THE VOICE OF THE CHILD IN SOCIAL WORK PRACTICE: A PHENOMENOLOGICAL ANALYSIS OF PRACTITIONER INTERPRETATION AND EXPERIENCE

Child protection policy, legislation and frameworks in England are informed by the need to consider the child’s voice in any decision - making forum. However, defining the child’s voice and applying these legislative requirements to practice is largely interpretative. There is a lack of practice guidance when it comes to conceptualising , capturing, and interpreting the child’s voice to inform decision-making. This study was conducted prior to the COVID-19 pandemic and begins to explore social workers’ experiences of how the child’s voice is understood and interpreted in child protection practice. Semi-structured interviews were conducted with five practitioners from a long-term, child protection team, and analysed using interpretative phenomenological analysis (IPA). Three super-ordinate themes were identified relating to (1) the ubiquity and ambiguity of the child’s voice, (2) the importance of relationships, and (3) the weight of the voices. Implications for practice include the need for more investment in practitioner-child relationships, and greater representation of the child’s voice in the final stages of needs assessments.


INTRODUCTION
The rights of children to have voice and influence in matters that affect them are enshrined in the UN Convention on the Rights of the Child (1989).General comment No. 12 (2009) further establishes that all children have the right to be heard and taken seriously in matters which affect them, and that their views must be taken into account to determine their 'best interests'.
Child protection policy, legislation and frameworks in England are informed by this requirement and the Children Act (2004) emphasises the need for practitioners to consider the views and interests of children in their assessments, particularly in relation to their safety, security, wellbeing and education (Holland, 2010).There has been an increased focus on the child's voice in UK social work practice since the publication of Eileen Munro's (2011) independent review of child protection, which stated too much focus on compliance with rules and procedures neglected the needs and experiences of children.In response to the review, UK government revised the statutory framework, placing more emphasis on direct work with children and families (Department of Education, 2011).However, conceptualising the child's voice and defining what constitutes their maturity and ability to form their own views is largely interpretative (Smith, 1996;Prout, 2003;Bruce, 2014).Furthermore, tension between the child's right to be heard and their right to be protected means that determining their 'best interests' is often a complex and ambiguous process (Komulainen, 2007).
Child protection social workers gather information from the child and significant people or organisations involved in their care.This information forms the basis of assessment, in which practitioners are required to make judgements and decisions about the child, including potential risks and needs in accordance with local authority thresholds (Taylor and Whittaker, 2018).
The first 'Revolving Door' report published by Action for Children (2017), revealed that in the year of 2015/16 over 158,000 children were referred to children's social care and assessed by a social worker, but didn't meet statutory thresholds for 'child in need' or 'child protection' status.Working Together guidelines (Department for Education, 2018) state that these children should be assessed to identify any early help which may promote the welfare of the child and prevent further referral.However, of those 158,000 cases in 2016, only a quarter were signposted to early help services.The second report (Action for Children, 2018) revealed that, over a two-year period almost one in three children who were referred but did not get statutory support were then re-referred the following year, suggesting earlier opportunities to help these children had been missed and raising questions about how the child's voice informs judgements and decisions in social work practice.
Decision making in child protection is recognised as a highly complex and ambiguous process, which often requires practitioners to rely on past experiences, personal intuition, or unofficial sense making strategies (Saltiel, 2016), which are often acquired 'on the job' (Kettle, 2018) and not easily supported or justified by policy and evaluation standards.In addition, these decisions are often made under significant time and resource pressures (Tupper, et al., 2016).
The decisions made by social workers in these cases often have a significant and lasting impact on the children and families concerned (Healy, 1998;Benbenishty et al., 2015).For this reason, social and psychological researchers have long been interested in examining the decisionmaking processes of social workers (Cuzzi et al., 1993;Helm and Roesch-Marsh, 2017;Cook and Gregory, 2019), and the Department for Education recently commissioned its own study of factors influencing the decision-making process (Tupper et al., 2016).The study considered factors including the child's; gender, age, disability and family history, which influence the decisions made in child protection practice.
Interestingly, despite the increased likelihood of becoming 'serious' cases, referrals for children whose parents had a history of domestic violence, drug use or alcoholism were 8% less likely to progress to further action.While the study did not look at whether or not any early help services were involved, these results highlight the complexities of decision-making in child protection practice.Findings from a series of focus groups with final-year social work students (Whittaker and Havard, 2016), suggest that practitioner anxieties lead to patterns of fear-based practice such as 'putting procedures before clients' or 'doing what is lawful over what is ethical'.Similarly, findings from interviews and focus groups with Australian social workers (Valentine and Katz, 2015) found that decisions to refer to early help or not were influenced by whether the professional approached this from a needs or risk perspective.This further highlights the complexities of child protection decision-making and points to tension between child participation and child protection.Valentine and Katz found that these decisions were also influenced by practitioner attitudes around intervention as mandatory or voluntary services.
A recent study involving in-depth interviews with child protection social workers in Scotland (Kettle, 2018), found that practitioners feel continuously challenged by the need to balance two forms of tension.'Closeness distance' refers to tension between becoming so close to families that they are enmeshed and unable to recognise when children need protection, and becoming so distant that they are ineffective in their attempts to affect change.'Power over-power together' refers to the tension between social workers using their power and authority over a family and using it with them in a collaborative way.These results highlight the potential influence that the relationship between practitioner and family can have on eliciting and attributing value to the child's voice.
Despite these findings, and the wealth of literature surrounding social workers' decisionmaking, there is limited research into how the child's voice is captured and interpreted by professionals in child protection assessments and decisions.In fact, simply defining what the child's voice is, how to capture it, and how to determine best interests as a result is largely subjective and interpretive.An international review of literature on how the child's voice is captured in early childhood institutions (Clark, 2005), suggest that observations, role-plays and creative art are effective methods of capturing the child's voice.However, these findings are specifically related to children's participation in learning.O'Reilly and Dolan (2016) explored child protection social workers' experiences of using similar play-based skills in their direct work with children.Practitioners reported that the application of these skills in practice gave them greater insight into the child's world, and led to greater representation of the child's voice in the assessment process, however the study did not look at the format in which the child's voice was represented and how it was interpreted by the practitioner beforehand.
As previously discussed, much of the literature on the inclusion of the child's voice within child protection proceedings, suggests that professionals often face difficulties in balancing the need to listen to the child's voice and the duty to protect them (Bruce, 2014).Warrington and Larkins (2020) recently referred to this as a 'false juxtaposition', arguing that protection must not be considered distinct from participation, and pointing to serious case reviews which repeatedly highlight failures to listen to children and therefore ultimately failures to protect them.When evaluating child participation in Norwegian social care and protection, Vis and Thomas (2009) found that children over ten years old were more likely to be included than under tens and that age was also a predictor to the weight attributed to the child's voice.
However, a later study (Vis and Fossum, 2015) found that while the child's age along with practitioner attitude, skills and knowledge influence child participation, organisational culture is more influential.Finally, Weatherall and Duffy (2008) found that while children's views and wishes are always sought in parental separation disputes, they are not accurately represented, and similar themes were found in a review of case proceeding documentation in Scotland (Bruce, 2014).Several examples were found where the section to include the child's voice was not filled in at all, as well as examples in which the child's voice was noted as 'not discussed.'There were files in which the child's voice section reflected the views of a third party, and examples of the child's voice being noted from the perspective of a sibling or copied from a sibling's file.In addition, where there was evidence that the child's direct views and wishes had been taken, they were often 'filtered' and interpreted by professionals, before being presented for assessment.
Little attention has been paid in research to how the child's voice is conceptualised, facilitated, interpreted and then represented in child protection practices.This is likely to be an individual process, influenced by several factors, which can only be given the attention it deserves through qualitative exploration of social workers individual experiences.Therefore, the aims of this study were to explore how the child's voice is conceptualised, understood and interpreted in child protection practice, from the perspective of social workers.The individual experiences of child protection social workers were explored in relation to the meaning and value that they attribute to the child's voice and their experience of how this can contribute towards child protection processes, as well as any barriers to child participation and factors which may influence the value and weight attributed to the child's voice.

METHODOLOGY
Ethical approval for this study was granted by the Psychology ethics committee at Manchester Metropolitan University on 12 th September 2018 (Reference number Pa_2017-8_29_1)

2.1.Design
Conceptualisation of the child's voice, along with how it is understood and valued, is largely subjective and interpretative, therefore an idiographic interpretative approach was required.As this study was interested in how individual social workers "create systems for meaningfully understanding" (Raskin, 2002:1) what the child's voice is, and how they make sense of it, the design was informed by constructivist interpretative phenomenology.Interpretative Phenomenological Analysis (IPA) is an idiographic approach to qualitative enquiry that is rooted in phenomenology and hermeneutics.Despite IPA being largely applied in health and psychological research, the aim of the approach is to understand how social actors make sense of their experiences and day-to-day consciousness (Houston and Mullan-Jensen, 2012).It was therefore considered an appropriate analytic strategy for exploring the individual's subjective experience of conceptualising and interpreting the child's voice in day-to-day social work practice (Smith et al., 2009).
All participants gave informed written consent to take part in the study.Their data was anonymised prior to analysis.Pseudonyms are used throughout to preserve the anonymity of participants.

Sampling
This was an exploratory study, interested in social workers' subjective experiences of including the child's voice in their day-to-day practice.Therefore, a small purposive sample enabled indepth exploration of how individuals conceptualise the child's voice and attribute meaning and value to it.Six participants were recruited from the same long-term child protection team to ensure homogeneity in relation to how the child's voice is framed organisationally, adhering to the principles of Interpretative Phenomenological Analysis (IPA; Smith et al., 2009).One participant had to end the interview early and this transcript was removed at analysis stage, once it was determined that the data set was sufficient.
The analysis followed the six analytic steps of IPA.After first-stage analysis of the final transcript, the first author applied the 'information power' concept (Malterud, Siersma and Guassora, 2016) to determine if the existing data set was sufficient.Considering the specificity of the sample, narrow aims of the study, strength of dialogue, and depth of analysis, it was concluded that data from the initial five transcripts was sufficient.The final sample of five participants was made up of four females and one male.The length of time within the team ranged from 2 months to 38 months and the average length of time within the team is 15 months.

Procedure and Materials
Recruitment materials such as posters and emails were shared within the long-term child protection team, and participants contacted the researcher directly to express interest in taking part.All participants were given a choice of dates, times, and location for their interview and each chose their place and time of work for convenience.Data was collected using semistructured interviews as they closely resemble naturalistic conversation (Wooffitt and Widdicombe, 2006) and allow for the collection of in-depth individual accounts (Madill, 2011).
Interviews lasting between 25 and 41 minutes took place in a private meeting room within the participants' workplace and followed a broad interview schedule.For example, one question asks, 'what happens when the child's voice is clear and understood?' with prompts including 'what does it mean to you?' and 'how do you know?'.Interviews were audio recorded for anonymised transcription.All participants were assigned pseudonyms to protect their identity.

2.4.Analytic Approach
Interpretative phenomenological analysis (IPA) was identified as the most appropriate approach for this topic as it is concerned with exploring lived experiences while giving special consideration to the individual and their cognitions (Smith et al., 2009).It is particularly useful when exploring topics that are 'complex, ambiguous, and emotionally laden' (Smith and Osborne, 2015), such as conceptualising and understanding the child's voice.IPA has its theoretical roots in phenomenology and hermeneutics, in that it recognises that researchers' observations of other people's experiences, will be understood from that researcher's position and therefore an interpretation is the best that we can offer (Larkin et al., 2008).For this study, a double hermeneutic approach was adopted, as the researcher aimed to make sense of the participants sense-making in relation to the child's voice (Smith et al.' 2009).
This inductive analysis was conducted over a number of stages: Firstly, the reading and rereading of transcripts allowed for familiarisation with the data and initial notes were made in relation to any content-, language-or concept-related observations (Smith, 1996;Smith et al., 2009).Emergent themes were developed, and then clustered into higher-order themes, for each case before moving on to identify any connections or comparisons across cases (Willig, 2008).
At this stage of the process, the researcher decided to put together 'pen portraits' for each participant using their own quotes, as a way of staying with the individuals' meanings and ensuring that the super-ordinate and sub-ordinate themes which were developing represented the participants' overall accounts.This process is in line with Smith et al's (2009) guidance which talks about the need to 'zoom-in' to and 'pan-out' from the data.From here, tables were created to represent the final super-ordinate and subordinate themes.

Reflexivity
The researcher is a white, working-class female, with no professional experience within child protection.At the time of the study, the researcher was a mature student completing a Psychology MSc, with a particular interest in person-centred and attachment focused practices.
From recruitment of participants through to the development of final themes, a reflective journal was kept in order to maintain an awareness of researcher positionality and any potential effects of this on the analysis and findings (Attia & Edge, 2017;Berger, 2013).For example, during the data collection stage, the researcher's position as a student on a psychotherapy pathway was evident following an interview with a participant who also had a background in psychology and an orientation towards therapeutic practice.Conscious not to favour this participant's viewpoint in the overall analysis and theme development, the transcript from this interview was analysed last.
Reflexivity was important when considering the researcher's position as a mother.For example, as one participant shared "to actually keep your children you only have to be good enough", it was important to explore their individual conceptualisation of what 'good enough' means and to maintain this as the frame of reference for any interpretation of this participant's transcript.To assist with this, pen portraits were created using direct quotes from each participant.This also allowed for any themes and corresponding quotes to be 'sense-checked' and reviewed by the second author.

FINDINGS AND DISCUSSION
Following multi-stage analysis, three superordinate themes (Fig. 1

THE AMBIGUOUS AND UBIQUITOUS NATURE OF THE CHILD'S VOICE
While none of the participants were asked 'what' the child's voice is, there were many attempts to define it as they shared their experiences of how it comes through in practice.Of course, to recognise it coming through, an awareness of what 'it' is would be necessary.This theme relates to the concept that the child's voice, while seemingly difficult to define, can be found

THE AMBIGUOUS AND UBIQUITOUS NATURE OF THE CHILD'S VOICE
The child's voice as everything, everywhere Hearing the voice

Ambiguous voices
Vicarious voices There is a real challenge though, in defining the child's voice: "it's kind of hard to explain" feelings and wishes of the child".However, the idea that these equally ambiguous concepts represent the child's voice, is something that is "drilled into us from day one" (Daniel: L106-107), suggesting it is more of an organisational construct than a personal one.This supports Vis and Fossum's (2015) findings that organisational culture is largely influential in terms of how the child's voice is included in practice.More recently, Dillon (2021) argues that this focus on gathering 'wishes and feelings' in child advocacy work, means that practitioners now view it as a task 'to do' rather than a meaningful form of participatory practice.

Hearing the Voice
This theme relates to the varying ways in which the child's voice may be facilitated and how can be 'heard' and understood.Jane (L52) suggests, "getting on the floor with the kids […] playing with them and exploring their environment" which gives the impression of meeting the child where they are, in their world, rather than asking them to come into ours.
Facilitating the child's voice requires "having that invested time with them".For some, its "got to feel natural" (Laura:L150), and is best achieved through everyday activities such as "just going out for a drive […] colouring or through playdoh", as more structured models feel "a bit forced" (Jane L52-57), and there's a belief that vital information would "never come out during a scheduled direct work session" (Laura:159-162).For others though, models like Three Houses offer a way to "completely validate whatever the child is saying" and can be used as "a checklist" to assess if their "needs are being met" (Daniel: L93-95).This dependence for models and checklists may support Whittaker and Harvard's (2016) findings that a fear of personal responsibility can lead to a reliance on procedural guidelines.However, Woodman et al, (2018) found that tools designed to assist in engaging children, can only increase practitioner confidence in communicating with children and that these tools are particular valuable for less experienced practitioners.
Sarah (L89-91) prefers for the child's voice to come through in observations of their "daily life, whether they go to school […] behaving at school […] engaging in anti-social behaviour" or observing their "interactions" or "bedtime routine at night" (Sarah: L114-118).This notion that the child's voice can come through from observing their relationships and routines is reiterated by Laura (L144-145) who talks of going to "school or nursery […] home" and observing "how they are in those different situations".
For many of the participants, play is a "powerful" way of facilitating the child's voice, as they describe examples such as using a "doll's house" or "police car" and noticing "the child's reaction […] relationship with, different objects" (Olivia: L159-163).This is an example of how unofficial sense-making and personal intuition are applied in practice (Saltiel, 2016) and supports the findings from O'Reilly and Dolan (2016) that the child's voice may be greater represented when it comes through in play.

Ambiguous voices
The above quote from Olivia is just one example of when the child's voice may be ambiguous, as it is safe to assume that the child's reaction to the toy car could have been interpreted differently, or even gone unnoticed, by a different social worker.This theme relates to the ambiguous nature of the child's voice, verbal and otherwise, and what it means to the participants when the child's voice isn't clear: "it's just confusing, […] makes me, generally, quite suspicious […] but what do you do with that?" (Olivia: L191-195).
Participants shared experiences of feeling that the child's voice has been 'schooled' (Daniel: L412) or 'coached' (Laura: L214), and a fear that when this happens "we almost automatically disregard the child's voice" (Daniel: L411-418).However, "knowing the child well" (Jane: L154) can enable workers to "pick up on when they might not be telling the truth" (Laura: L212-219).This appears to support the findings from Kettle (2018), in so much that a balance between closeness-distance is needed in order to recognise the authenticity of the child's voice.
) were identified as; (1) The Ambiguous and Ubiquitous Nature of the Child's Voice, (2) The Importance of Relationships, and (3) The Weight of the Voices.Each of the themes are introduced along with a discussion of the corresponding sub-ordinate themes.

Figure 1 :
Figure 1: Representation of themes and comes through in many forms, sometimes not even directly from the child themselves.3.1.1.The child's voice as everything, everywhereParticipants shared their understanding of the child's voice in terms of what it is, what it means to them and their experience of how it comes through in their practice."I think it's everything…" (Jane: L30), emphasising the personal meaning and importance attributed to it.
Vis and Thomas' (2009)fore knowing when it's coming through: "it's just the way that the child's presenting isn't it […] more than words[…]smiles or frowns".In addition, there is a sense that the child's feelings are "what we look for in younger children, that are non-verbal" (Olivia: L207-210), but not as much when words are offered.This is in slight contrast toVis and Thomas' (2009)findings that participation was more likely to be facilitated for older children.