Copyright Sociological Research Online, 2000

Jeni Harden, Sue Scott, Kathryn Backett-Milburn and Stevi Jackson (2000) 'Can't Talk, Won't Talk?: Methodological Issues in Researching Children'
Sociological Research Online, vol. 5, no. 2, <http://www.socresonline.org.uk/5/2/harden.html>

Abstract

In this paper we explore some current issues in, what has come to be called, the new sociology of childhood and how these relate to the process of researching children's lives in general, and to our own research in particular. We discuss the developmental model of childhood, before going on to explore ideas about children as, on the one hand, inhabiting a relatively autonomous realm and, on the other as part of the same social world as adults but with different sets of competencies. The implications of these differing positions for researching children will be assessed prior to a discussion of the design of our current research, on children and risk, and the wider implications of our reflections on the research process.

Keywords:

Childhood; Children; Methodology; Research Methods

Introduction

1.1

The assumptions and concerns we bring to the research process shape the techniques or methods we employ. Within the sociological study of children, the key issue upon which methodological debates turn concerns the nature of childhood. The way in which researchers conceive childhood will shape the research in which they engage. Indeed, the extent to which researchers embrace or reject the idea of children as 'different' shapes the nature of their research. Childhood is primarily a relational term, grounded in its relationship of difference with adulthood (Jenks 1982).

1.2

For at least the last hundred years children have increasingly become a project not just for their parents (Beck and Beck-Gersheim1995) but also for the state and its various emissaries. As Nikolas Rose puts it:

The modern child has become the focus of innumerable projects that purport to safeguard it from physical, sexual and moral danger, to ensure its 'normal development', to actively promote certain capacities or attributes such as intelligence, educability and emotional stability (Rose, 1989: 22)

1.3

Sociological research is, of course, not external to these processes and children have increasingly become objects of the sociological gaze. New sociological approaches to childhood suggest that rather than viewing children as future adults in the making we should focus upon children's lives and perceptions and activities. This entails a shift away from the idea of a child as 'becoming' an adult to the 'being child', conceptualised as an active social agent (Qvortrup 1994). While these new perspectives are essential in challenging adult-centred views, we should not forget that children's lives are largely bounded by adult surveillance. There is no free and autonomous realm of childhood outside the social relations in which childhood in general, and particular individual childhoods are forged (Scott et al 1998).

1.4

In this paper we explore some current issues concerning the process of researching children's lives in general, and for our own research in particular. We begin with a discussion of approaches to researching children framed by differing conceptualisations of childhood. We will discuss the developmental model of childhood, before going on to explore ideas about children as, on the one hand, inhabiting a relatively autonomous realm and, on the other, as part of the same social world as adults but with different sets of competencies. In the second section we will discuss in detail the design of our research, on children and risk. Finally, we will address the wider implications of our reflections on the research process. This research from which the discussion in this paper is drawn entailed interviewing children, aged between 9 and 15 and their parents from 30 Households (15 in an urban area and 15 in rural and semi-rural areas) in The South East of Scotland. The project was designed to assess the salience of the concept of 'risk anxiety' in these households and to investigate the ways in which matters relating to risk, danger and safety were negotiated on an everyday basis (for more detailed discussion see Harden et al 2000 and Harden 2000) ^[1].

Understanding Childhood: Approaches to Researching Children

2.1

In this section we will discuss differing conceptualisations of childhood and their implications for research strategies.

Adults in the Making?

2.2

Developmental studies of childhood stress the difference between children and adults. Children are, at different ages, in different stages of development towards adulthood - the older they become, the more adult they become. One indicator of this is in terms of language. Moreover, the assumption that such differences are inherent has encouraged the development and use of standardised tests to assess child development. Indeed, the whole educational system is premised on the institutionalisation of age grading, the result being that children are held to be incapable of doing what they are, in fact, not yet permitted to do (Thorne 1987).

2.3

James et al. (1998) argue that while there is much agreement that such an approach is problematic, the influence of the developmental model in our everyday thinking is so pervasive that it can and does influence even those who take a paradigmatically different view. The clearest example of this is the researcher's choice of age group to study, which, they point out, is often not explained or reflexively addressed. While not denying that there are differences between children at different ages, they are critical of the pre-sociological approach for assuming that such differences are natural (James et al. 1998: 174). They argue that we should conceptualise age as a social rather than natural variable in research (James et al. 1998:175). Solberg (1996) argues that in order to genuinely explore how childhood is constructed it is necessary to avoid letting such assumptions about age influence our approach. She advocates 'a certain ignorance of age' in favour of an emphasis on 'the situational contexts within which children act and suggests that we move our attention away from 'being' to 'doing' (Solberg 1996: 54).

2.4

We would suggest that it is also useful to think reflexively about the everyday effect of the developmental model. Childhood may be socially constructed, but it is constructed through culturally and historically specific beliefs and assumptions, at the core of which in late modern Western societies is the developmental paradigm. These beliefs and assumptions do not simply operate at the level of ideas but have material and practical effects on social organisation and interaction. Thus it is important to analyse these processes as well as deconstructing them. For example, many contexts (schools, play schemes, family activities etc.) are organised in age specific ways and this in turn is likely to produce behaviour which is then defined as age specific, at least within a given context. The interesting questions, sociologically, relate to how age appropriate behaviour is learnt in different contexts and how it varies across settings. Our research specifically set out to explore what children and parents saw as age-appropriate behaviour and activities in relation to risk and safety. We noted that the institutionalisation of childhood in relation to the transition from primary to secondary school has now come to be seen as one of the important boundaries between childhood and adolescence.^[2]

A Different Species?

2.5

Other approaches conceptualise childhood as 'different', not in the sense of being a stage towards adulthood, but as another culture, a world relatively autonomous from the adult world. As Waksler (1991) notes,

In everyday life we adults take for granted that children as a category know less than adults, have less experience, are less serious and are less important than adults in the ongoing work of everyday life. I suggest that for the word less we as sociologists try substituting the word different and consider the theoretical and methodological implications.

2.6

James et al. (1998) have described this as the 'tribal child' approach. Such an approach, based on the idea of children's social competence, has been important in raising awareness of the importance of studying children in their own right. Nevertheless, James et al. (1998) argue that there has been a tendency to focus on difference, for example by concentrating on play as a part of childhood but not of adult culture (James et al. 1998: 181). This conception of childhood tends to reinforce the view of children as special and of childhood as a special time, thus reinforcing the distinction between children and adults. Moreover, the emphasis on being child-centred has implications for the research sites within which children are studied. This has produced a view which suggests that the exploration of children's worlds is best achieved through the study of children in child-centred settings, for example the playground (Opie and Opie 1969). Thus the family tends to be excluded from studys of children's worlds (James et al. 1998:182). More generally this approach ignores the extent to which children live their lives under adult surveillance. Hence the child-centred approach minimises the consequences of adult control and structuring of children's lives and the extent to which the spaces for autonomous play are often actually bounded by adult control (James 1993).

2.7

From this perspective it appears as though the only appropriate method, for studying children and childhood, is participant observation because it allows a supposedly complete and untarnished picture of children's culture to emerge. Clearly, participant observation is problematic in a culture where children are used to seeing adults as different and are therefore unlikely to be prepared to accept them as one of themselves or to ignore their presence. Even observation is constrained by practical issues of access. For example, it would be very difficult if not impossible to undertake sustained observation in a household context in Britain, given the privacy associated with family life. This does not necessarily mean that other methods, in particular interviews, should be seen as 'second best'. Interviews can offer unique insights into the experiences of respondents, allowing them to describe and explain their own social worlds, and it would be wrong to assume that through observation the researcher can gain the 'real' picture while in interviews only a public account is given. Rather than judging methods as tools through which we can uncover the 'truth', we should instead examine the relationship between the methods used, the assumptions behind these and the data collected (Bourdieu 1977).

Equal but Still Different?

2.8

James et al. (1998) argue that children should be understood, not as lesser beings than adults, but as having different competencies which the researcher must address. The recognition that children can actively participate and communicate their ideas in research challenges the belief that children are somehow less competent that adults. Indeed there is also an underlying presumption that adulthood itself confers competency as a research participant. Nevertheless different methods have been developed for research with children such as, encouraging them to express themselves through drawing and writing in school, based on the assumption that they find this an easier way to communicate than one-to-one verbal communication. As James et al. argue,

Engaging children in what might be called 'task-centred activities' which exploit children's particular talents and interests might provide a better way of allowing children to express their ideas and opinions than the use of more 'talk-centred' methods such as interviews or questionnaires (James et al. 1998: 190).

2.9

However, this presupposes that sociologists have the skills to analyse responses given, for example, through drawing. Indeed there is a danger of reading in meanings which may not be there for the children themselves (Backett-Milburn and McKie, 1999). Moreover, there is a tendency to assume that these techniques will revel some truth not accessible through talk.

2.10

Such use of 'task-centred' activities, now common-place in individual or group interviews with children, has been influenced by techniques developed in anthropology such as Participatory Rural Appraisal (Thomas-Slater 1995), and by practitioners working with troubled children (Hill et al. 1996). Children are being asked to express themselves through, for example; sentence completion exercises (Morrow 1998); taking photos and drawing pictures (Punch 1998). Often a combination of these methods is employed. On the one hand, such methods are used to encourage children to express their views and opinions on the topics raised by the research, while, on the other hand, it is expected that these techniques will foster a rapport between the child and the researcher. However, as Hill et al. (1996) point out, there needs to be a balance between the use of such techniques and straightforward conversation, the key to which is a flexibility and sensitivity on the part of the researcher to the individual children being interviewed. These sorts of methods are, in our view more valuable and less problematic if they are utilised as an occasion or stimulus for talk rather than as evidence in their own right.

2.11

There are several issues concerning the question of children's competencies and the direction of task-centred methods towards these. First, it is possible to question the idea that children have different competencies, or that the 'child' competencies such as drawing are more effective ways to communicate with children. Backett and Alexander (1991) found that children's limitations in their drawing ability shaped what they 'said' in their pictorial representations. Furthermore while for adults, children's involvement in drawing may be associated with 'fun', we cannot assume that this is the case for children. Though children are encouraged to express themselves in this way in school, not all children are comfortable with it. It is important that children do not become entirely 'other' in the research context – because we were all once children does not make us experts on all children. Nevertheless, the fact that some adults have memories of finding 'art' a confusing medium makes it likely that some children will feel similarly now.

2.12

Second, we can question the underlying assumptions of the researcher which may lie behind the use of 'task-based' methods with children. The perceived need to create special techniques in order to communicate with children is premised on the belief that this is more difficult than is the case with adults. However, as Alderson and Goodey point out,

It is only more complicated to speak to children if one assumes in them a certain degree of taxonomic remoteness from ourselves. However, for complication in the subjects, read fear in the researchers… My complicated and complication-inducing nervousness exists in the face of what I feel to be alien and threatening, and is increased by the extent to which I cannot consider children to be on an equal footing with myself (Alderson and Goodey 1996).

2.13

Third, the wider research context must be acknowledged. There is increasing emphasis in the development of innovative methodologies in order to secure funding for research. The ESRC programme 'Children 5-16: Growing into the 21st Century', stipulated this as an aim of the programme. While, this is not necessarily problematic, it creates an environment in which research with children is being directly associated with a hierarchy of inventiveness not seen to be necessary when researching adults, thus constructing children as 'other' in methodological terms. This can lead us back to notions which sociologists ought to be questioning such as developmental and tribal models of the child and the idea of children as inherently different or special.

Interviewing as 'Equals'?

2.14

Where interviewing is used with children it tends to be seen as a specialised activity different from other forms of interviewing. Children tend to be as socially distant from the researcher whatever their gender, ethnicity or socio-economic background. However, it is important to remember that adults were all children once. The assumptions that a researcher brings to any project are rooted in our biographies and personal experience as well as being mediated through sociological knowledge and training. Thorne argues that this poses problems for the researcher,

To learn from children adults have to challenge the deep assumption that they already know what children are 'like' both because, as former children, adults have been there and because as adults they regard children as less complete versions of themselves. When adults seek to learn about and from children, the challenge is to take the closely familiar and to render it strange. (Thorne 1993: 12)

2.15

For Mandell (1991), our epistemological assumptions about childhood - whether we view it as different or the same as adulthood - determine the role that we, as researchers take in ethnographic work with children. She advocates taking the 'least adult role', which entails suspending 'all adult-like characteristics except physical size' (Mandell 1991). However James et al (1998) question whether it is desirable to take on such a role (James et al. 1998:183). They argue that researchers do not need to pretend to be children to understand them or to argue from their perspective. Rather, the researcher could aim to be friends with the children they are studying. Nevertheless James et al. argue that this may prove difficult given the highly structured and primarily unequal nature of adult-child relations, as a result of which, children may feel uneasy with any attempt by adults to befriend them (James et al. 1998:189).

2.16

We feel, however, that the assumption that it is possible to be friends with your research subjects is in itself flawed. For example, there have been debates, in the context of feminist research, about the possibility of a special and relatively equal relationship between female researchers and female respondents (Oakley 1982). This position has been criticised on methodological (Malseed1987) and ethical grounds (Finch 1985). Friendship is essentially a relationship between equals (Jamieson 1998) and the structure of research generally militates against complete equality.

2.17

It has been noted that one of the features of child research is that inequalities of power between children and adults are duplicated in the research process. One of the implications of this is that children may say what they think the researcher wants to hear, particularly in one-to-one interviews (Hill et al. 1996; Mahon et al. 1996). One proposed solution to this problem is that the interviewer and interviewee be 'matched', by encouraging children to be interviewers, in order to minimise power differentials and facilitate a closer relationship (Alderson 1995; Mahon et al. 1996).

2.18

The idea of minimising the social distance between researcher and participant in order to create rapport has also been discussed in debates about feminist methodology (Oakley 1982; Finch 1984). However, Puwar (1997) points out that 'matching' interviewee and interviewer in terms of what the researcher perceives to be the key criteria of identification (such as gender, race, age sexuality) does not necessarily facilitate a closer relationship in the interview. The building of rapport can depend on less obvious identity markers, for example the places where one has lived (Puwar 1997). We cannot assume that age is the key difference between an adult researcher and child respondent, or the only one which matters. There may be commonalities which can override this difference or other differences which may be, in this context, more significant than age. Furthermore, 'matching' cannot change the fact that one person is the interviewer and one is the interviewee, a form of inequality inherent in the interview process (Collins 1998).

2.19

Moreover, involving children in interviewing may create as many problems as it solves, not because the interviewers are children, but because of their relationship to the researchers who are making use of their services. In the first place, employing interviewers from outside the core research team – whether they are children or adults – is problematic in terms of the process and practice of qualitative interviewing. If interviewers are not party to the thinking behind the research design and the main research questions being addressed, it is very difficult for them to engage in the requisite 'thinking on the feet' entailed in probing more deeply where necessary, following up on issues hinted at but not fully explicated, or referring back to significant points made earlier in the interview. Second, integrating children into the research as interviewers may simply create a new level of hierarchy in which the child interviewer is positioned between adult researchers and the children researched. The child researcher thus becomes a conduit through which adult researchers (who remain at the top of the hierarchy) gain access to children's views of the world.

2.20

Third, as Mayall (1994) argues, it is at the analysis stage in the research process that the power differentials between children and adults are most clearly manifested.

However much one may involve children in considering data, the presentation of it is likely to require analysis and interpretations, at least for some purposes, which do demand different knowledge than that generally available to children, in order to explicate children's social status and structural position (Mayall 1994: 11).

2.21

Once again this is not a problem which is specific to research on children. In researching adults attempts to involve subjects in the process are inevitably limited at the analysis stage. If academic research is to produce anything more than lay understandings it must involve access to concepts, theories and scholarly knowledge unavailable to most research subjects. Indeed such prior conceptual understanding is likely to have influenced the aspects of lay knowledge on which the researcher chose to focus prior to and during the fieldwork. Everyday experience cannot deliver a full understanding of the conditions which produce that lived experience (Smith 1988)

2.22

The issue of power relations between children and the researcher has also been seen as an ethical consideration. In the child research literature there seems to be a greater concern over ethical issues than in other areas. While the importance of ethics in the research process is not denied, the immediate association between children and ethics must be questioned. Alderson and Goodey (1996) argue that the idea of a 'child-centred' ethics is grounded in spurious assumptions. If we reject the Piagetian view of children as not yet competent or fully functioning moral beings, then we should be aware of an equally questionable implication, that children constitute a separate species for ethical purposes. It is a short step from saying that children think or act in a characteristically non-adult way, to 'treating' them as stereotypes of childhood with one ethics for adults and another for children.

2.23

Morrow and Richards (1996) argue that the ethical debates around children centre on the extent to which children are perceived as being vulnerable on the one hand, and as incompetent on the other. In Western cultures children are conceived of as a 'protected species' (Scott et al. 1998) and, as such, it is not surprising that in research the same belief is held. In addition, the prevalence of developmental discourse,which presents children as less than adults in all respects and ensures that children are often seen as less competent than adults (Hutchby and Moran-Ellis 1998). Both these issues have ethical implications in terms of gaining access to children and in the actual process of data generation. A key issue here is the tendency for children to be seen as innocent and in need of protection from researchers as well as other predatory adults (Scott and Watson Brown 1997). It is, however, crucial for social researchers to be reflexive about these issues and not simply to take them for granted.

2.24

While ensuring children's rights in the research process is clearly a delicate matter (Alderson 1995) there is no simple formula to persuade us that research with children always carries a greater ethical burden than any other. The main ethical issues should not revolve around children's innate difference but relate to children's social location as subordinate to adults. For example, 'informed consent' is problematic not primarily because of children's lack of understanding of research, but because their participation in any research project is dependent on adult gatekeepers. In our study of children, risk and safety, gaining access was initially dependent on schools and later on parents and it was clear that the children themselves had rarely been consulted about their involvement. It is important therefore that we see negotiations with gatekeepers, for example Local Education Authorities, schools, teachers and indeed parents as aspects of the research process and therefore as data on researching children.

Reflections on Talking to Children about Risk and Safety

3.1

In our initial discussions about this project, prior to submitting the proposal to the ESRC, we were very clear that, while two of the three grant holders were able and wished to be involved in fieldwork they would only be interviewing parents. This decision was taken despite the fact that all three applicants had a wealth of qualitative research experience, including with teenagers and on sensitive topics. Stevi and Sue both felt, at this stage, that they did not have the necessary skills and attributes to successfully interview children. In the job description we made it clear that we wished to appoint a research fellow who did indeed have the requisite skills and attributes. This indicates the extent to which we had unreflexively accepted the idea that researching children is a special activity requiring special skills and self-presentation. We went on to assumed that Stevi and Sue did not possess such skills, or at least not in sufficient abundance to be confident interviewers of young children. This assumption was based on the following reasoning: as women in their 40's who did not have children themselves and who had never interviewed anyone under 14 Sue and Stevi felt that they lacked the necessary knowledge and language to communicate effectively with children, and that they might appear too much like teachers, or parents, or even grandparents! Reflecting back on these assumptions, we now feel that, while we would not wish to suggest that interviewing children is easy or that any researcher could successfully undertake it we do think researchers should be reflexive about both their own and wider social assumptions in this context. For this reason we feel it is appropriate to present some reflections on these issues by Jeni Harden, the research fellow, and it is to these that we now turn.

Preparing for the Field

3.2

For the remainder of the paper we move between Jeni's first hand experience, in relation to interviewing children, and the more general concerns shared by the whole team. Jeni's first person account of her fieldwork is rendered in italics and we return to roman text to make the wider connections.

3.3

In the course of this project I have interviewed children and young people from eight to sixteen years, the vast majority of whom have been articulate and enjoyable to talk with. I would have felt fortunate to have such a group in any 'adult'- based project. Yet at the start of the project I experienced greater concerns about the process of data collection, and self- doubt about my own research skills, than I have ever felt before. These concerns were rooted in my own lack of experience with children which led me to make assumptions about difficulties in communicating with them. On the one hand, these worries were related to the feeling that I should be able to talk to another person despite an age difference. On the other hand I also had professional concerns. Most of the literature, rightly, refers to children as competent social actors, and discussed the techniques being used successfully in researching with children. I had concerns that I would not be able to match this success.

3.4

Other than the children of one close friend, I had little contact with children in my daily life. Yet when this project began I started to see and hear children everywhere. I became much more aware and, I suppose, interested in their presence in my adult world. The idea of trying to understand children's perspectives became very appealing - in part I saw them as a group discriminated against by the adult world, and by giving them the time and space to talk, I could help to address this problem. I also saw them as an 'unknown' that I wanted to find out more about.

3.5

But how to go about it? Interviews, as far as I was concerned were about talk. How did you get children to talk? Though my friend's children chatted quite freely with me, they had known me all their lives. To the children I would interview, I would be a complete stranger. Though I have interviewed adults in several different contexts, before beginning this project I had never interviewed children. I felt confident in my skills as an interviewer and had experienced many difficult situations. For example, my work with parents of chronically ill children involved a great deal of patience and sensitivity as the process of talking about the illness was often very upsetting. Interviewing doctors in Russia posed the challenge of being seen by the interviewees as an outsider and often treated with suspicion.

3.6

Thorne (1993) warns of a form of ethnocentrism in research with children - we think we are closer to children than we really are because we too were once children. However, Jeni described her concerns in a different light.

3.7

In some ways I felt closer to a Russian adult than to Scottish children. Perhaps this stemmed from limited experiences with children, as an adult, or from limited memories of my own childhood. While this may have been useful in avoiding some assumptions about the experience of being a child, and the implications this has for research, it brought with it other difficulties.

3.8

What if they don't say anything? What if they just walk off when I'm talking? What if they just want to play all time and won't answer any questions? What if they can't understand what I ask them or I can't understand their replies? These were just some of my 'what if' nightmare scenarios. Having had limited experience in any context with children, I was clearly under the impression that children would find the interview situation far harder to deal with than adults. Though I was well aware that not all adults were comfortable in such situations, I assumed that most if not all children would find it difficult to cope with and so would make my job a much harder one. Therefore in this respect I began from the position that children were different from adults.

3.9

It was with these concerns on my mind, that I sat waiting in the hairdressers one day. Two sisters, of around 8 and 10 years were waiting for their mother. I glanced over at them and they caught my eye and started laughing. I smiled at them and one came over and sat next to me and asked me who my favourite Spice Girl was. I was honest and told her that I didn't like any of them, which she didn't seem to mind. They launched into a debate as to the various merits of Sporty and Scary, and I asked them questions about each one. Later I emerged from the hairdressers, not only pleased with my new haircut but also triumphant in the knowledge that I could do it - children would talk to me. First Contact had been made. To a certain extent I am trivialising what were real concerns at the time and which shaped the methods we chose to use in our interviews with children.

Reflections on 'Special' Techniques

3.10

The use of 'task-centred' methods in research with children was discussed above. While we were not convinced that children, as a group, had different competencies from adults, there were several reasons why we included some written and some verbal exercises in our interviews. First, we wanted to explore these techniques with children and to assess their applicability for use with a wider population. Second, we felt that it was important to try to create an atmosphere in which our participants could feel comfortable and importantly, not get bored. Given reports that children find one-to-one interviews, with adults, intimidating (Hood et al. 1996), we felt that structured activities may reduce any pressure on them to talk or to maintain eye contact with the researcher. In this way they acted as a 'time out' device to enable the participants to relax. Third, we wanted to use the exercises as a springboard for discussion both in the abstract, about children in general, and about their own personal experiences.

3.11

We gave the children several different tasks: ?

• Compiling lists of things that they thought were risky and not risky for children and young people.
• Sentence completion exercises. 'I feel scared/safe/guilty/annoyed at my parents when…'.
• Cards with commonly referred to risky situations, for example 'talking to a man you don't know', and to group them into categories depending of whether they thought they were not risky at all risky for both boys and girls, more risky for girls, or more risky for boys.
• Vignettes: four short written scenarios were given to the children to read and then comment on.

3.12

The artificiality and strangeness of any interview process is likely to be heightened by adult-child power relations and by the fact that children are less likely than adults to have been exposed to analogous experiences. Children's accounts of themselves, particularly as presented to adult strangers and 'experts' is frequently mediated through known adults. For example, parents usually accompany children on visits to the doctor's surgery and explain their symptoms for them. In a variety of everyday contexts adults speak for children. Thus children are not used to giving an account of themselves to unknown adults. In using task-centred research techniques in order to facilitate rapport and put children at ease, we were compensating for the limitations of children's social experience not for any innate incapacity. In what follows Jeni reflects on the differences she perceived between children and adults in the interview process.

3.13

There were differences between the interviews with children and with adults in the nature of the conversational process. The adults I interviewed tended to answer questions at length, or were more likely than the children to have their own agenda for what they wanted to discuss. Interviewing the adults proceeded with fewer needs for question probes and longer monologues from the interviewee. This is not to say however that interviewing adults was easier, simply that it posed different challenges. The longer monologues given by adults often contained many different points of interest and it proved difficult at times to both remember and to be able to bring the interview back to cover all points raised. The rapid dialogue that tended to characterise the interviews with children, made it easier to pick up on and probe further, almost everything the children said. Both child and adult interviewing present different challenges to 'thinking on your feet', one of the key skills involved in qualitative interviewing (Mason 1996).

3.14

The different conversational styles characteristic of the interviews I undertook with children and parents have implications for the techniques used in the interview process. I found it useful with the children to use task-based exercises as prompts for discussion. The readiness of adults to talk at length and to give examples to illustrate their remarks seemed to reduce the need to use such tasks as vignettes to encourage discussion, though they were still useful in focusing adults on a particular issue and in eliciting responses to a standard scenario. So we may use the same tools with children and adults but they can serve different purposes.

Reflections on an Old Technique: Rapport Building in Relations between the Interviewer and Interviewee

4.1

Miller and Glassner (1997) argue that rapport building is essential to minimising the impact that social distance may have on the interviewee. This involves making the interviewee feel comfortable and competent to talk freely in the interview (Blumer 1969). Here we return to Jeni's reflections on the fieldwork.

4.2

I tried to create a rapport with the hope of letting both the child and myself feel relaxed and able to enjoy the interview. When we sat down together, I usually chatted with them about their room, what they had been doing, what I had been doing and so on, before going on to talk about the project with them and what they could expect during the interview.

4.3

I think it is especially important to try to establish a connection with the interviewee. There are some children with whom I found it easier to 'connect' and whom I felt relaxed with very easily. There were others whom I felt I was unable to form any kind of link during the interview. Creating a rapport can mean sharing experiences with them before and during the interview. For example I play computer games and was able to talk to some of the children about these - and get some handy tips! One boy said to his mum at the end of the interview that he was surprised because I liked all the same things as him. Another girl had horses so I was able to tell her about my own experience with horses. The key to establishing a connection was to pick up on the things they were interested in and to give information about yourself. The conversations I had with children on topics such as horses or computer games may not have been entirely relevant to our research but I feel that they were worthwhile. They helped the children to relax and foster a dialogue because they eased the pressure on them to talk about themselves all the time.

4.4

Of course children are not a homogenous category. As with adults, there are key factors which can influence children's experiences, such as gender, class, age ethnicity and so on. While what we are primarily concerned with here is they ways in which children as a group are defined as different from adults in the context of research it is nevertheless useful to note Jeni's experiences of the differences between children.

4.5

I felt most rapport with the teenage girls. In this case age and gender were both significant factors, given that I remember that I remember that part of my life (with the help of diaries kept at that time) far better than my earlier childhood. But it is by no means as simple as that. I did not necessarily feel more at ease with older children or with girls. Nor did I find it' easier' to interview older children than younger ones. I used the same tasks with them all, regardless of age or gender. Moreover their readiness to discuss the most sensitive topics, for example sexual risk, did not seem to depend on age or gender – a ten year old girl talked freely about the risk of AIDS whereas an older girl was embarrassed in such a discussion. Perhaps one difference was in the language used by the children. For example, younger children tended to speak of 'Stranger Danger' in terms of 'bad things' happening to children, whereas older children would refer to rape or sexual assault.

4.6

We would not want to over emphasise the problem of developing rapport with children, because in many respects all successful interviewing involves framing topics in ways which encourage people to talk about themselves, and using interpersonal skills to put respondents at ease. One reason why it was relatively easy to develop rapport with the parents was because, in the main, they already viewed their children as part of their own life project (Beck-Gernsheim 1996) and were thus keen to talk about them and about the process of parenting. Nonetheless rapport was not always automatic and the three of us involved in interviewing parents (Sue, Stevi and Jeni) felt that there were some we connected with better than others. In our ongoing analysis we have discussed with each other interviews where we felt both very comfortable and very uncomfortable.

4.7

The appearance of the researcher, and also their presentation of self, will play a part in the process of creating a rapport (Measor 1985; Scott and Porter 1983). Those of us interviewing parents reflected on appropriate modes of dress and presentation of self, wanting to establish our professionalism as researchers but also to take account of the domestic settings in which the research was conducted. We were interviewing parents at home in the evenings and at weekends and they were frequently dressed in tracksuits or jeans and slippers and in one case even in a dressing gown. We would therefore have felt out of place wearing business suits. However Jeni felt that there were particular issues of self-presentation involved in interviewing children while at the same time establishing credibility with their parents. This is how she has reflected on this issue.

4.8

Some years ago, as a graduate researcher, interviewing dentists, I became very aware of both age differences and differences in professional standing when I was mistaken by the dentist for his replacement receptionist on a Youth Training Scheme. During our current study, I was also very aware of the different selves that can be presented by the interviewer, because I was in contact with both parents and children. I was often asked by the parents what my role in the research is - often they presumed I am a student. With the parents therefore, I was conscious of trying to be 'adult'. As clothing is a significant factor in the initial presentation of self, when interviewing adults I consciously wear smarter 'work' clothes. For interviews with children I did not feel this would be necessary and I was able to wear jeans, trainers - often clothes quite similar to those which the children themselves wore. This was not an attempt to be 'least adult' by copying them, rather it was that in interviews with adults I was trying to be 'most adult'.

4.9

It is important to consider what creating a rapport in an interview setting is intended to achieve. Most obviously it is hoped that this will enable participants to relax and so to communicate more freely. However, inherent in this, is the idea that interviews are 'search and discovery' missions and interviewees are seen as 'vessels of answers' (Holstein and Gubrium 1997:116). Creating rapport then becomes a means by which to maximise the potential for the extraction of information from interviewees. It is hoped that a relaxed interviewee will talk more and will talk more honestly. This position rather begs the questions what constitutes good interview data?

Interviewing Children: What are Good Data?

5.1

Bourdieu (1977) argues that interviews are the weakest form of research methodology because participants give official accounts based on what they think they ought to say. It is also argued that we should expect interviews with children to be particularly problematic because the power relations between adult researchers and children are likely to exacerbate the tendency to give public rather than private accounts in the interview setting. This argument assumes that interviews can produce the 'facts' of children's – or anybody else's – lives. It is, we would suggest, more useful to consider, and, where possible, to research, the range of versions which children would give to researchers, teachers, social workers or even to their parents.

5.2

From a positivist perspective the aim of the interview is to collect objective 'facts' with as little bias arising from the interview setting and from the researcher as possible. It is argued that bias can be limited by interviewers strictly controlling their participation in the interview (Gordon 1987). Much of what is involved in creating rapport would, in this context, be viewed as increasing potential sources of bias and so distorting the data. In a similar vein, but from a different perspective, it has been argued that participant observation can provide a truer account of children's lives than interviews because the researcher can establish a relationship with the child and so gain a fuller picture of 'what they really think' (Ennew 1994).

5.3

However, most sociologists would now dispute the notion of 'facts' emerging straightforwardly in the context of an interview. It is argued that there can be no knowledge about a child's 'reality' because the interview is an interaction in which narrative accounts are created by the participants (Miller and Glassner 1997). Holstein and Gubrium (1997) argue that both interviewee and interviewer are active participants in a social process. 'Respondents are not so much repositories of knowledge, treasuries of information awaiting excavation, so to speak, as they are constructors of knowledge in collaboration with interviewers' (Holstein and Gubrium 1997: 114).

5.4

Nevertheless, while rejecting the positivist belief in interviewees as receptacles for 'facts', which the researcher should release, it is not necessary to deny any connection between the accounts given by people in interviews and their wider social worlds (Miller and Glassner 1997). Interviews can provide access to the way people make sense of their own experiences and social worlds. 'Narratives which emerge in interview contexts are situated in social worlds, they come out of worlds that exist outside of the interview itself. We argue not only for the existence of these worlds, but also for our ability as researchers to capture elements of these worlds in our scholarship' (Miller and Glassner 1997: 105).

5.5

When interviews are understood in this way, we can begin to question the attempts to encourage children to talk more openly in interviews. Children, like other research participants give accounts of their experiences and beliefs to the interviewer even though they may tell different stories in different ways to their friends, or to their parents or to another interviewer. Nevertheless these stories are based in their 'real' worlds, and on their knowledge and experience regardless of whether they are accurate representations of any particular event (Thomson and Scott 1991; Holland et al 1992).

5.6

It is possible from this perspective to challenge the view that children will not talk. Mahon et al. (1996) noted that in their project, 'less 'successful' interviews, in terms of depth of response and rapport established were conducted with those in younger age groups and with boys'. This assumes that data = talk and the more talk the better the data (Brownlie 1999). It would be easy to conclude from this that children may not provide such rich data as adults because they may talk less in interviews. However, whether answers given by interviewees are yes/no or are lengthy monologues, they are still part of the interview interaction and are equally useful and important forms of data. Holstein and Gubrium (1997) note that,

The narratives that are produced may be as truncated as forced- choice survey answers or as elaborate as oral life histories, but they are all constructed in situ, as a product of the talk between interview participants (Holstein and Gubrium 1997: 113).

5.7

In a previous study Jeni interviewed a woman who saw the interview as a chance to give, in detail, eighteen years of health care experiences. This interview produced three hours of talk. On the other hand one boy interviewed for this project was very quiet and said very little, the interview lasted only thirty minutes. Does the interview with the woman necessarily provide 'better' data? As in life, it is often the case that those who talk the most in interviews are not always the most interesting. Moreover, silences are also data. In our interviews with children, they rarely say much about sex and sexual risk whether spontaneously or when asked directly. Can infer from this that children are embarrassed to talk about sex or that sex does not map onto their own personal experiences, it is not within their 'system of relevances' (Giddens 1991)? Or is it rather that they do not have easy access to a public language through which to raise and explore these issues (Holland et al 1998).

Conclusion

6.1

Interviewing a child may, in some ways, be a different experience from interviewing an adult, but we must be reflexive about our understanding of such differences and about the choices and decisions taken in the process of researching children. There are many factors other than age which are central to research strategies and outcomes, including other forms of social differentiation, the situational contexts of interviews, and the subject area being explored. Moreover, while not denying that age is and should be a consideration when planning to interview children, we have argued in this paper that it is children's social location rather than anything inherent in being a child, which merits our attention. It is from this perspective that attempts to develop new and reconsider old methods of research, in relation to researching children, should be understood.

Notes

¹The study referred to is 'The Impact of Risk and Parental Risk Anxiety on the Everyday Worlds of Children', funded as part of the ESRC 'Children 5-16: Growing into the 21st Century'. The views expressed in the paper are, however, those of the authors and do not necessarily reflect those of the funders or employing organisations.

²However we also noted that this 'boundary' differs in age terms between England and Scotland, with the transition occurring one year later in Scotland.

References

ALDERSON, P (1995) Listening to Children: Children, Ethics and Social Research. Ilford: Barnardo's.

ALDERSON, P., AND C. GOODEY (1996) 'Research With Disabled Children: How Useful Is Child-Centred Ethics?' Children And Society. 10 (2), pp 106-116.

BACKETT, K. AND ALEXANDER, H. (1991) 'Talking to Young Children About Health: Methods and Findings', Health Education Journal 50(1) pp 34-38.

BECK, U and Beck-Gersheim, E. 1995 The Normal Chaos of Love, Cambridge: Polity.

BROWNLIE, J (1999) Between Discourse and Practice: Creating the therapeutic subjectivity of the young sexual abuser Unpublished PhD Thesis, Department of Applied Social Science, University of Stirling, UK.

BOURDIEU, P. (1977) Towards a Theory of Practice. Cambridge: Cambridge University Press.

BLUMER, H. (1969) Symbolic Interactionism: Perspective and Method. New Jersey: Prentice Hall.

COLLINS, P. (1998) 'Negotiating Selves: Reflections on 'Unstructured' Interviewing' Sociological Research Online, 3( 3).

ENNEW, J. (1994) Street and Working Children: A Guide to Planning. Development Manual 4. London: Save the Children.

FINCH J. 1985 'It's great to have someone to talk to' In Bell, C. and Roberts, H. eds Social Researching: Problems, Politics and Practice, London: RKP.

GIDDENS A. (1991) Modernity and Self Identity. Self and Society in the Late Modern Age. Cambridge: Polity.

GORDON, R.L. (1987) Interviewing: Strategy, Techniques and Tactics. Homewood: Dorsey.

HARDEN, J. Backett-Milburn, K. Scott, S. and Jackson, S. (2000) 'Scary Faces, Scary Places: Children's constructions of risk' The Health Education Journal.61 (1) pp34-51.

HARDEN, J. (2000) 'There's No Place Like Home….' Childhood 7(1) pp43-59.

HILL, M., LAYBOURN, A. AND BORLAND, M. (1996) 'Engaging With Primary-Aged Children About Their Emotions And Well-Being: Methodological Considerations', Children And Society 10 (1) pp 129- 144.

HOLLAND, J. RAMAZONAGLU, C. SCOTT, S. SHARPE, S. AND THOMSON, R. (1994) 'Methodological Issues in Researching Young Women's Sexuality' in M. Boulton (ed) Challenge and Innovation: Methodological Advances in Social Research on HIV/AIDS, Basingstoke: Falmer.

HOLSTEIN, J.A. AND GUBRIUM, J. F. (1997) 'Active Interviewing' in D. Silverman (editor) Qualitative Research: Theory Method and Practice. London: Sage.

HOOD, S., KELLEY, P., MAYALL, B., OAKLEY, A. (1996). Children, Parents And Risk. London: Social Science Research Unit, Institute Of Education.

HUTCHBY, I AND J. MORAN-ELLIS (1998) 'Situating children's social competence' in I. Hutchby and J.Moran-Ellis (eds) Children and Social Competence: Arenas of Action. Basingstoke: Falmer.

JAMES, A. (1993) Childhood Identities. Edinburgh: Edinburgh University Press.

JAMES, A. JENKS, C. PROUT, A. (1998) Theorizing Childhood. Cambridge: Polity.

JAMIESON, L. (1998) Intimacy: Personal Relationships in Modern Societies. Cambridge: Polity.

JENKS, C. (1982) 'Constituting the Child' in C. Jenks (ed) The Sociology of Childhood: Essential Readings. London: Batsford.

MAHON, A., C. GLENDINNING, K. CLARKE, G.CRAIG (1996) 'Researching With Children: Methods And Ethics', Children And Society 10 (2) pp145-154.

MANDELL, N. (1991) 'The Least Adult Role in Studying Children', in F.C.Waksler (editor) Studying the Social Worlds of Children. Basingstoke: Falmer.

MASON, J. (1996) Qualitative Researching. London: Sage.

MAYALL, B. (1994) 'Introduction' in B. Mayall (editor) Children's Childhoods Observed and Experienced. Basingstoke: Falmer.

MILLER, J. and GLASSNER, B. (1997) 'The 'Inside' and the 'Outside': Finding Realities in Interviews' in D. Silverman (editor) Qualitative Research: Theory Method and Practice. London: Sage.

MORROW, V. (1998) Understanding Families: Children's Perspectives. London: National Children's Bureau.

MORROW, V. AND M. RICHARDS (1996) 'The Ethics Of Social Research With Children: An Overview', Children And Society 10 (1)pp 90-105.

OAKLEY, A. 1982 'Interviewing Women: a contradiction in terms' in Roberts, H. ed Doing Feminist Research London: RKP.

OPIE, I AND OPIE, P. (1969) Children's Games in Street and Playground. Oxford: Oxford University Press.

PUNCH, S. (1998) Negotiating Independence: Children and Young People Growing Up in Rural Bolivia. Unpublished PhD Thesis, School of Geography, University of Leeds, UK.

PUWAR, N. (1997) 'Reflections on Interviewing Women MPs' Sociological Research Online 2 (1).

QVORTRUP, J. (1994) 'Childhood Matters: An Introduction' in J. Qvortrup, M. Bardy, G. Sgritta and H. Wintersberger (eds) Childhood Matters: Social Theory Practice and Politics. Aldershot: Avebury.

ROSE, N.1989 Governing the Soul: The Shaping of the Private Self , London: Routledge.

SCOTT, S and PORTER, M 1983 'On the bottom rung: Women and women's work in Sociology' Women's Studies International Forum 6(2) pp 181-206.

SCOTT, S. AND WATSON-BROWN, L. (1997) 'The Beast the Family and the Innocent Children: Feminism and Fairytales' Trouble and Strife ,No 36.pp 7-13.

SCOTT, S. JACKSON, S. BACKETT-MILBURN, K. (1998) Swings and Roundabouts: Risk Anxiety and the Everyday Worlds of Children. Sociology 32(4) pp 689-707.

MALSEED, J (1987) 'Straw Men: A note on Anne Oakley's treatment of textbook prescriptions of interviewing' Sociology 21(4) 629-31.

SMITH, D. E. (1988) The Everyday World as Problematic, Milton Keynes: Open University Press.

SOLBERG, A. (1996) 'The Challenge in Child Research: From 'Being' to 'Doing'' in J. Brannen and M. O'Brien (eds) Children in Families: Research and Policy. Basingstoke: Falmer.

THOMAS-SLATER, B. (1995) 'A Brief History of Participatory Methodologies' in R. Slocum, L. Wichhart, D. Rocheleau, and B. Thomas-Slayter (editors) Power, Process and Participation- Tools for Change. London: Intermediate Technology Publications.

THOMSON, R. AND SCOTT, S. (1991) Learning about Sex: Young Women and the Social Construction of Sexual Identity, London: Tufnell Press.

THORNE, B. (1987) 'Re-visioning Women and Social Change: Where are the Children?' Gender and Society 1(1): 85-109. THORNE, B (1993) Gender Play: Girls and Boys in School, Buckingham: Open University Press.

WAKSLER, F.C. (1991) Studying the Social Worlds of Children. Basingsstoke: Falmer.

Copyright Sociological Research Online, 2000

36