to life at the shelter.
The mothers were invited to talk about the phases in chronological order, for example, 'How do you think the separation and move [to the shelter] affected your child/children?How did they feel?'The interviewer could then ask follow-up questions, depending on how the interview developed and what each mother talked about and highlighted in her narrative.The mothers' narratives varied in length and, to some extent, in focus.For example, some of the interviews focused more on the experience of violence and its consequences, while others stressed the health of the child/children and the support given to them when arriving at the shelter.
The interviews lasted between approximately 40 and 90 minutes and were carried out in a place where the participants felt comfortable and safe, for example, at the shelter, in their home, or by telephone.The women were recruited with the help of staff at domestic violence shelters in Sweden, who were asked if they could give information about the study to women with children who were staying at the shelter or had stayed at the shelter within the last three years.Remembering events after three years can be difficult, but is not impossible.The time aspect could influence the narratives because of possible memory issues, but this choice was made for ethical reasons, to give the participants time to process the escape from violence, and because of difficulties in the recruitment process related to the COVID-19 pandemic regarding social distancing and shelter restrictions.

Ethics
Ethical approval was obtained from the Swedish Ethical Review Authority (Dnr.2020:04561, 2021-03242, 2021-06928-02).The women who showed an interest in participating in the study were provided with written information by the shelter staff, after which the researchers contacted the women to give them information, answer questions, and schedule an interview if the women were still interested.The written information was repeated verbally before the interview began, and the women were again given the opportunity to ask the interviewer questions to clarify any aspects of the study.Before the interviews, the women were asked for their informed consent and reminded that they could withdraw from the study at any time.Furthermore, to protect the identities of the women and their children, the names used in the article are pseudonyms.Specific details such as names of places and specific organizations have been removed.
For the sake of the mothers' and children' well-being, we have attempted to minimize any safety risks and the possibility of triggering negative feelings or trauma related to their experiences of violence.Hence, as interviewers we did our best to be sensitive to the mothers' situation and wellbeing during the interviews.The women were also informed about where they could seek help if the interview caused them any distress.Moreover, our contacts at the shelters, who helped us with the recruitment process, were also helpful in fulfilling the women's and children's needs for support, safety and well-being.

Data analysis
The interviews were constructed around the five phases of living at a domestic violence shelter presented by Rösare (2015).With this in mind, we started our thematic analysis inspired by Braun and Clarke's (2006) six phases of thematic analysis: (1) familiarizing yourself with the data, (2) generating initial codes, (3) searching for themes, (4) reviewing themes, (5) defining and naming themes, and (6) producing the report.We do not view these phases as steps that follow a strict sequence.Instead we view them as a structure, in which the analysis can go back and forth between the different phases based on what is identified during the process of coding the data material, individual reflections among the researchers, and joint discussions regarding the codes that generate the themes.This means that reflexivity and discussion were prominent throughout the analysis process (see Braun andClarke 2013, 2019;Clarke and Braun 2018).In our case, we (the two authors) first familiarized ourselves with the data material by alternating between doing the interviews and transcribing them (i.e. the person who did not conduct an interview transcribed it).We also reread all transcripts multiple times when generating the initial codes and creating themes based on these codes.This process was done individually at first, with the researchers then discussing their codes and interpretations, both as part of the analysis process and also as a way of validating the codes generated.
During the first step of reading the transcripts and the initial coding it became clear that the analysis needed to be divided due to the richness of the data material.We therefore used the phases identified by Rösare (2015) to generate a general thematization.Rösare's first phase -the arrival and first day at the shelter -was chosen as a central organizing theme (see Braun and Clarke 2013), based on codes related to the start of living at a shelter.As a result of this, we returned to the transcripts, reread them and reviewed the initial coding, generating the themes presented in the next section.For example, codes such as difficulty with separation, reactions, and symptoms of distress created the theme trauma and fear.
In the final phase of the analysis, the themes generated were analysed in relation to a children's rights perspective, with a particular focus on children's right to protection and participation, viewing children as competent actors regarding their own situation and how they define their need for support.

Findings
Based on the analysis of the mothers' narratives of their arrival and first day at the shelter, three themes emerged, a new environment, trauma and fear, and building trust.A common feature of all three is arriving in an unfamiliar place, which creates feelings of uncertainty about the shelter itself and the people there, and regarding the future.

A new environment
Analysis of the narratives shows that the shelter represented a new environment for the mothers and children.The mothers and their children did not know how long they would be staying there, whether they were safe from violence, or whether they would get along with the other women and children living there.The mothers also spoke of a sense of loneliness, about ending up in a room alone not knowing what would happen.At the same time, some mothers said that they preferred being alone, and not having to answer other people's questions or compare stories of violence.Nevertheless, several of the mothers mentioned the importance of settling in, or more specifically 'landing' at the shelter.Cecilia and Rebecka described this as follows: So when we arrived at the first shelter, we came late at night.It was way past her bedtime [. ..].Then we got to meet one of the people responsible for the children, and then she, her colleague, became [the child's] personal counsellor.(Cecilia) The first day, the staff were, they were there and we also had a few volunteers.I don't remember so much from the first day, but we settled in.That's what I remember, that we arrived in the morning, we came in the late morning.We had a little time to talk, settle in, sign some papers, the kinds of things you need to do.Cook some food, get some clothes.(Rebecka) Both quotes illustrate the importance of having time to settle into the shelter, but also to get to know the people there -both the other residents and the staff and volunteers.It is not possible to just continue life as it was, before, and the experience of arriving at the shelter could be overwhelming.The mothers and children therefore needed time together as a family unit to settle into their new environment.In addition to the more practical and social aspects of living at the shelter, Amelia described feeling a sense of freedom when she arrived: One day everything looks fine, the next day you wake up in one of the social services' domestic violence shelters.And I remember that I felt so free.It felt much better waking up in this . . .an empty apartment.(Amelia) Here, Amelia expressed that she felt free waking up at the shelter apartment.Later in the interview, however, she described that although she felt free in the empty apartment, the emptiness became a problem, especially as there were no toys for her child to play with.Furthermore, in many cases the mothers stated that their children did not know why they were there or did not understand that their mothers were unable to give them the answers they needed at that particular time.As a result, it could be a frightening time for the children, as expressed by Helena: I didn't understand the seriousness of it when we moved, that we would never come back.So, it was very hard for the children to leave everything, friends, family, their things, their beds, their safety, in a sense.Now, all of a sudden, we were among loads of kids and women whom we didn't even know, in a room and sharing a kitchen.It was . . . it was scary and new, somehow.It's not something you do every day.(Helena) Entering a new environment, not knowing why or for how long, and leaving everything familiar behind, can be frightening for anyone.Several of the mothers stated that the children did not know about the violence, suggesting that the move to a domestic violence shelter may have been even more of shock for them.It meant that their world was changing, and they had no power over the situation.This lack of power and control could make the situation at the shelter even more frightening for the children, especially if they were not given any information about the shelter environment or what was happening.Therefore, the mothers and their children need time to settle in when arriving at the shelter.
Based on the mothers' narratives, children's participation in the process of leaving a violent home and arriving at a shelter can be questioned.Several of the mothers expressed that during the first day they informed their children about why they were there, which can be interpreted as a way of involving their children in what was happening and viewing them as competent actors who could understand the situation they were in (cf.Article 12 UNCRC).However, the children had little to say about their situation, and they were seldom given information ahead of time about the escape to the shelter.This was expressed as being necessary both as a consequence of their age and for security reasons, and it might have been the only option, as described by Cecilia: As the quote indicates, mothers might not have the option to involve their children in the process of leaving even if they want to, as it could put their lives at risk.For this reason, during the first phase of living at a domestic violence shelter, it may be important from a children's rights perspective to move from protection to participation.More specifically, this would involve going from the initial stage of protecting the child (and mother) during the process of fleeing a violent home, to letting the child receive information and participate in what is happening at the shelter.This can be understood in terms of the shelter environment, as the shelter is supposed to be free of violence, with less risk of retaliation by the violent partner.However, based on the mothers' descriptions, children's participation appears to be limited.

Trauma and fear
A common theme in the mothers' narratives is the children's reactions to the violence they or their mothers had been exposed to, which became apparent to the mothers after they arrived at the shelter.For example, the mothers described the children's fear of violence and of the new and unfamiliar situation.There were also signs of regression.Rebecka described the initial time at the shelter as follows: She was very, very afraid.We had stopped using diapers, but it has affected her psychologically, everything that has happened, so she started using a pacifier and diapers again for a while.My other daughter, she became scared and very shy.(Rebecka) In the quote, Rebecka first described her 4-year-old daughter, then her second daughter who was around one year old.The quote and the other mothers' descriptions show that their children had been affected by the violence, even though some of them were young when they fled, and that this became clear to the mothers when they arrived at the shelter.However, younger children might not have been able to express this verbally, and instead showed their feelings through non-verbal communication.Similarly, during the first period at the shelter, the children seem to have been afraid of losing their mother, which manifested itself in different ways, such as the child trying to protect the mother or never wanting to leave her side.Erika and Sofia described this as follows: He protected me a lot there at the beginning, so there was a lot of me showing him I was OK, even though I was in a support session.[The staff member] went with him and opened the door to where I was sitting in counselling to constantly show him that I was still there.So, there has been a lot of showing him that he can trust the people that worked and lived there, I would say.(Erika) He exhibited both somatic and psychological symptoms.He played dead on one occasion, 'come and look mummy, there's blood'.He was afraid that someone would come and take me away, the police.He had been threatened with the idea that the police would come all the time.That is, if you don't do this, if you don't do that, then the police will come.[. . .].So he thought the police would come and take me away, and that someone would come, that I would disappear.(Sofia) Erika's statement shows that when they moved to the shelter her son needed to know where she was at all times.Otherwise he would panic, as she described it, which could be related to his experiences of violence at home.For example, she could not go to the shop next door to buy something without bringing him along.Similarly, the reason why Sofia's son pretended to be dead and said he was bleeding could be because he had witnessed some of the violence.His fear of losing his mother may be related to the new environment and the fact that the shelter is a safe place.This would mean that it is not only the violence that contributed to the appearance of these symptoms, but also the move to the shelter and the child's fear of losing the most important person in his life at that time -his mother.The behaviour expressed by the children can be described as continual monitoring of the mother, which can also be seen as non-verbal communication of their feelings.
The interviews suggest that the violence had affected the children's trust, for example their trust in unfamiliar people (or places) and their trust that their mother would return and continue to be there for them.This means that it is important that both mothers and children receive support to strengthen children's trust in their surroundings (their mother, the shelter staff, etc.).

Building trust
The narratives show that the children have often had to leave behind possessions and relationships on which they depended, and that they may need help to rebuild their sense of trust in other people.This may take some time, but it could begin right away.Erika and Helene describe this as follows: He accepted it as a home, he did.He started calling the shelter home after a while, he really did.He felt very much at home there.He actually did really well there.We still go there for support sessions, and it is really hard to get him to come back home with me.(Erika) Well, the woman who was our support person, she came to be very important for all of us.(Helena)

NORDIC SOCIAL WORK RESEARCH
In the quotes, Erika and Helena talked about the shelter and the staff, illustrating the importance of supportive relationships.Erika described how, after some time, her son came to view the shelter as a home, suggesting that he felt safe and began to trust the people working and living there.Together with Helena's description, this depicts a period of adjustment, during which the children's initial distrust transforms into trust in the shelter staff.However, it is not enough for the children to trust the people at the shelter; the mothers also need to be able to trust them.As Jacqueline described: It took some time before I felt comfortable leaving her with the staff, because they had staff who would help out by looking after the children when you had a counselling session or something like that.But it took a long time before I dared to let her out of my sight.I think we had lived there for 2-3 months before it really felt OK. (Jacqueline) Although the quote concerns Jacqueline's lack of trust -as she was not comfortable with letting her children out of her sight -it also can be interpreted as showing that the mothers' reactions and behaviours during the stay at the shelter might also affect how the children react and experience the shelter.A lack of trust by the mother can affect whether the children trust the staff, as children often look to their parents for guidance and meaning.The mothers' feelings can unconsciously be transferred to the children.For this reason, both mothers and children need to build trustful relationships with the shelter staff and receive the support they need to take the next step towards a life free from violence.

Discussion
The aim of the study was to analyse mothers' narratives of the initial time at a domestic violence shelter for mothers and their young children, with a focus on children's right to protection and participation.The findings show that the children do not always know they are going to a shelter, which makes it an unfamiliar place where unfamiliar people live or work.They often arrive without personal belongings that can provide comfort (e.g.toys or pets).This is consistent with previous research showing that the escape to a shelter was seldom discussed with children in advance (Øverlien, Jacobsen, and Evang 2009).This leads to uncertainty about where they are, what is happening, and how long they will stay (e.g.Øverlien, Jacobsen, and Evang 2009), so having time to settle in and receiving information are of great importance.Especially as they are not just escaping the violence; they are also leaving behind family, friends, and things that they hold dear and might never be able to get back.This is a loss that needs to be processed at the same time as the violence is being processed, and it can be interpreted as an additional source of stress for the mothers and their children.
Several children were described as showing signs of regression and fear of losing their mothers during the initial time at the shelter.This is also in line with previous research showing an initial increase in children's problematic behaviour (Fredland et al. 2014;Troensegaard, 2014), although it has been shown to decrease over time and when receiving support at the shelter (e.g.Fredland et al. 2014).Furthermore, building trustful relationships and making the children feel safe in their new environment are also highlighted as important by the mothers.Previous research also shows that this is one of the most important aspects that mothers look for when seeking support from a shelter (Sullivan and Virden 2017).Supportive and empowering relationships with the people living and working there can thus be understood as being of great value to the children.Research also shows that children can react negatively to violence in the home, for example by exhibiting symptoms of post-traumatic stress and anxiety (e.g.Cater et al. 2015;Cater, Andershed, and Andershed 2014;Holt, Buckley, and Whelan 2008).Both mothers and shelter staff need to keep this in mind when interpreting children's verbal and non-verbal communication.
We have also sought to discuss our findings in relation to children's right to protection and participation, as set out in the UNCRC.Our findings show that it is important to consider the age differences between children, who in this study ranged from zero to six years old.For example, our findings show the importance of giving children the opportunity to participate and receive information about what a shelter is and why they are in one.Explaining the situation to the children and allowing them to participate in the transition can make their arrival at the shelter a more welcoming experience and help them feel safer (see also Øverlien 2011a;Överlien and Hydén 2009).Exactly how this is done, however, depends on each child's age and communication skills, as well as the security measures in place, and it may not be possible to explain the situation to all children.Furthermore, as previous research has shown (Øverlien 2011a;Øverlien, Jacobsen, and Evang 2009), such information can be difficult for a child to handle, especially if it needs to be kept secret (e.g. plans to escape from home, or the address of the shelter).
In conclusion, our research shows that respecting children's rights in cases of domestic violence that require them to live at a shelter is a balancing act.Children have the right -and, we argue, should have the opportunity -to participate by expressing their opinions as much as possible.It is also important that they receive age-appropriate information to help them form their opinions.At the same time, they need to be protected from violence, and this may require withholding information from them for their own safety.Depending on the specific situation of the individual child, protection may need to take priority over participation.On the basis of our findings, we cannot definitively say what is in the best interest of a child who must flee to a domestic violence shelter, other than that it is always in their best interest to get away from the violence.Receiving information is of important, unless it poses a risk.Even when information is withheld for security reasons, it should be provided as soon as possible once the child (and mother) are no longer at risk and the child can handle the information.
From a children's rights perspective, protection (Articles 6 &19 UNCRC) and participation (Article 12 UNCRC), need to go hand in hand, so that children and their mothers are safe and feel included when arriving at a domestic violence shelter.Adults (e.g.mothers and professionals) need to trust the children's abilities and address their support needs (rather than focusing only on the mothers' needs).At the same time, there may be situations where participation would place additional stress on the children or put them (and their mothers) at risk.In such cases, participation may be judged not to be in their best interest (Article 3 UNCRC).It is therefore important to make an individual assessment of each child in relation to the specific situation, and to revise the assessment regularly as the situation changes.Whatever the circumstances, however, it is important that all children receive the help and support they need to cope with the situation and the violence they have experienced.

Limitations
The present study has some limitations that should be taken into account.First, it has a small sample, which means that generalizations from the findings should be made with caution.However, the aim of the study was to analyse mothers' narratives about the initial time at a domestic violence shelter, not to make generalizations that represent a wider population.The findings should be viewed as a way to further build on Rösare's (2015) phases in relation to mothers' and children's experiences of moving to a domestic violence shelter, and to contribute to existing knowledge.Secondly, although the study focuses on the young children's experiences, it does so from their mothers' point of view.This means that the children's experiences are interpreted through other people, and the article only captures the mothers' narratives of their young children's situation during the first phase of their stay at the shelter.

Implications for policy and practice
Based on our findings in this study, it is important that children are involved in the process of moving to a shelter.Giving the children information and involving them can help them to build up trust and a new sense of safety regarding the shelter, their families, and themselves.Of course, this needs to be done in a way that considers their level of maturity and potential risks.The insecurity of living at a shelter may never be eliminated, but it can be reduced by involving children as much as possible in the process, given the risk factors, their maturity and how involved they want to be.There also needs to be a welcoming environment and as much stability as possible, to ensure that the children can build trustful relationships with their mothers and shelter staff, as well as with other residents.Mothers may also need support to help them give their children support.For example, they might need information about common reactions in children experiencing domestic violence, and about the kinds of help that are available.Shelters need to be adapted to accommodate children of various ages, from newborns to teenagers.There also needs to be staff who can focus on the children and help them process what has happened, as well as helping the mothers so that they can support their children in turn.Small things, like asking how they are doing, listening to their verbal and nonverbal communication, and learning how they express themselves and their opinions are important steps for building trust.Finally, as Rösare (2015) points out, it is important not to forget the basics: making sure the children are fed, have the clothes they need, and are given a teddy bear or other toy to comfort them when they arrive.
She [the daughter] would have said something immediately.It would have put our lives in danger.(Cecilia)