1 Introduction

This paper explores steps taken within Young Lives to develop a qualitative child-focused and participatory research methodology to capture local understandings of child poverty, as experienced by children, their caregivers and communities in the study countries. To this aim, we provide a review of participatory child-focused research approaches and provide critical reflections from our teams’ recent experiences implementing a selection of methods to research aspects of child wellbeing. ‘Wellbeing’ and ‘transitions’ are the main themes of Young Lives qualitative research, explored through the following research questions:

  • What do children and their carers in contexts of poverty understand by child ‘ wellbeing ’ and how do their views compare and change over time?

  • What are the key moments and markers of change during childhood (i.e. childhood transitions ), and how are they experienced by children in contexts of poverty?

Placing children’s wellbeing at the core of qualitative research moves beyond traditional poverty research agendas, offering a focus for interdisciplinary enquiry centred on children’s (and caregivers’) experiences, perspectives and aspirations. Capturing children’s ‘standpoints’Footnote 1 in contexts of poverty is also respectful of their diversities and their capacities to think, feel, and aspire beyond ‘survival’ (Ben-Arieh 2005). This is not to underestimate the difficulties children face, but instead promotes a holistic view of their experiences of both wellbeing and adversity that could be used to inform more effective and integrated interventions. While childhood is marked by diversity, children’s lives and development do share some common features, notably that they share a largely marginalised structural position in relation to adults, although this generalisation breaks down at an individual level (Alanen 2001; Alderson and Goodey 1996; Boyden 2006; Boyden and Ennew 1997; James 2007; Jones and Sumner 2007; Harper et al. 2003; Mayall 1994, 2002; Punch 2002b; Woodhead and Faulkner 2008). Children are often the most affected by adverse circumstances because of their relative immaturity and their lack of social power (Boyden and Mann 2005, p. 3). They have frequently been amongst the least visible groups in social research. Child-focused research positions children at the centre around which key research questions, descriptions, interpretations and analyses are made. Involving children at different points in this process affirms children as competent social actors, the ‘experts in their own lives’, and therefore valid sources of data (Langsted 1994). This involves recognising their agency and vulnerabilities, as well as their potential for resilience in the face of adversity (Boyden and Mann 2005; Schoon 2006; Ungar 2005).

Young Lives is a major international project on child poverty (2000–2015) funded by the UK Department for International Development (DFID) to follow the lives of 12,000 children growing up in contexts of poverty in Ethiopia, Andhra Pradesh (India), Peru and Vietnam. It seeks to improve our understanding of the causes, dynamics and consequences of child poverty, and how specific policies affect children’s wellbeing. Young Lives was initiated as a ‘millennium study’ and recruited 8,000 children born at the turn of the millennium (2000/2001), along with 4,000 children who were 8-years old at the time (born 1994). Together they comprise the two study ‘cohorts’ who, along with their caregivers, are participating every few years in a data-gathering survey that collects information on diverse aspects of their lives and livelihoods. The first survey round took place in 2002 and provided essential baseline information about Young Lives children, their households and their communities. Separate survey instruments are administered to older cohort children, their caregivers and community members. The completion of the second round of data collection in 2006–2007 and subsequent rounds scheduled every few years through to 2015 will track changes in children’s circumstances and enable longitudinal analyses.

The qualitative component has only recently been introduced (2007) and was designed to be an integrated sub-study, using qualitative research methods to explore in greater depth the lives of some 200 Young Lives children across the four study countries over the remainder of the project. In each country, the qualitative research is focused in four sites, except in Ethiopia where five sites are being covered to capture ethnically-based regional difference. In each site, equal numbers of boys and girls from the younger cohort, now aged 6/7, and from the older cohort, now aged 12/13, have been selected on the basis of additional criteria, including ethnicity/caste, religion, socio-economic status, household structure and pre-school attendance. The aim is to generate qualitative case-level data that can be integrated with the larger sample survey data, thus creating the potential to produce 200 longitudinal embedded case studies.Footnote 2

2 Organisation of the Paper

The first section of the paper provides an overview of child-focused, highlighting key literature and examples in this field. The second section describes three core methods used to explore subjective wellbeing within Young Lives (Wellbeing Exercise, Life-course Timeline and Body Mapping) and reviews the experiences of piloting these with younger and older children in rural and urban sites in Ethiopia, India, Peru, and Vietnam. The final section reflects on the lessons learned, the challenges of piloting these methods and implications for developing a broader child-focused methodology.

3 Literature Review

3.1 Conceptual Background

Young Lives qualitative research is grounded in an anthropological approach, emphasising holism, context, process and the diversity of children’s experiences. Reynolds (2006, p. 295) contends that the “manifest advantage of the anthropological perspective is that it specifically looks at practices, how they come about and how they relate to moral categories and ideas of a good life on the ground, where everyday life is enacted.” At the same time, Young Lives qualitative research is informed by interdisciplinary insights which link changes in children’s lives and wellbeing to their development, understood as “a process of change in which a child learns to handle ever more difficult levels of moving, thinking, speaking, feeling and relating to others” (Myers 1992, p. 4). Our methodology aims to reflect these approaches for understanding children’s lives.

Wellbeing is an important but somewhat elusive concept. Like the concept of poverty, it is open to numerous definitions, conceptualisations and methodological approaches. Our underlying assumption regarding child wellbeing mirrors our view of childhood—that wellbeing is a socially contingent, culturally-anchored construct that changes over time, both in terms of individual life course changes as well as changes in socio-cultural context. Imposing concepts and measures based on what it means for a 2-year-old boy in Los Angeles to be ‘doing well’ on a 12-year-old girl growing up in rural Ethiopia would be a weak starting point for analysing child wellbeing, as even if both children are growing up in poverty, differences of cultural and historical context and individual life phase are highly significant. In the qualitative work, we want to start with as few assumptions as possible regarding emic (or ‘local’) understandings of the potentially diverse ways in which wellbeing is experienced by different children in the study contexts so have used open-ended questions and semi-structured methods to explore these. Emic perspectives are valued as researchers “work with people to define their reality and challenge imposed knowledge” (Veale 2005, p. 254).

That poverty has many dimensions is now widely accepted (Boyden 2006; Harpham et al. 2005; Narayan et al. 2000; White et al. 2003, p. 381). Yet much of the literature on childhood poverty focuses on the experience of loss or deficits—of income, material possessions, shelter, health, education and so on—and therefore on what children lack (Boyden et al. 2003). While poverty is fundamentally about deprivation of one kind or another, efforts have been made to highlight the important things that children who grow up poor may have, such as their resourcefulness, resilience, and optimism, with an emphasis on their sense of agency (Save the Children 2001, in Streuli 2007). Despite growing up in adverse conditions, some children come out ‘okay’ while others do not (for example, Garmezy 1993; Rutter 1987; Werner and Smith 1998). What differentiates the child who is vulnerable to negative outcomes from adversity, compared with the child who appears to cope despite adversity has been described as ‘resilience’, the ability to ‘bounce back’ from stress, or “to recover from, adapt, and remain strong in the face of adversity (Masten 2001 in Boyden and Mann 2005, p. 6).” There are other particular factors that moderate and mediate poverty experiences and outcomes (Boyden and Cooper 2007) such that even children growing up in the same household, under the same material conditions, and reared by the same caregivers may experience different trajectories in relation to poverty (e.g. Elder et al. 1993). A child-focused approach acknowledges that children are diverse in their capacities for resilience, in the impact of risk and protective factors, and that similar experiences and conditions may nonetheless yield different wellbeing outcomes amongst any group of children.

According to Fattore et al. (2007, p. 7), recent efforts to make child wellbeing research more child-focused have “moved the thinking on children’s well-being from a focus on survival and basic needs to beyond survival…from negative to positive, from traditional domains to new domains, and from focussing on preparation for adulthood (well-becoming) to the present lives (well-being) of children.” The focus of Young Lives qualitative research is on subjective and psychosocial wellbeing which capture individual perceptions and the meanings people give to different aspects of their lives (Denzin and Lincoln 1998, p. 3). These perceptions and meanings encompass both present and future lives (i.e. ‘being’ and ‘becoming’) which are important dimensions of children’s subjectivity as they are of adults’—caregivers, teachers, etc (Uprichard 2008). The initial research efforts of the qualitative team have focused on generating meaning(s) (and less on measuring outcomes) as the aim at this stage of research is not to produce an ‘index of child wellbeing’ to be used across Young Lives study countries. Instead, the impetus is to use qualitative methods to generate local understandings of the perceived risks and opportunities children face on a daily basis and in the long-term, and what they need to do well now and in the future. The idea is to document contextualised indicators of what it means for ‘these’ children, in ‘this’ place, to be doing well (or not), and to explain changes in expectations and experiences over time.Footnote 3

Intergenerational dynamics is another key concept in Young Lives, especially in relation to the transfer of poverty across the generations, but also related processes whereby children learn and negotiate deep-seated assumptions and values, expectations, habits and orientations, especially through their day to day interactions with parents and other family. This is a reciprocal process, however, poignantly illustrated by findings reported by Ridge (2003) from the Listening to Children Study which involved participatory research with a group of 10–17-year olds growing up in low-income families in the UK. The study found that children struggle to protect their parents “from the realities of the social and emotional costs of childhood poverty in their lives”, just as parents, especially mothers, worked hard to protect their children from the worst effects of poverty (p. 8). Young Lives research will similarly be enriched through in-depth study involving a spectrum of adult and child actors who can offer differing perspectives on children’s lives, as well as providing opportunities for triangulation of data. Involving caregivers and other adult respondents recognises the influence and decision-making power they often have over children, and respects that their views and wishes may not be the same as those of children. Children and adults may differ in what they consider ‘risky’ behaviour, ‘useful’ skills, important transitions, the value of school versus economic activity, as well as in the decisions they consider in children’s best interests (Boyden et al. 1998, p. 30; Woodhead 2001; Harpham et al. 2005).

3.2 Methodological Approaches

The Young Lives qualitative methodology may be characterised as:

  • Qualitative and longitudinal

  • Child-focused and participatory

  • Multi-actor

  • Flexible and reflexive

  • Mixed- and multi-method

  • Responsive to ethical issues

This methodology has informed the development of a toolkit of methods that can be applied in diverse cultural contexts, marked by variations in children’s daily lives, their relationships with adults (including adult researchers), and preferred ways of communicating their ideas and feelings. The toolkit includes methods based on drawing (e.g. ‘Lifecourse draw-and-tell’ and ‘Happy day/Sad day’), writing (e.g. ‘Daily activity diary’), talking (e.g. semi-structured interviews) and on other activity-based techniques (e.g. photo-elicitation and child-led tours of the neighbourhood). The individual methods for data gathering form part of a broader methodology that was inspired by several recent strands of research developing child-focused participatory techniques. A ‘children’-focus acknowledges that research with young people may pose special challenges for adult researchers, particularly in highly hierarchal societies that marginalise children’s views (Boyden and Ennew 1997; Punch 2002b). Children’s participation has therefore been linked to the goal of respecting children’s rights, embodied in the UN Convention on the Rights of the Child (1989) and in relation to their ‘evolving capacities’ to participate, which is viewed as an enabling rather than a restrictive concept (Lansdown 2005).

Hart’s (1992) well-known ‘ladder of participation’ emphasises participation as a continuum, and that children may participate to varying degrees at different stages of a project. Within the context of Young Lives research, participation is a form of communicating and engaging with children to forge a relationship of mutual respect and trust. Beazley and Ennew (2006, p. 192) argue that “no method is inherently ‘participatory’…it depends on how a method is used” and is therefore tested ‘in action’. They observe that in Development Studies, the group being researched is almost always a vulnerable group (in this case, poor children), which presents many opportunities for misunderstanding and exploitation (Beazley and Ennew 2006, p. 189). Building child participation into the research agenda may work to break down some of these power differences based on age and material privilege.

Other child-focused methods have been influenced by Participatory Rural Appraisal (PRA) techniques that have been used with adults in developing country contexts (Chambers 1994). These are generally task- or activity-based and can be highly visual, involving, for example, mapping, ranking, and photography. Selecting methods for use with children is unique in the sense that their potential ‘fun factor’ is often considered, possibly reflecting the (in some ways questionable) assumption that ‘children’ have lower ‘attention spans’ for research than do adults. Furthermore, methods for use with children often mirror the activities they are thought to experience as part of their daily lives in school, at play, or at home, although in some cases this may be based on researchers’ ethnocentric assumptions about what children do and like, informed especially by Western developmental psychology and child-centred peadagogies. Data have been generated with children through the use of, for example, diaries (Baxter 2007; Frankel 2007; Nesbitt 2000; Tekola 2007), photography (Barker and Weller 2003; Clark and Moss 2001; Einarsdottir 2005; Gabhainn and Sixsmith 2006; Orellana 1999; Punch 2002b; Morrow 2001), draw-and-tell techniques (Driessnack 2006; Ennew and Morrow 1994; Pridmore and Bendelow 1995; Punch 2002b), mapping (Darbyshire et al. 2005; Veale 2005), time-use charts (Christensen and James 2000), drama and role play (Armstrong et al. 2004; Veale 2005), video clips from soap operas (Punch 2002a) and the use of radio workshops and recordings (Frankel 2007), among many other examples and creative techniques.

One of the advantages of using visual data, such as photographs and drawings, or ‘active’ methods, such as child-led neighbourhood tours, is that children with limited literacy may participate in the research (Clark and Moss 2001). Visual data enable the researcher to ground discussion in children’s experiences and social environments thus making the interpretative process more collaborative (Hart 1992; Mauthner 1997; Smith et al. 2005, p. 484). For example, photo-elicitation techniques involve children using cameras to capture aspects of their lives for discussion. For example, that they can take pictures of whatever they choose may affirm their agency and ultimately influence the direction of research through the images they capture, as discussion and interpretation is grounded in the visual data they produce (Einarsdottir 2005, p. 527). Children are also able to capture spaces and aspects of their social worlds to which adult researchers may not have access, such as intimate home environments or ‘hidden’ spaces of child labour (e.g. Young and Barrett 2001).

Other methods are more heavily reliant on literacy, such as the use of diaries which some children may consider ‘too much like schoolwork’ and therefore not enjoyable. However, diary writing can also be a preferred medium for young people who like that it is ‘like schoolwork’, as Tekola (2008) has documented in her research with children in Addis Ababa, Ethiopia. Theis (1996, p. 72) described an experience in Vietnam where “One 12-year-old girl…sat quietly in a corner of her house and did not respond to the interviewer’s question. She was, however, quite happy to write her daily activity schedule on a piece of paper. When the interviewer later tried to talk to the girl again, she stood up and left the room.”

3.3 Younger Children

Young Lives is innovative in its inclusion of relatively young children in qualitative research (ages 6 and 7). Young children have until recently been neglected when it comes to involving them in research about them and the things that matter most to their wellbeing (Alderson 2000; Clark et al. 2003; Clark 2005; Clark and Stratham 2005; Cousins 1999; Lancaster 2006; Miller 1997; Mortimer 2004; Reynolds 1989). Failure to include them in research may be due to lack of creativity and appropriate adaptation of methods on the part of researchers, a belief that young children cannot be reliable sources of data, or an assumption that caregivers ‘know best’ so can speak on behalf of young people. Boyden and Ennew (1997, p. 33) go further to suggest that not including young children in research reflects a resistance on the part of adults to “the idea that small children might be allowed to take major decisions about their own and other people’s lives…they imagine that giving children the opportunity to voice their opinion is the same thing as allowing them absolute control, letting them take over the world”.

One of the areas where young children’s views are increasingly being sought is in relation to their experiences around key early childhood transitions, particularly their entry into formal schooling and other institutional settings (Clark and Moss 2001; Dockett and Perry 2005; Lancaster 2006; Ledger et al. 2002; Miller 1997). For example, the Mosaic Approach, developed by Clark and Moss (2001) is a framework for ‘listening’ to children’s views, drawing on the image of a mosaic as a visual analogy of how children’s and adult’s views may come together to offer a fuller picture of children’s lives—each method and/or each person’s perspective representing a unique tile in the ‘mosaic’. It was developed with 3- and 4-year olds over an 18-month period in a UK pre-school as a way to involve young children in reviewing their ‘early years’ services. The Mosaic Approach offers a portfolio of tools to involve both children and adults in gathering ‘documentation’ for the research, with each piece of information providing one piece of the ‘mosaic’. The methods used in the study included observation, child conversations,Footnote 4 child photography, child-led tours and mapping of the care setting, role play and collective interpretation with children and adults based on the documentation generated. The authors argue that it is the mosaics that are created through the research that offer the richest and most useful data and not the individual ‘tiles’ from which they are built.

Other studies on children’s perceptions of their early childhood care and education settings, primarily in industrialised countries, have relied on ‘talk-based’ methods as the primary data-gathering tool (Dupree et al. 2001; Campbell-Barr 2003; Einarsdottir 2005, p. 526). In Sweden, Sheridan and Samuelsson (2001) interviewed 5-year olds in order to understand how they perceive their opportunities to exercise influence in their pre-schools and how their experiences relate to the quality of their institutions (as evaluated by ECERS, the widely used Early Childhood Environment Rating Scale, e.g. Harms et al. 2005). In England, children’s perceptions of their nursery school were elicited through interviews with 4-year olds about why they attend, what they like and dislike about nursery, using creative methods such as play telephones (Evans and Fuller 1996, 1998).

3.4 Combining Methods in Child-Focused Qualitative Research

Any one of the above methods described thus far may be effective tools for accessing children’s perceptions and experiences. However, there is a growing tendency towards multi-methods approaches within child-focused research, as in the Mosaic approach (See Table 1 for examples). The design of multi-method child-focused research has been aided by the publication of methods guides, such as Boyden and Ennew’s (1997) Children in Focus: A manual for participatory research with children and extensive reviews of the successes and challenges of piloting specific methods, for example, Armstrong et al.’s (2004) Piloting Methods for the Evaluation of Psychosocial Programme Impact in Eastern Sri Lanka. Johnston (2006) produced a review of child-focussed methods tested during the course of Young Lives pilot work in Peru which formed the basis of further piloting in the four study countries (2007). Multi-method approaches recognise that, as a social group, children possess a broad range of capacities and preferences for expressing themselves. In practical terms, combining more than one qualitative technique may help to sustain children’s interest in the research ‘moment’, providing varied opportunities and modes for expression. In terms of data quality, it enables triangulation of research data.

Table 1 Examples of child-focused studies combining methods
Full size table

Researching wellbeing with children lends itself to a multi-method approach, given that wellbeing is an umbrella concept that captures individual perceptions and the meanings people give to different aspects of their lives (Denzin and Lincoln 1998, p. 3). Its holistic character means that many qualitative methods could be used to explore aspects of wellbeing and it is the complementarity of the three techniques discussed here—the Wellbeing Exercise, Life-course Timeline, and Body Mapping—that make them key to our data gathering strategies with children around the issue of wellbeing.

Examples of other studies that have used these methods are well-documented. Hubbard and Miller (2004) originally used interviews to elicit local conceptions of wellbeing (alongside a community needs assessment) as part of an evaluation of ecological mental health interventions in refugee communities. Hubbard’s method was developed further and used as the basis of a group discussion with parents and children in five villages in Eastern Sri Lanka (Armstrong et al. 2004, pp. 44–45). The purpose of the discussion was to elicit local understandings of what constitutes wellbeing and illbeing (rather than its causes) using the question “what is it about the person that tells you that they are doing well?”. There were some differences between children’s and adults’ responses, for example, children emphasised the quality of relationships (i.e. “being loving or kind (anbu)”), while parents focused on mixing well with others and emphasised obedience over good habits.

Variations of the Life-course Timeline have also been used with children. For example, James (2005) describes how two different kinds of timelines were used in research carried out with 10-year-old working class children in England (see also Christensen and James 2000). The two paper charts, entitled respectively ‘My Life’ and ‘My Next Year’ captured children’s subjective understandings of the life-course, of the ageing process and the ageing body. In another study, the “Sibling Relationships in Middle Childhood” project, time charts were one of the tools used with a group of children aged 8–12 in England.Footnote 5 The time charts enabled children to write or talk about specific events in their lives and how their relationships with their siblings had changed over time.

Body maps have also been used with young people and in a variety of settings. They were recommended by Ennew and Plateau (2004) as a ‘visual stimulus’ in research on children’s experiences of physical and emotional punishment; for example by asking focus groups of schoolchildren to identify ‘places used for punishment’, ‘vulnerable places’, ‘private places’, and ‘places that hurt’ (p. 215). They were also used with both children and adults in an impact evaluation of a psychosocial programme in Eastern Sri Lanka to look at what caused pain or sickness, common conditions in the community, and how people could stay healthy (Armstrong et al. 2004). Adults in particular placed physical problems in their socio-economic context, for example, a cut foot was linked to women worrying about financial problems whilst clearing forest as labourers was something that was now necessary because their spouses had been murdered or disappeared (p. 16).

4 Studying Well Being in Young Lives’ Pilot Research

Adaptations of the three methods described above were tested in eight pilot studies carried out in the four Young Lives study countries. The pilots involved younger and older children, caregivers, teachers, and other community members. Most pilots covered the full range of group and individual methods that had been shortlisted for inclusion in a shared methods toolkit, although some teams worked intensively with the ones that were most likely to cause problems. Group-based work was intended to establish rapport with children, to explore consensus (Lewis 1992, p. 420), and to enable children to choose their own levels of disclosure (Hill et al. 1996, p. 139). The presence of peers may also shift the power-balance between adults and children, giving children the confidence to speak and participate (Hill 1997, p. 175). Group-based research was followed-up with individual or paired interviews to investigate personal experiences in greater depth.

The pilots took 2–3 days per site and were undertaken in one rural and one urban location in each study country (or in Vietnam one rural and one ethnic minority site, as this is an important source of difference). They were preceded by 3–4 days training in the selected methods and in ethical considerations in fieldwork, which included adapting them to the local context. Each research team built in time for methodological reflection and skills reviews during fieldwork, and had a full debriefing session at the end of the pilot period, which resulted in a set of pilot reports. These formed the basis of discussions for agreeing a shared methods toolkit to be implemented across the four study countries in the first round of data gathering (August–December 2007). The data presented below is for illustrative purposes only, as the pilot reports focused mainly on detailed cohort-specific assessments of each method.

5 Wellbeing Exercise (Group Discussion or Wellbeing Draw-and-Tell)

The Wellbeing Exercise explores what children and adults consider a ‘good’ or ‘bad’ life for children of the same or a specified age and gender, living in their community, and what they identify as sources of risk and protection.

5.1 Process

With adults, this was piloted as collective interviews with caregivers of case study children and with groups of community members. It was also a core activity with older and younger children (with modifications), where it was primarily carried out as a group ‘draw-and-tell’ exercise, with subsequent one-on-one follow-up interviews.

The method was carried out with older children in various steps, beginning with an informal ‘focusing’ activity to get them ‘thinking’ about the research topic. In Peru and India, for example, after an ice-breaker game, participants stood with the facilitator in a circle with their eyes closed and were asked to ‘picture in their mind’ a child they knew in their community who they thought of as experiencing ‘wellbeing’. Teams tried to capture ‘wellbeing’ using different phrases appropriate to local languages, for example (translated into English as), a child ‘for whom life is going well’ (in Peru), a child who has a ‘good life’ (in Ethiopia), a child who looks ‘good’ (in India) and a child who is ‘happy’ (in Vietnam). Rather than being discouraged by the variation in how the concept was back-translated into English, we accept this as a reflection of the diverse socio-cultural and linguistic contexts in which we are working. Even in English it is rare to speak directly about a person’s ‘wellbeing’. We were therefore aiming to capture children’s understandings of differences between children in terms of what kinds of children are doing well or not and the basis for that differentiation. We acknowledged that variations in the way ‘wellbeing’ gets translated across the different contexts may capture different aspects of our wellbeing concept.

After initial work to establish a shared concept of wellbeing amongst a group of older children, one facilitator in India incorporated another ‘focusing’ activity whereby the group were asked to imagine 100 people in their community lined up side-by-side. The children were asked if everyone was the same. “They responded that ‘[T]here is a difference… some are having money, some are not, some look good, some look bad’” (India pilot report 2007). The children were then asked to think only about the young people in their community, and to think of one child who ‘looks good’ (is experiencing well-being) and one who doesn’t ‘look good’ (illbeing) and what the differences between the two might be. The children then sat down to draw their individual images of a child experiencing wellbeing and a child experiencing illbeing, followed by discussion. In Peru the exercise also focused on children’s drawings and group discussion around them. The groups of older children collectively ranked their ‘wellbeing’ drawings on the one hand, and the ‘illbeing’ drawings on the other, which provided a sense of those clusters of indicators they associated most with doing well or not. Note taking recorded the results of the ranking, as well as the individual stories and group discussions that emerged from drawings. However, in Ethiopia children were reported to prefer to have a collective discussion with their suggestions written on a flip chart divided into columns for ‘good life’ and ‘bad life’, even though this gave the exercise the dynamics of a lesson (which is not surprising, considering the sessions took place in classrooms). While in Peru children ranked drawings representing child wellbeing and child illbeing, in Ethiopia they ranked individual indicators of wellbeing and illbeing according to desirability.

In all of the countries, the facilitators found it challenging to implement the method with young children in the same way as it was being carried out with the older groups. Neither individual drawings, nor group discussion seemed to be very effective techniques. The Peruvian team experimented by offering a higher level of ‘scaffolding’ to support children’s constructions of well-being. The facilitator gradually drew the wellbeing and illbeing pictures, but the choice of features was entirely under the children’s direction. Children were asked about every detail of the drawing to be done to answer the general question “how is a child for whom life is going well/not well?” (for example, what is their face like, their clothes, things they have, family situation, school attendance, etc). The drawing was shown to them frequently (not only at the end) to check the facilitator was depicting their ideas correctly. In practice, the children and the facilitator most often started by co-constructing a representation of illbeing as this was less abstract and meant that the exercise could finish reassuringly on wellbeing.

5.2 Data Generated

Older children’s conceptions of child wellbeing and illbeing in Vietnam often focused on their school environment, friends and studies. Wellbeing was seen as having a healthy and happy life with no sickness, and being loved by their parents and people around them. Illbeing meant a life without family, without a house, and without love from parents (for example, being beaten by parents and not being cared for while they were sick). Children wanted their siblings to be able to go to school, their families to have enough to eat, and for their parents not to quarrel.

In Ethiopia good experiences included attending school like their friends and being healthy; bad experiences related to family and education, for example, parents divorcing or becoming ill, having to live separately from their parents (for example, to attend school), having a long journey to school or needing to leave school early to start work, and being beaten. Children described wanting to complete their education with good results (many rural children didn’t expect to continue beyond grade 8 because high school was too far away and expensive), help their parents, have a harmonious family, and marry and have children. Indicators of child wellbeing mentioned by adults and children in Ethiopia included having a harmonious family and good relationships with others, fulfilling basic needs (including having sufficient food), and behaving well. Children wanted to attend school and have support in doing so (for example, time to study and sufficient materials), and also to have time to play and resources such as footballs to use in games with other children. Children characterised as experiencing illbeing were orphans or from a poor or ‘broken’ family, had insufficient food, needed to start work at a young age, and behaved in a dishonest or disorderly way. Perceived threats to children were not restricted to family, school or community, but also related to environmental and global issues (one young child from the urban site described how “when I see a program on HIV/AIDS, I usually turn off the TV set because it disturbs me”).

Family was the central source of wellbeing for all children in Peru, specifically the presence, love and support of parents, as lack of one or both parents put girls especially into an insecure and vulnerable situation. The notion of interdependence was apparent in sessions with older children, as they stressed the importance of each family member understanding and fulfilling their complementary roles. Children described as experiencing illbeing in the urban site did not have parents or siblings; in the rural site they had parents, but their parents did not take proper care of them, for example, they did not wash their clothes or comb their hair and often beat them. Education was also an important indicator, followed by material things like nice clothes and housing. Urban children presented particularly vivid images of illbeing (e.g. pirañas and gang members), and discussed poverty and lack of basic services (piped water, electricity), as well as drug and alcohol use. Older girls in the rural and urban sites described their fear of sexual abuse, for example, having to work alone in the fields, which puts them at risk of being raped. Supernatural beings were also linked with illbeing; for example one young boy drew a boy that was snatched by a soul (alma), implicitly as a punishment for wandering around in the street alone, and urban children frequently mentioned a movie character called Chucky (a demonic doll) who gives them nightmares. A clear association was made between being good, i.e. respectful and obedient, and being well. Social and affective indicators also seemed to be stronger than, or inseparable from material indicators; for example, having old and dirty clothes indicate not only lack of material resources, but also that children do not have a family that is concerned about them, their appearance, and their needs.

Comparison of responses from adults and older children in India demonstrate that children were able to provide detailed and wide-ranging indicators of wellbeing and illbeing, which were embedded in local environments (for example, playing in drainage ditches) and spanned social, physical, economic, and cultural characteristics (Table 2). Interestingly, the caregivers’ indicators related primarily to the behaviour and appearance of the child, whereas the children’s encompassed other family members, primarily their parents, and relationships within the family as a whole (See Table 3). In Ethiopia adults and children seemed to have a shared understanding of children’s wellbeing and respondents attributed this to interdependence (if children were successful they could support their parents in the future) and socialisation where children are ‘taught to expect what their parents could afford.’ When children in the Ethiopian sites were asked what wellbeing was for them and how it could be achieved, they replied in terms of things that were valued within their community, good for their family, or achievable within their context. Older boys, for example, said ‘we demand what parents can afford, because they always tell us to ask for what they can do.’ Nonetheless, in the urban site some children wanted things they saw in the community (for example, extra tuition or karate classes) even if they knew their parents could not afford them. The gap between children and adults was also apparent when they discussed future expectations—while most children wanted to focus on their education, parents believed that they should work for the family as well and playing and study time were rarely considered as important).

Table 2 Indicators of child wellbeing and illbeing generated with adult caregivers, India
Full size table
Table 3 Indicators of child wellbeing and illbeing generated with a group of older boys, India
Full size table

5.3 Advantages of the Wellbeing Exercise

The method was engaging and relatively easy to use with older children, and as a basis of group discussion with adults). When carried out initially in a group setting, it raised further issues for discussion in individual interviews which produced more detailed information. It generated a variety of information from different perspectives, especially relating to how children’s lives differed from their ideals. For example, in Ethiopia it produced data relating to economic, social, environmental and personal issues, although some children focused on morality and behaviour. The Wellbeing Exercise was also useful for exploring gendered differences in adult and child conceptions of child wellbeing and illbeing.

5.4 Challenges of the Wellbeing Exercise

The concepts of wellbeing and illbeing proved difficult to translate, especially in Vietnam, although in India researchers were able to find equivalent phrases in Telugu. In Ethiopia researchers talked about good and bad lives rather than wellbeing or illbeing as this was more easily understood, however, this meant that some children’s accounts had a moral tone. Similarly, there was initial confusion in India between children who were experiencing wellbeing and ‘looking good’, and in Vietnam where children who didn’t understand the concept of wellbeing were asked to draw the ‘best’ children in their class or neighbourhood. Some children initially drew other children they liked or disliked rather than children who were experiencing wellbeing or illbeing, raising ethical concerns. Ethical issues were also raised when group discussions highlight different ‘groups’ of children (e.g. ‘poor’, ‘not doing well’, ‘doing well’) given participants are themselves growing up in contexts of poverty and are differentiated. Likewise, it is sensitive to discuss children’s family lives in a group setting, although disclosures could be explored further in individual interviews. Potentially, the method might reinforce social divisions inadvertently. Children found it difficult to talk about wellbeing/illbeing or good/bad lives in the abstract, but drawings helped to ground discussion in children’s images. Sessions required active facilitation, including continual probing to explore different layers of meaning, and it was sometimes difficult to maintain children’s concentration beyond 20–30 min. Some individuals can dominate the conversation, while others are reluctant to talk about anything that could reflect badly on their families. The group method is therefore appropriate for gathering shared understanding rather individual or personal stories.

6 Life-course Timeline

The Life-course Timeline explores children’s life experiences, focusing on what they remember as the important moments of their past (both happy and sad) and why these were memorable. The method enables exploration of how children feel about their current situation, their future expectations, and the extent to which their aspirations are achievable and shared by their parents.

6.1 Process

Teams experimented with different ways of generating life-course timelines with children. In some cases, children were asked to draw a timeline from birth to 30, or to start by drawing themselves at their current age on an existing time line. In others, pre-drawn images representing a growing child were provided on individual timelines that were folded in half, the left side representing the past and present (which was filled in first), the second half representing the future. In India, older boys insisted on working together in pairs (or ‘teams’, which they named for competitive purposes, insisting the facilitator judge their drawings at the end of the session!). The pairs worked on shared timelines which they used to draw images of themselves in the past and as they see themselves growing up. In Ethiopia, the older children wrote their ‘good’ memories above the line and their ‘bad’ ones below it. In Vietnam, the children drew a ‘happiness’ graph indicating how they felt overall during the period represented.

Few children used calendar years to indicate the passing of time, while age was a more useful marker. The future was always less detailed than the past/present, though they could generally indicate changes and events anticipated over the next 5 years; for example, what they want to do (e.g. go to secondary school, get married), what support they’ll need to do this, and the extent to which their parents share these aspirations. Although the exercise was carried out as a group method in the pilot, it was considered more suitable to plan this as a one-on-one activity with individual children to explore in-depth personal biography.

6.2 Data Generated

Most data generated in Vietnam with the older cohort of children related to school, even though the sessions actually took place in local community centres or ‘cultural houses’. For example, some children described an increase in pressure to perform well in school following the transition from primary to secondary school. They also felt less able to discuss their experiences or ask for support from parents (e.g. in doing homework) and now relied on friends/class mates. Happy memories also related to school, for example, receiving awards, going on picnics and study tours with teachers. Unhappy memories from Ethiopia covered a wider range but were mostly related to their families, for example, death and illness in the family (including parents), family breakdown, loneliness due to lack of siblings, beatings, neglect, hunger, leaving school, and robbery.

In Ethiopia children described pleasure at educational achievements, but also at eating good food, helping parents, recovering from illness, receiving new clothes and other presents from parents, celebrating holidays, and travelling. In the future, many Vietnamese respondents wanted to be policemen as they saw them as having a heroic role in protecting society, while Ethiopian respondents hoped to be doctors, teachers, or pilots (“I want to be a pilot in order to eradicate poverty from my family and country and introduce my country to the world”, 12-year-old girl in Ude). More generally Ethiopian respondents described wanting to be loved, help their parents (“I want to discover a medicine for asthma to cure father who suffers from it”, 12-year-old girl in Kolfe), and be good students. Ethiopian respondents also spoke about their fears, namely death, separation, educational failure, early marriage, HIV/AIDS, and in one case environmental catastrophe (a theme that also emerged in the Wellbeing Exercise).

6.3 Advantages of the Life-course Timeline

The method was useful in generating cross-cutting information on both childhood transitions and child wellbeing. In Peru, a life-course timeline was created with a group of mothers in Peru as part of a discussion on important childhood transitions and wellbeing that can be compared with data generated from their children. Children found the exercise relatively easy and some enjoyed illustrating their timelines with colourful drawings; others enjoyed not having to draw as they were given the option of using pre-drawn timelines illustrated with stick figures.

6.4 Challenges of the Life-course Timeline

Although the Life-course Timeline was piloted as part of a group activity, it was later agreed that discussion of children’s individual biography should take place individually with the researcher. It was therefore not suitable as a group method. The method would also require adaptation in contexts where time is not conceived linearly. Children generally found it easier to describe the past than to talk about the future. Nonetheless, nearly all of them found it difficult to remember the first 5 years of life, which made the exercise particularly challenging with younger children. Few children recalled the year or age they were when events happened. In Vietnam and Ethiopia it worked better to ask about the things children remember the most, rather than “important” or “major” events as this sounded overly important. For example, in Vietnam only two children said they could remember anything important; one child mentioned his parents’ distress at the loss of their sugar cane harvest, and being very hungry aged 8 as his parents didn’t have any rice and they had to eat cassava instead.

7 Body Mapping

Body mapping is the exploration using a common pictorial image of a body (a ‘body map’) of what makes children feel good or bad, where these feelings are located on their bodies, how they make themselves feel better and who, if anyone, helps them with this. It is a group-based activity which also generates accounts of specific experiences that can be discussed or followed up individually during subsequent interviews.

7.1 Process

The facilitator asks for a child volunteer, or if they aren’t willing the facilitator volunteers, to lie down on a large piece of paper and be drawn around. Younger children in Peru were able to practice this by drawing around their hands. Children were then asked to think about parts of their body where they felt good or bad (for example, a headache or a full stomach). After they had thought for a few minutes, they wrote these experiences on separate self-adhesive notes, which they placed on the respective part of the body map. In Peru this was preceded by a game of ‘Simon says’ to make sure that the younger children could identify body parts. Children were then asked to talk about these experiences, for example, what happened and why, what did they do to make it better and who helped them, and were there any positive outcomes? The last question, ‘were there any positive outcomes?’ produced interesting responses in Ethiopia where children talked about getting better food and being encouraged to rest when they felt ill. Children were also asked to rank their experiences according to frequency and impact (for example, headaches were common, but not regarded as serious). Children from the older cohort described both expected and unusual sources of physical illbeing, for example, their feet were affected by the long walk to school and by having to fetch water and grind grain, as well as playing football. However, more surprisingly, their knees were affected by having to kneel down in front of teachers and being required to clean the school office and toilet. Older children also described what they did to make themselves feel better (“I drink pepper with water when I feel abdominal pain”, 12-year-old boy from Ude village) and how they were treated with herbs and other traditional medicine due to the cost of formal healthcare (“when I was bitten by a dog, my mother gave the dog injera with my blood. She told me if the dog eats it you will be cured. She was right”, 12-year-old girl from Ude village).

7.2 Advantages of the Body Map

The method is multifaceted and relatively easy to use with older children. It can be a starting point for exploring key events, relationships, important sources of support, and risky places. In Vietnam, for example, it provided surprising amounts of information about children’s relationship to their school and the attitude of their teachers, also on relationships within the family, parenting practices, daily activities, close relationships and influences, and local service access. The method identified diverse sources of illbeing, many of which were related to education, for example, beatings at school, subjects that need to be learnt by heart and require a lot of homework (Vietnam), and punishments for not doing homework or getting a bad grade (Peru, urban community). Children seemed to enjoy it as it was relatively physical and interactive, although younger children could become frightened or embarrassed (e.g. when lying down to be drawn) and tended to imitate other children in their responses. The discussion was generally lively and wide-ranging, although younger children found it hard to explain why things hurt or remember how they were made better.

7.3 Challenges of the Body Map

Body Mapping requires active facilitation, for example, discussing answers collectively and generalising them to the group (e.g. ‘have you experienced this?’). Otherwise it could be time consuming or boring if children separately write experiences on post-it notes and tell their own stories to the facilitator at length without interacting with the other children. It was sometimes difficult to shift the focus from physical illhealth to explore more diffuse or positive experiences (often a problem of translation). The Peruvian team explored things that made children feel good, which were usually located in the heart (for example, playing with pets, parents buying a toy or something for them, mother cooking for them, etc). Children in Peru also described the embodied experience of other emotions such as fear in their belly (for example, of robbers, or movie characters like Chucky) or sorrow in their heart when their parents wouldn’t let them go out to play.

8 Methodological Reflections—How Children Told us About Their Wellbeing

The holistic character of the wellbeing concept meant that it could be explored from different angles using a variety of mutually-reinforcing methods. The Wellbeing Exercise was used with children and adults to capture intergenerational understandings of what it means for girls and boys (aged around 6 and 12) living in their communities to be doing well/not well, or to have ‘good’ or ‘bad’ lives. To a lesser extent, it was used as a basis in individual interviews for gathering case-level data on children’s experiences of wellbeing. The Life-course Timeline added a temporal dimension and enabled children to reflect on how important moments and markers of change in their past impacted on their wellbeing. It also framed discussions on children’s expectations for the future, and what could support and/or prevent them from achieving them. The Body Mapping technique generated information with children about their embodied expressions of wellbeing and illbeing, the risks they face and the people and/or services that help them. All the methods generated contextualised understandings of child wellbeing and important life-course transitions and were also effective in eliciting information on child poverty, without having to ask narrowly focused questions about the material things that children do or do not have.

We found that a flexible multi-method approach strengthened our ability to engage diverse groups of children and adults on the topic of child wellbeing. The pilot work resulted in an agreed methods toolkit that balances less-structured, activity-based data collection (e.g. drawing, photography, etc) with ones that are more obviously adult-guided (e.g. structured observations of educational settings, collective interviews with caregivers, etc). While group activities are less-powerful settings for gathering case-level data, these were critical for building relationships and familiarity. They were also important for observing peer interactions and showed how group dynamics can impact on data collection, with one or two children ‘dominating’ the session. As a result, we incorporated ice-breakers, games and physical movement into the research protocol to build rapport and encourage equal involvement. Another challenge to group work was finding ways to create productive peer interaction in order to avoid conducting one individual interview after the other which can be time-consuming and risks children getting bored and distracted. This often related more to the skill of the adult facilitator than to the children themselves.

Involving the 6-year-old cohort in the research was at times challenging, especially when applying the methods in ways more appropriate for the older children. We therefore incorporated significant ‘scaffolding’ into the activities, including allowing for more ‘guiding’ and supporting questions with younger children than would be appropriate or necessary with older ones. We also invited additional fieldworkers to join the groups in order to assist individual children (usually the ratio was one fieldworker to two or three children). In some cases, it was the facilitator and not the children who did the drawing, but this was always based on the children’s responses. We also felt involving the younger cohort in research activities was critical to building rapport with them for future rounds of data collection in which they will be involved. This is important given the longitudinal nature of the study and plans to follow these same children as case studies of the qualitative component until they are at least in their early teens.

Other challenges were more conceptual, including struggling to make the abstract concept of wellbeing meaningful, especially to the younger children. It also required translation into multiple languages and embedding into diverse cultural contexts. Further, we needed to ensure that the questions used to capture wellbeing would ultimately be consistent, easily understood by different groups of children, and comparable across countries. Our approach was that assessments of children’s wellbeing need to be contextualised, or ‘culturally anchored’ and should begin at the level of individual children’s experiences in their social contexts. For example, by asking children to express what it means to be doing well in relation to particular risks, spaces, people, expectations and life changes, and using this to elicit information on the processes that protect them from such risks. Using drawing was particularly useful in providing a ‘concrete’ focus for children to discuss these issues, and maintained their interest by framing group discussions around the stories they told about their drawings (Mauthner 1997, p. 26).

9 Conclusion

This paper has explored recent work within Young Lives to develop a methodology based on child-focused, qualitative and participatory approaches to capture what children, caregivers, and other community members understand by child wellbeing and how these understandings change over time. Placing child wellbeing at the centre of the poverty research agenda captures children’s experiences beyond ‘survival’ and ‘deprivation’ and recognises their potential for resilience in the face of adversity. Involving children directly in research is the only way to grasp how they perceive difference within their local contexts, both between groups of children and between children and adults. Our focus on subjective and psychosocial wellbeing moves beyond psychological approaches that emphasise individual needs and goals and that often fail to capture the collective and interdependent dimensions of children’s experiences. Interpersonal relationships were shown to be important for children’s subjective wellbeing; how well they are doing as individuals depends on the quality and strength of their immediate social relationships with their family and peer group, and in many cases was also tied to their institutional service access, especially in relation to schooling. However, an emphasis on children’s interdependence does not mean that policies aimed at households will necessarily impact positively or equitably on all the children living within them. Given that what is considered to be best for children may differ between (and among) adults and children, it is especially important to seek multiple perspectives on the changing sources of risk and protection for children’s wellbeing. Young Lives qualitative methodology reflects children’s interdependence by situating them as key actors within the broader social, economic, and political processes that shape their everyday lives, aspirations for the future, and pathways over time.