Edinburgh Research Explorer Reimagining child welfare outcomes: Learning from Family Group Conferencing

This article reports on a qualitative study, which sought to retrospectively understand the contribution family group conferencing (FGC) makes to longer ‐ term outcomes for children at risk of entering State care and their families. Eleven case examples of FGC were studied from five local government areas across Scotland. Each example included the perspectives of different stakeholders in the process: children, their parents/carers and extended family ( n = 32), and professionals ( n = 28) involved with them. The study found a number of interconnecting issues in relation to FGC outcome contribution. First, the personal experience of process matters to the service user and his/her opinion of the outcomes they achieve. Second, what professionals do and how they do it can impact the outcomes of people requiring support. Thus, the relationship between professionals and service users is central to understanding why and how families achieve longer ‐ term outcomes. Finally, who defines outcomes and to what purpose is significant when understanding outcomes. This article reports on two sets of FGC outcomes identified within the study: personal and professional. Arguing for a more nuanced understanding of outcomes in child welfare this article begins to explore, and challenge, the manner in which outcomes are identified and valued in child welfare.


| INTRODUCTION
Over the past 20 years in Scotland and across the United Kingdom, the integration and modernisation of children's services has resulted in an outcome-led discourse in child welfare policy and practice (Davis & Smith, 2012;Frost & Stein, 2009: see also Mitchell, 2018). Canavan, Coen, Dolan, and Whyte (2009) argue that outcome measures have become a key way of defining and measuring the quality of services since the 1990s, focusing on what organizations achieved rather than what and how they did it. During that time there has been an increased requirement from social service agencies in the voluntary and public sectors to demonstrate outcome success in order to obtain funding and to meet government set targets (Burns, 2017;Richards, Tolbot, & Munro, 2015). Evaluating effectiveness is wholly desirable to assess whether services are meeting identified needs, are cost effective, and whether change is required to make improvements in the quality of life for service users.
Similar to many countries around the globe, in Scotland (the focus of this study), there have been considerable legislative and policy change in the child welfare arena, calling for the modernisation, change, and improvement of social work to address perceived inefficiencies and lack of responsiveness of the needs of those people using social services (Christie, 2011;Scottish Government, 2015. Children's services are being challenged to become more relational and find new ways to engage with children and families amongst a backdrop of financial constraint (Audit Scotland, 2016;Ofsted, 2015). A challenge for current child welfare policy and practice has been to view parents and children as partners in finding solutions, while also being part of the problem that needs fixing (Morris & Featherstone, 2010). In child welfare, a child's right to participate is often at odds with his/her right to protection, and those parents with whom partnership is required are also those identified as being in need of support, direction, and correction (Barnsdale & Walker, 2007). Families with care and protection needs are often caught in conflicting policy and practice expectations: parents are expected to take on responsibilities for care, while being positioned as failing (Featherstone, White, & Morris, 2014). This dichotomous positioning can often be at odds with child welfare outcomes discourse, impacting on the way work with children and families is approached. This article argues for an outcome focused discourse to consider both the complexities of the experiences of service users and the quality and effectiveness of the processes delivered to them. The paper begins by discussing empirical literature on FGC before discussing the findings of a qualitative study which sought to understand the contribution FGC makes on longerterm outcomes for children at risk of going into State care and their families, who have been involved in the process.

| FAMILY GROUP CONFERENCING
FGC (also known as Family Group Decision Making) is a family-led decision-making process internationally known for involving families in decisions to safeguard the welfare of the family's children. Originating in New Zealand in the 1980s, FGC sprang from debates regarding the over representation of Maori children in the welfare and justice systems. Like many western countries, New Zealand's child welfare system was seen to be underperforming, where child protection processes were too focused on risk rather than family engagement and support. As such, it was recommended that families needed to be more involved in decision-making when the State becomes involved in the private realm of the family (Marsh & Crow, 1998).
FGC in child welfare seeks to improve child safety and reduce the need for professional care and intervention (Dijkstra, Creemers, Asscher, Deković, & Stams, 2016). When there are concerns, a meeting of the child's extended family and social network is convened to involve them in making decisions and plans regarding the welfare of the child. The meeting is organized with the help of an independent FGC coordinator. The independence of the coordinator implies that he/she has no other professional role other than to facilitate the FGC service with the family (McKillop, 2016). There are several distinct stages in the FGC process after the initial referral is received.
The first stage prepares the family and professionals working with the family for the meeting. This time is used to share information with the family and give participants an opportunity to talk through the issues and think about how they might contribute to solutions. The second stage is the meeting or conference. At this start of the meeting, family members and professionals, including the social worker, re-share information and reflect on their concerns, purpose of the meeting, and the decisions required. Following information sharing, there is family-only time during which the family discuss their concerns, without professionals present, and develop a plan. The plan is then discussed and agreed between the family and professionals. The final stage of the FGC process is one of review, where those involved in the original meeting (family and professionals) consider the progress of the plan and amendments are made to the plan, if it is considered necessary.
The approach has spread worldwide since the 1980s, including to the United Kingdom (see Edwards & Parkinson, 2018, pp. 3-5 for discussion on FGC history in the United Kingdom). More recently in Scotland, the National Guidance for Part 12 of the Children and Young People (Scotland) Act 2014 states that families should be supported to make decisions prior to children being placed in care and that an example of "good practice" is Family Group Decision Making (Scottish Government, 2016, pp. 9-10). This inclusion in the Scottish Government Guidance has resulted in an interest in FGC nationally, a pattern mirrored in other parts of the United Kingdom as well (Munro, Meetoo, Quy, et al., 2017).
Research studies on FGC outcomes have been the subject of evaluation and meta-analysis, some of which have found positive effects, such as: high levels of participant satisfaction; bringing family members closer together and strengthening positive family ties; keeping children safe through the delivery of a plan which protect and safeguard children and parents/carers; improved partnership working between families and social work services; achieving more timely permanency and existing out of home care more rapidly (Pennell & Burford, 2000;Merkel-Holguin, 2003;Holland, Scourfield, O'Neill, & Pithouse, 2005;Marsh, 2013;Metze, Kwekkeboom, & Abma, 2015: see also Mitchell, 2018). Other studies have had mixed results identifying either neutral or negative effects of FGC (Berzin, 2006;Sundell & Vinnerljung, 2004).
More recently, Dijkstra et al. (2016) undertook meta-analysis, investigating the effectiveness of FGC in youth care from 17 different studies. This analysis found that FGC did not significantly reduce child maltreatment, out of home placements, and involvement in youth care. Hollinshead et al. (2017: 286) undertook a control trial study investigating a common triad of outcomes related to child welfare involvement after FGC. This study did not find a statistically significant effect on the likelihood of a referral, re-referral with substantiation, or placement in out of home care. As Fox (2018) points out, these studies had difficulties using a quantifiable statistical approach to understanding FGC, for example a change in the factors affecting the research groups, sample sizes, activities undertaken, and definitions used in research changed over time. Qualitative research literature strongly indicates evidence of programme satisfaction by family members and efficacy of FGC programmes-there is little known about longer-term outcomes for children and family members (Edwards & Parkinson, 2018;Munro et al., 2017). These studies tend to be smaller in size and focusing on smaller geographical areas.
Empirical literature regarding the study of outcomes of FGC has been critiqued for being methodologically weak (Crampton, 2007;Frost, Abram, & Burgess, 2014). Isolating the effects of FGC from the influence of other services that are typically offered alongside FGC appears difficult to assess (Crampton, 2007). Importantly, children appear to be primarily overlooked as a source of knowledge regarding FGC (Holland & Rivett, 2008). Those studies that do involve children tend to focus on programme satisfaction rather than their perceptions of outcomes (Bell & Wilson, 2006). The local variations of FGC practice raise questions of what authentic FGC practice is and how it is measured. Which child welfare outcomes FGC is expected to improve, agreement regarding the aims of the approach, defining clear and evaluative standards, and measuring associated outputs are also problematic (Barnsdale & Walker, 2007;Crampton, 2007). Randomized trials have also been operationally difficult to undertake, as one of the challenges of implementing randomized trials of FGC programmes is the inability to recruit families into the study (Crampton, 2007).
There is extensive literature relating to FGC process but few studies report on outcomes in the longer term (Frost et al., 2014;Frost & Jackson, 2018). Studies considering process have remained positive, although there have been mixed findings in relations to outcomes that is outcomes have been neutral, positive, or negative. The empirical data regarding FGC outcomes appears to be limited, somewhat dated, and is, at times, contradictory, which seem to be linked to the different methodological approaches of the studies undertaken (Frost et al., 2014;Fox, 2018: see also Mitchell, 2018). Building knowledge regarding FGC practice and outcomes is challenging for many reasons, including those related to the conceptualisation of what outcomes may mean for different stakeholders and the difficulties with outcome attribution in complex contexts (Mayne, 2008;Patton, 2012). Who defines outcomes and how they are measured have an impact on what is considered important to know about FGC (Mitchell, 2018). It is within this debate that the current study of FGC outcomes sits, aiming to qualitatively capture the contribution of FGC by considering the retrospective perspectives of those involved in FGC.

| THE RESEARCH PROJECT
The data for this paper has been drawn from a collaborative PhD study funded by the ESRC and Children 1st, a national child welfare

| Methodology
The study design involved a retrospective, qualitative case study approach. The use of qualitative methodology provided a deep understanding of personal experiences of FGC. The study sort to build a rich understanding of the FGC process and outcomes from the different perspectives of those involved in it and as such children, family members, and professionals were interviewed. FGC documents were also examined for each of the case examples, providing data that was created at the time of the FGC.
The unit of analysis under study is the family's FGC experience, where the family had a child who was at risk of going into State care.
The study sought to examine a number of examples (n = 11) of FGC who were purposefully selected. These purposeful examples might be described as "good" examples of the FGC experience. The consideration of good cases in the study builds an understanding of FGC practice from which an exploration and extrapolation can be made (Lewis, 2003;Snape & Spencer, 2003).
All families involved in the study had been referred to FGC by social work services because a child or children within the family was at risk of going into State care at the time of referral. Access was negotiated with 11 families involved in FGC across five local government authorities in Scotland. The fieldwork took place in 2014 and 2015. Given the sensitivity of the study and the involvement of children and young people in the research, a more intensive level 2 ethical review was required and approved by the University of Edinburgh.

| Data collection methods
Data was collected using two methods: in-depth semi-structured interviews and document analysis of FGC files held by either the local authority or Children 1st. This was a retrospective study, so all participants were interviewed at least 1 year after their initial FGC meeting.
Some of those interviewed chose to be interviewed at the same time, either in pairs or small family groups. In total, 60 people-children (10)

| Coding and analysis
Analysis proceeded with themes being identified from the accounts of research participants rather than from any hypothesis (Bazeley, 2013).
Initial codes were developed inductively while transcribing the interviews. Further codes were generated as the analysis progressed with the assistance of Nvivo software. Analysis was carried out on an intra case and cross case basis, with close reading and re-reading of the data (Braun & Clarke, 2013). The data was analysed thematically and identified themes were moved around and organized into a smaller number of more conceptual categories (Bazeley, 2013;Ritchie, Lewis, & Ellam, 2003).

| Limitations to the study
The researcher was aware that those individuals who agreed to participate in the study might have a more positive attitude towards their FGC experiences than those who would not have agreed to participate in the study, suggesting the possibility of a positive bias in the cases under study. As Mason (2002, p. 65) points out, it is not possible to eradicate bias in qualitative research, rather, the aim is to try and understand complexities and interactions and to "develop an understanding of how the context and situation work" rather than suggest that the key aspects of the research can be controlled. The study is small and, as such, provides a spot light on the relationship between process and identified outcomes. Further, although the researcher attempted to gain children's views, there was only a small number of children (n = 10) who agreed to be interviewed in the study, as such, data from child respondents has been incorporated into the category "children and family members." Historically, those who are supported by and involved in social work services have been described as "cases." In practice, there are more than the social work service user involved in an FGC. Individuals participating in an FGC might include: the child, members in his or her "family" and their extended social network, social worker and/or other identified professional(s), and the FGC coordinator. To help clarity in the study, each example of FGC or case describing the broader FGC grouping has been identified as an "FGC pod" (Ney, Stoltz, & Maloney, 2011). Findings discussed below focus on data from the qualitative interviews.

| FINDINGS
This study found that outcomes experienced by children and family members were different to those experienced by the professionals who are working with and supporting families involved with FGC.

| Personal outcomes (outcomes important to children and family members)
The study found that children and adult family members encounter FGC as a continuous, often emotional, experience that could influence their capacities and identity while involved not only in the FGC process, but also beyond the meeting, impacting some individual's quality of life. Building on the personal outcomes research and typology, Miller (2011) and Cook and Miller (2012) developed from empirical research on adult social care, this study found that personal outcomes-those important to children and family member involved in FGC-fell into three categories: process, learning and change, and quality of life (see Table 1 below).
Process outcomes are those children and family members credit to the FGC process. Those experiences where the children and family members felt, for example, recognized within the process as an individual with unique concerns, feeling supported to participate in the process, and understanding the perspectives of others.
In this study, many family members expressed a variety of emotions about their experience from the FGC process. Jill (all names are pseudonyms), a single mother discussing her experiences within the FGC, was one of these:

It felt good to be talking positively about Dillon (her son)
and not just seeing him as a problem. We were looking forward and that was a good thing. (Jill,mother,Pod 8) Here, Jill reports that the process allowed her to express her feelings and thoughts about her son in a different, more positive, way. The solution-focused FGC practice, reflected in her statement, assisted her and other children and family members to feel motivated about the process and more hopeful about the future.
Daryl, a father, whose two children were at risk of being permanently removed from his and his wife's, Moyra, care. Daryl has a long and strained history with social work services. For example, prior to his involvement with FGC, the police were often asked to be present at meetings with social services to help reduce the potential of conflict towards the authorities. Here, Daryl describes his experience of the FGC process, particularly the family-only time within the meeting: Children and family members interviewed in the study considered their outcomes were linked to their personal experience of FGC. Families expressed outcomes in terms of process and/or learning and/or a change in their quality of life.
The second set of outcomes identified in the study were professional outcomes.

| Professional outcomes
Professionals in the study included FGC coordinators, social workers, health workers, and a teacher (n = 28). Similar to empirical research carried out by Marsh and Walsh (2007) and Munro et al. (2017), professionals in this study tended to use the process to provide a window through which to view the family dynamic and an opportunity to offer an opinion on family situations at a given moment. Further, professionals defined outcomes in relation to their own organizational purpose, for example, whether a child was accommodated, or a meeting was organized, and if a plan was developed. As such, professional conceptualisation of outcomes in the study often differed qualitatively from those who used the service and were defined through a professional lens.
Given professionals in the study tended to conceptualize outcomes in a manner that often reflected professional or organizational requirements and goals, professional outcomes were identified in a number of categories and subsets. To begin, professional outcomes had two primary categories: practice outcomes and organizational outcomes. Practice outcomes were further divided into two subcategories: process outcomes and change and learning outcomes-see

| Practice outcomes
Practice outcomes are those outcomes that reflect the professional's interactions with and observations of the service user and family members, aiding them in undertaking their professional role with the family. These outcomes have been divided into two subcategories within the data set: process outcomes and change/learning outcomes.
Process outcomes are those where the professional reflects on how the FGC process has assisted them in working with the family members.
Outcomes identified by professionals in this subcategory can help professionals understand the FGC process and identify key professional tasks. One of those professionals was Fran, Jake's FGC coordinator: I checked that out with Jake (young person) and yes that is what he wanted. He wasn't at all happy about coming The Professionals' experience of the process assisted reflective professional practice, where, for example, a professional might observe a family supporting its members and undertaking tasks in ways that have not previously been observed. In social work practice, these are important observations and can help provide an understanding of a family's situation for assessment or intervention, while also offering evidence of theory in practice, as in this example: strengths-based approaches (Payne, 2014 Christine, a social worker in Pod 11 was one of the professionals who also discussed her FGC experience: My memory of the first meeting was really, really good.
Professionals' interview in the study considered that the FGC process enabled them to observe families working together in ways not previously experienced. Not only did this aid involvement in the decision making by both service users and professionals, but the experience also supported professionals to see a different side to family dynamics and knowledge as expressed in the evidence above.

| Organizational outcomes
Finally, organizational outcomes are those that relate to the objectives or purpose of the work being undertaken with the service user (often reflected in the original reasons for referral). They are those outcomes deemed important by the professional and the organizations within which they work in the medium to longer term. These are determined, in part, from the original reasons for the referral to a service. Although all the families involved in the study were referred to FGC services at a point where a child in the family was at risk of being placed in State care, each family also had different contextual issues and stressors relevant to individual family members, for example (but not exclusively), violence; bereavement; mental health problems; alcohol and/or drug abuse; impact of being disabled; non-attendance at school; impact of life transitions; and coming to the attention of the police within the community. Referrals in this study also identified concerns such as: poor family communication (n = 6); concerns regarding the safety of a child within the family (n = 8); family conflict which included, in some cases, domestic abuse or aggressive argument between family members (n = 10).
The professional outcomes framework (Table 2) identifies a number of organisationally-defined outcomes evident across the 11 FGC pods in the study. The organizational outcomes which FGC contributed towards include: improved child safety and, in some cases, improved safety of other members of the family; family members' increased sense of control of their lives; an increased commitment of family network and family functioning, influencing improved interfamily relations; and improved use of professional support, and in several cases, a social work no longer being involved with the family. An indepth discussion regarding these organizational outcomes can be found in Mitchell (2018, pp. 237-244).

| DISCUSSION
An outcome, simply understood, is the result of an activity or support or service which can be attributed to it (Frost & Stein, 2009). Yet, the concept is not straightforward. What an outcome is and who conceptualizes it is often contested in social work practice and literature. This study found that there are a number of challenges in relation to the use of the term "outcome." These challenges, although not explicitly linked, can, together, suggest a number of assumptions regarding outcomes that may limit the manner in which work with children and families is approached, highlighting the need to rethink how the term is used in FGC and more widely in childcare and protection. Miller (2011, p. 3) highlights that particular services may narrowly emphasis specific types of outcomes-creating a false separation of process from outcome identification. It is important to take account of what the users of services would see as desirable outcomes and the measures chosen by researchers and/or professionals may not be those that are important to children and families. Empirical research has indicated that social work service users do not separate the process and how they engage with it from the outcomes they experience (Beresford & Branfield, 2006;Frost & Stein, 2009). These findings were replicated in this study. The different experiences and perceptions of family members and professionals affect respondent's conceptualisation of outcomes. As such, both personal and professional outcomes result from respondent's involvement in FGC.
Despite the policy focus on outcomes, measuring them remains challenging. It is important to distinguish between those outcomes important to the individuals involved in the service and those outcomes defined by professionals and the service on behalf of the service user. The purpose of understanding personal outcomes, from the service user's perspective, are varied. To begin, one cannot assume that the outcomes identified by children and family members will be the same as those identified by professionals. As explored in this study, the personal experience of process matters to the service user impacting on his/her perception of the relevance of support being focuses on what makes a difference to those using the service (Miller, 2011).
The simplicity of the frameworks outlined above belies the complexity and relationships within and between the frameworks, that is, any attempt to impact one outcome may affect others; there is an interactive dependence on the activities, inputs, and changes which can lead to a change in the quality of life for service users (Mitchell, 2018). Evidence from this study suggests the interaction of the outcomes experienced by children, family members, and professionals build towards improving quality of life outcomes for children and family members. Complexity of the experience for children, family members, and professionals is evident. These frameworks offer a starting point to understand the complexity of outcome contribution for FGC and for child welfare more broadly. The application of the frameworks will require further investigation including whether children conceptualize outcomes differently from adult family members.

| CONCLUSION
This article makes a significant contribution to understanding outcomes in child welfare. The analysis presented argues for a reimagining of FGC outcomes, where those outcomes important to family members are considered alongside those conceptualized by professionals. The consideration of outcomes in this way ensures a broader, more nuanced, understanding of FGC outcome contribution, reflecting an understanding of the complexity experienced by those involved in child welfare services. It is of importance to maintain a focus on those using the service and the professionals who work with them when conceptualizing outcomes. FGC can provide a mechanism for a shared dialogue about how the different expectations of family members and professionals can be brought together and are seen as being mutually important for progress to be made. These findings have implications not only for the measuring and recording of outcomes for FGC but also more broadly in children's services as it adds to the international debates about the outcomes the child welfare system is seeking to achieve, the importance of generating evidence to help us understand what outcomes might be from different perspectives, and how we might gauge whether such outcomes are achieved (Shlonsky & Benbenishty, 2014). The recording and reporting of outcomes need to reflect the varied experiences of those involved in the service (including children) to be captured and expressed in ways that are meaningful and acceptable not only to those who commission, manage, and staff services but also to the service users too. In this way, the complexity of family's lives can be understood in the context of their own outcome identification. Effective services require the increased involvement of service users (including children), alongside professionals, in defining outcomes in child welfare.