‘If you look,you have to leave’: Young children regulating research interviews about experiences of domestic violence

Children in refuges in Norway

Refuges for abused women have existed in Norway since 1979. Since then, they have been a haven where mothers seek safety and can isolate their children from abusive fathers and partners. Norway has a population of close to 4.7 million. In 2008, 1895 women stayed at women’s refuges (Sentio Research Norge, 2011), and during the same year, 1725 children stayed in refuges with their mothers. We can see then that refuges have almost as many children as women under their roofs. The length of stay varies greatly. Some of the children stayed one night, while others stayed several months (Øverlien et al., 2009).

Children who experience domestic violence

Children who experience domestic violence have increasingly been recognized as a vulnerable group. Several researchers have shown a connection between domestic violence and various forms of maltreatment (Edleson, 1999; McGuigan and Pratt, 2001). Children who experience domestic violence are at increased risk of other forms of violence, such as physical child abuse (Herrenkohl et al., 2008; Rumm et al., 2000). Moore and Pepler (1998) found that 42 per cent of the children in their study who had experienced domestic violence had also been physically abused. Children who experience domestic violence are also at risk of developing a number of emotional and behavioural problems, both short and long term (Holt et al., 2008; Øverlien, 2010).

Particular concerns for young children

Limited research has been conducted in regard to domestic violence and younger children. However, the results of this research give us reason to be particularly concerned. Findings suggest that younger children are especially exposed and vulnerable (Cunningham and Baker, 2004). Researchers have found that children as young as 1 year old are negatively affected by their fathers’ violence. Bogat et al. (2006) found trauma symptoms among 1-year-old infants who had been exposed to severe domestic violence (i.e. heard or seen the violence). DeJonghe et al. (2005) found that 1-year-old infants exposed to domestic violence were more likely to display distress in response to verbal conflict than children who had not been exposed. They also had additional risk factors, being more dependent upon their caretakers and having fewer means to express their concerns than older children and youths. Because younger children are more often physically close to their mothers, they are more likely than older children to experience the violence visually (Fantuzzo and Mohr, 1999; Fusco and Fantuzzo, 2009).

Young children as informants

As informants in qualitative research, children, and young children in particular, have been under-represented. A recent review of the literature found only 18 phenomenologically oriented qualitative interview studies with children published between 1991 and 2008, and only three of them included children as young as 4 years old (Spratling et al., 2012). Much of the qualitative research on domestic violence relies on mothers’ reporting. Mothers tend to both under-report and over-report what their children may have seen, heard or in what way they may been affected by in terms of violence (Apple and Holden, 1998). One example is Litrownik et al.’s (2003) study of 692 children and their mothers. They found that 14 per cent of the mothers reported that their children had witnessed a family member being hit, while one in three of the children reported the same.

Still, as a number of researchers have observed, the last three decades have seen growing recognition of the value of listening to children’s voices, children’s trustworthiness as informants and their rights to be included as participants in research (Dockett and Perry, 2007; Lahman, 2008). This growing recognition is part of a general increase in focus on children as social actors and individuals with rights (Jans, 2004), in which the United Nations Convention on the Rights of the Child (UNCRC) has been of central importance. According to the UNCRC, children have the right to be heard in issues that concern them (article 12). With the introduction of the new sociology of childhood/social studies of childhood in the late 1980s came an understanding that children need to be taken seriously as social agents and as active constructors of their own social worlds. As such, they also need to be included as research participants (Corsaro, 2005; Hutchby and Moran-Ellis, 1998; James and Prout, 1990). Our research follows this approach as we not only include young children in our study but also analyse their discourse in order to gain knowledge about them as research informants.

Furthermore, there is a growing body of literature concerning how research with child informants is best conducted, in terms of both adapted methodology (Einarsdottir, 2007; Fargas-Malet et al., 2010; Moore et al., 2008; Wohlwend, 2009), and how to meet the extra ethical challenges it raises (Dockett et al., 2009; Lahman, 2008). Eriksson and Näsman’s (2012) interview study with children who experience violence combines these issues when they explore ways to negotiate the tensions between victimized children’s vulnerability and their agency and rights to participation. There seems to be considerable consensus that children can articulate their experiences in research interviews in ways that make them competent informants in qualitative research (Dockett and Perry, 2007; Spratling et al., 2012).

Using children as the source of knowledge about domestic violence has opened up quite new perspectives in our way of understanding the phenomenon and its implications. An example is how children have been found to be active participants during the violent episode rather than passive witnesses to the events (Edleson, 1999; Fusco and Fantuzzo, 2009; Källström Cater, 2004; McGee, 2000; Mullender et al., 2002; Øverlien and Hydén, 2009). Studies such as these underline the importance of including children as informants, not only as a rights issue but also to provide valid and nuanced understandings that can inform the development of interventions and treatment. However, the aim of this article is not just to show how young children can provide important information about their life situation but also to propose how young children in fact may co-regulate, limit and even take the lead in the research interview. To do so, we utilize a concept originally developed to describe infant–mother interaction.

Infant intersubjectivity and self-regulation

The concept of infant intersubjectivity has been outlined by the British psychologist Colwyn Trevarthen and several other authors over a series of publications ranging from the early 1970s to today (Aitken and Trevarthen, 1997; Bråten, 1998; Murray and Trevarthen, 1985; Stern, 1971; Trevarthen, 1974). Some of these are independent but related scientific works, rather than a unified theory, but they still make some common theoretical claims about preverbal infants: they have a rudimentary sense of self, a capability for intersubjective interaction and an ability to regulate their own involvement in such interaction (Trevarthen, 1998).

At the basis of the theory lie observations of non-verbal interactional patterns between 2 -month-old infants and their mothers. These interactions were first highlighted and termed proto-conversations by Bateson (1971) and are characterized by turn taking, reciprocality, initiative and response, moments of joint attention and affective attunement to the other. Equally importantly in these dialogical patterns, infants also periodically turn their gaze away, and this is considered necessary for the infant to integrate and compute impulses, and to prevent becoming overwhelmed and overstimulated (Stern, 1985; Trevarthen, 1998). Infant intersubjectivity theories thus understand the infant’s involvement in proto-conversation as closely connected to their self-regulation of internal emotional states (Schore, 1994). They further claim that the pattern of dialogue described in the infant–mother interaction remains the fundamental building block of more complex communication later in child and adult development (Trevarthen, 1998). In fact, the theory has inspired several normative models for communicating with older children in clinical ¹ and pedagogical settings (Aarts, 2000; Øvreeide and Hafstad, 2007).

If infants can co-regulate their interactions with others and display behaviour that serves to protect themselves against over-stimulation, could these abilities be transferred into the verbal interactions that young children engage in? In this study, we ask more specifically, how do young children regulate their involvement in research interviews about sensitive and potentially traumatizing aspects of their lives?

Background

This article draws on a larger study of children in refuges for abused women in Norway. The study by Øverlien et al. (2009) was the first nationwide research study conducted in Norway on children in refuges. The aim of the study was to collect systematic knowledge about the children and adolescents in the refuges. A further aim was to obtain a better understanding of their experience, both before and during their stay at the refuge, as well as their thoughts and wishes for the future. When planning this study, a decision was made to include young children (4–7 years of age), despite the extra challenges this entailed in terms of support from ethical boards, adjustment of the interview situation and analysis of data. In this article, we present and discuss the findings from this subsample.

Participants

Voices from six girls and one boy, aged 4–7 years, were analysed in this study, and four of them are presented in this article. ² They all lived, or had recently lived, at four different refuges in Norway. All the children in the study had experienced violence towards their mother from a partner, sometimes the child’s father and sometimes not. Some of the children had experienced direct physical violence. Although we know that children who experience domestic violence are at risk of sexual abuse, none of the children talked about having been subjected to sexual violence. The selection of children was made on the basis of age and psychosocial situation, and they were recruited with the co-operation of the refuge staff.

Interviews

A decision was made to conduct tape-recorded interviews with the youngest informants in the study, although researchers are increasingly using video cameras as a tool in participant observation to study young children’s lives (Corsaro, 2005). The interviews took place in a quiet, separate room at the refuge, such as the kitchen. The interviews were between 20 and 90 minutes long. An interview guide was developed (adjusted for the age group 4–7 years) and used primarily when preparing for the interview rather than as a check-list with questions that had to be answered. The aim was an interview in which the child was in focus, and in which the interviewer had the position of listener rather than interrogator. Most often, the children were interviewed alone, but on one occasion, at the request of the mother, the child’s mother was present. Interviews were tape-recorded and transcribed by the interviewers. The interviews were conducted by three researchers, two women and one man, all trained and experienced in different types of professional conversations with children. The excerpts chosen were translated into English by the researchers and proof-read by a professional translator. The transcriptions of data can never be a mirror image of the interview situation itself. On the same lines, translated excerpts are not replicas of the original wording used by the informant. However, great efforts have been made to keep the English translation as close as possible to the original language.

Ethical concerns

The study followed the ethical principles recommended for the social sciences in Norway and has been subjected to standard ethical investigations by the Norwegian Social Science Data Service. During the fieldwork at the refuges, the staff, alone or together with the researchers, informed the mother and the child about the project, its purpose and its research team. The children were informed of their right to terminate their participation in the study at any time. This information was provided verbally and in writing and was translated into the mother and the child’s native language when needed. Interviewers were especially observant for non-verbal signs of distress and were prepared to terminate interviews should the children show such signs. Furthermore, the mothers were informed of the obligation of the researchers to inform the Child Protection Services if a child’s well-being was at stake. The mothers and those children who were able to write were asked for written consent. The children were asked for verbal consent, and the information letter was read to them.

Analysis

The interviews were analysed by applying elements and concepts from infant intersubjectivity theory. Three elements characterize this analysis and distinguish it from the phenomenological analysis often used in interview research. First, the interviews were treated as whole entities of meaning-making processes as opposed to breaking them down into partial elements considered to represent the child’s experiences. Second, drawing on sociolinguistic methodology (Mishler, 1986), the interactions between child and interviewer were viewed as social interactions, rather than merely linguistic exchanges. Finally, the interactions between the children and the interviewer were compared with the infant intersubjectivity model of proto-conversation.

Using these analytic tools, each interview was listened to, and transcripts were re-read repeatedly, together with the field notes taken by the interviewers. Patterns of verbal and non-verbal interaction were identified and categorized in terms of the children’s involvement in the interview themes.

Negotiating involvement – ‘can we be calm now?’

The following excerpt is from the interview with Oda, who stayed in the refuge with her mother for over 3 weeks. The interviewers were told by the refuge staff that she was 5 years old, but during the interview, she made it clear that she was 4. She is in fact the youngest child in our sample:

We see here how Oda is able to participate in a dialogue with the interviewer and at the same time negotiate other activities, including her own breaks. She digresses from the subject several times, seemingly dodging the interviewer’s questions, but in fact, the interactional pattern is one of turn taking and dialogue. In a sense, Oda is leading the conversation, while the interviewer is supporting her with open-ended questions, emotionally attuned utterances and summing-up of the content of Oda’s last contribution. This seems to make it easier for Oda to talk about her experiences.

Thus, Oda is an able informant about her own experiences, and at the same time, she shows skills in regulating her own engagement and involvement in the interview. A closer examination of this interaction, using proto-conversation between mothers and their infants as a model, may help us to understand better what happens in these regulatory processes.

The interviewer introduces a theme from Oda’s own story (Yellow house), as a mother would engage an infant by addressing something in his immediate surroundings. She awaits the response from Oda (yes) and follows it up with an open invitation to expand (Tell me). Oda begins her story, and a series of turn-taking responses between Oda and the interviewer follows, as in proto-dialogue between mother and infant. The interviewer acknowledges Oda’s statements with emotional attunement (Oh no, and In the stroller. So, you were sad). This allows Oda to expand her story even more, this time with descriptions of her mother’s emotional reactions (She is scared. She cried like me). At this point, Oda diverges and turns her interest to the tape recorder (Can I hear it talk?), like an infant pointing towards something to engage his mother in joint attention. The interviewer follows Oda’s focus, makes a comment about it and returns to the issue she is interested in (We will talk some more. What was it like before you came here?). This makes Oda follow her lead again, and this time with an emotionally laden description of her father (He was so very, very mean). Following that is the request for a break, or, at least, the interviewer’s interpretation of such a request (Can we be calm now?). In light of our model, this can be understood as a verbal equivalent of an infant withdrawing her gaze. Oda’s request is respected by the interviewer.

Applying the theory of infant intersubjectivity to our data allows us to see interactional patterns similar to the rhythmic dialogue described between infants and their mothers. We can see in Oda’s behaviour the equivalents of an infant pointing to invite joint attention and intermittently withdrawing her gaze, thus regulating her involvement and her own emotional state. The role of the interviewer also becomes more evident. A sensitive, emotionally attuned interviewer follows the focus of the child and respects the child’s need to defocus and diverge, and thus enters easily into a rhythmic dialogue with the child, resulting in a relatively ‘rich’ interview.

Control by redefining –‘now I hear a police car’

Other children act differently, but with similar results, in regulating their involvement in the interview situation. In the following transcript, the interviewer is trying to engage Dina, a girl aged 6 years, in a conversation about the people she has met since she moved into the refuge with her mother and little sister 3 days earlier:

To the growing frustration of the interviewer, Dina is not sticking to the issue at hand. Rather, she seems to be avoiding the questions and invitations of the interviewer and introducing her own topics instead, such as chipmunks from a children’s movie, and friends from the past. Several different interview strategies introduced by the interviewer are ignored, and Dina effectively avoids the issue of how she has experienced meeting other children and their mothers in the refuge. Her experiences on this topic remain obscure during the whole interview.

On other issues, Dina does indeed stay on track and replies readily to the interviewer’s questions. She does, for example, give quite a lot of detail about how they got from their home to the refuge. There does not seem to be anything lacking in her ability to participate in a dialogue about events in her life, or even about sensitive issues. Rather, Dina exemplifies a way of avoiding some of the issues addressed by the interviewer, and she does it in a seemingly effortless way. She does not show particular signs of distress or discomfort, and her tone of voice is light and calm. She does not seem burdened by the fact that she breaks the communicative rules set by the interviewer.

By conceptualizing what happens in the interaction between Dina and the interviewer in terms of intersubjectivity, this appears to be more than just avoidance. Dina can be seen as regulating her own involvement in the interview, by moving ‘in and out’ of the theme at hand. She takes control over a situation in which she initially had little control. Her communicative divertive behaviour can be considered a communicative skill, much like infants avoiding eye contact with their mothers for a while during non-verbal interaction. In any case, the effect is that she takes control over the conversation and redefines what is supposed to be the issue.

Avoiding follow-up – ‘I don’t know’

Other children regulate their involvement in yet other ways. Simone, for example, is 7 and moved to a refuge 1 month before she was interviewed. She now lives in transitional housing owned by the refuge. Simone’s mother does not want the interviewer to use the word ‘refuge’ during the interview, so they agree on using the phrase ‘this house’. Simone is described by the interviewer as uncertain and confused during the interview, and the interview with her is characterized by a different kind of avoiding behaviour:

Simone seems quite ambivalent in this excerpt, and she regulates her involvement in the conversation by silences and ‘I don’t know’ answers, as well as introducing topics of her own. Her ‘I don’t know’ answers can be understood as a form of communicative skill, as they provide the responses the dialogue requires, but without revealing herself. At the same time, she seems to regulate her own feelings. Her tone of voice is low and sad, but at the same time, she describes her experiences in positive words (happy). She seems to want to turn the subject towards a more positive direction. She does not answer any questions about her father. Although mother and child have been informed of the purpose of the interview and consented to participating, Simone’s signals of avoidance could be interpreted as if she has changed her mind and no longer wishes to participate. In fact, her signals of avoidance are so clear that the interviewer terminates this interview in advance, without Simone having to ask for this herself.

The effect is that Simone gains some control over the interview. Her strategy is effective in the sense that few of the issues introduced by the interviewer are followed to any degree of depth. The interviewer interprets Simone’s diversions as signals for her to stop her line of questioning, either trying another approach, as shown in the example, avoiding the issue altogether, and finally finishing the interview in advance. Again, we can see the similarity with an infant avoiding eye contact as a sign that she needs a break from stimulation.

Using play to gain control – ‘if you look you have to leave’

The last voice to be presented here is the voice of Olivia. Olivia is 5 years old and had lived in the refuge for nearly 1 month when she was interviewed. She is described by the interviewer as active, self-assertive and seeking a lot of contact, especially before the interview and during breaks. Olivia shows us an example of just how creative and playful children can be in their effort to gain control and to regulate their involvement. In this excerpt, she has introduced a game in which the interviewer must keep her eyes closed while she is drawing:

The punishment for opening her eyes, Olivia explains, will be the termination of the interview. Play is central in many of the interviews and in different ways. All interviews contain playful elements. The children are encouraged to play during breaks, and experiences with play are also a central theme in the interviews. In this transcript, Olivia shows us how play can also be an effective way of gaining control over, and regulating, the interview setting. The interviewer is forced, at least for a while, to abandon her agenda of gathering research data, and to take part in Olivia’s game instead. Furthermore, the game that Olivia invents not only gives her control over her own involvement but also includes taking control over the interviewer’s behaviour. The observation that play can make a child ‘seem as he were a head taller than himself’ (Vygotsky, 1978: 102) seems very appropriate.