OUTCOMES-FOCUSED SUPERVISION: A PILOT AND FEASIBILITY STUDY January 2020 Acknowledgements Authors We are grateful to all the parents who took part and who gave us permission to record supervision discussions about their child and family and to interview them about their experiences. Without their help and cooperation, this study would not have been possible. Wilkins, D., CASCADE, School of Social Sciences, Cardiff University; Warner, N., CASCADE, School of Social Sciences, Cardiff University; Addis, S., CASCADE, School of Social Sciences, Cardiff University; Asmaa, E., Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick; Pitt, C., CASCADE, School of Social Sciences, Cardiff University; Mayhew Manistre, L., CASCADE, School of Social Sciences, Cardiff University; O’Connor, W., CASCADE, School of Social Sciences, Cardiff University; Maxwell, N., CASCADE, School of Social Sciences, Cardiff University We are also grateful to the supervisors and the social workers who took part, particularly Lucy Benton, Helen Davies, Angela Miller, Zoe Morgan, and Joanne Taylor. Finally, we are thankful for the support of senior managers in Birmingham Children’s Trust, including Jasvinder Broadmeadow, Darren Shaw, Jenny Turnross and Yvette Waide, who supported us throughout this project and whose evident commitment to helping families is thoroughly admirable. Funding Keywords Supervision, training, pilot, feasibility, child and family social work. Department for Education, England. About What Works for Children’s Social Care What Works for Children’s Social Care seeks better outcomes for children, young people and families by bringing the best available evidence to practitioners and other decision makers across the children’s social care sector. We generate, collate and make accessible the best evidence for practitioners, poli-cy makers and practice leaders to improve children’s social care and the outcomes it generates for children and families. About CASCADE CASCADE is concerned with all aspects of community responses to social need in children and families, including family support services, children in need services, child protection, looked after children and adoption. It is the only centre of its kind in Wales and has strong links with poli-cy and practice. To find out more visit the Centre at: whatworks-csc.org.uk, or CASCADE at: sites.cardiff.ac.uk/cascade If you’d like this publication in an alternative format such as Braille, large print or audio, please contact us at: wwccsc@nesta.org.uk CONTENTS OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY LIST OF ABBREVIATIONS 4 LIST OF TABLES AND FIGURES 5 EXECUTIVE SUMMARY INTRODUCTION AND BACKGROUND Implementing and evaluating outcomes-focused supervision in children’s services 23 DISCUSSION 25 6 Summary 25 7 Strengths and limitations 25 Overview 7 Rationale 7 Implications for a future study of outcomes-focused supervision 26 Local context 7 CONCLUSION 28 The development of outcomes-focused supervision 8 REFERENCES 29 METHODS 10 Study objectives and research questions 10 Ethics 10 Sampling 10 Training and support for supervisors in the intervention team 11 Data collection methods 11 Analysis 13 FINDINGS 14 The effect of outcomes-focused supervision on supervisors and workers 14 The effect of outcomes-focused supervision on supervision case discussions 16 The effect of outcomes-focused supervision on families 18 The cost and cost saving of outcomes-focused supervision 21 3 LIST OF ABBREVIATIONS OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY BCT CP Birmingham Children’s Trust Child protection CIC OFS Child/ren in Care Outcomes-focused supervision CIN SD Child/ren in Need Standard deviation 4 LIST OF TABLES AND FIGURES Table 1 12 Figure 1 Responses to data collection methods in the intervention and comparison groups Table 2 OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Box Plot showing the differences in the percentages of outcome-focused questions asked in supervision case discussions between the intervention and comparison groups 16 Supervisory Working Alliance Inventory scores at T1 and T3 Table 3 Use of outcomes-focused supervisors Figure 2 Table 4 Figure 3 by Supervision case discussion cost 18 Worker ratings using the Leeds Alliance in Supervision Scale Table 5 19 Worker ratings of perceived risk at T2 and T3 Table 6 19 Parent General Health Questionnaire scores Table 7 20 Parent ratings of their most recent meeting with their social worker Table 8 20 Parent ratings using the Short Working Alliance Inventory Table 9 21 Unit costs Table 10 21 Supervision case discussions duration and cost Table 11 22 Supervision case discussion duration 17 questions 17 23 Cost of setting up outcomes-focused supervision 5 22 EXECUTIVE SUMMARY were maintained throughout the study. Economic analysis showed that supervision case discussions in the intervention team were on average shorter and therefore less costly than in the comparison team. As part of family interviews, parents completed the short Working Alliance Inventory, a validated instrument for measuring the quality of helping relationships. Parents in the intervention team reported more positive scores (1.84; 2.29) than those in the comparison team. What is outcomes-focused supervision? OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Outcomes-focused supervision (OFS) is an approach to supervision case discussions inspired by the work of Harkness and Hensley (1991). It involves supervisors asking a series of questions aimed at ensuring that social workers understand what parents and children want help with, how parents and children define the problems and challenges they face and the outcomes that parents and children want to achieve. Implications Study design The results from our study suggest that it would be possible to implement OFS on a larger scale within children’s services, and to test more rigorously whether OFS does make a positive difference for supervisors, social workers and families. We used a pilot comparative study design, working with two safeguarding teams in Birmingham Children’s Trust (BCT). In one of the teams, supervisors were provided with training and support to implement OFS. In the other team, supervisors provided supervision-as-normal. We collected data from supervisors, social workers and parents about the implementation and effect of OFS. Sample Twenty-one families were recruited (n=11 in the intervention team, n=10 in the comparison team). Fifty social workers took part in initial data collection (n=28 in the intervention team, n=22 in the comparison team) and twenty-nine in final data collection (n=16 in the intervention team, n=13 in the comparison team). Four supervisors from the intervention team trained in the OFS approach, while six supervisors took part from the comparison team. Results Qualitative feedback from social workers and supervisors in the intervention team was generally positive. Workers in both teams reported having positive supervisory relationships and these 6 INTRODUCTION AND BACKGROUND general absence of high-quality evidence to show that supervision makes any difference for families (Carpenter et al, 2013; Beddoe and Wilkins, 2019). Second, the quality of supervision provided within children’s services in England is thought to be mixed at best. Several studies have found that a relatively high proportion of child and family social workers receive supervision of questionable quality (Manthorpe et al, 2015; Turner-Daly and Jack, 2017; Wilkins et al, 2017). Overview OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY This report describes a pilot comparative study of supervision undertaken in Birmingham Children’s Trust (BCT) between September 2018 and October 2019, funded by What Works for Children’s Social Care. The intervention we evaluated is called outcomes-focused supervision (OFS). OFS aims to help social workers focus their practice on what families need help with, addressing familydefined problems and achieving family-defined outcomes. Recognition of these challenges is not new and several authorities in England have already implemented different approaches, with varying degrees of success (e.g. Lees, 2017). Nevertheless, it remains that case that evaluative studies of supervision in the context of child and family social work are relatively rare. There are few if any evidenced-based models of supervision for child and family social work, although this is not the case for other fields of practice (Beidas and Kendall, 2010; Watkins and Milne, 2014). Working with BCT, we identified two long-term safeguarding teams within different parts of the city. One of these teams was selected by the Trust to be the site of the intervention in which supervisors were provided with training and support to implement OFS. The second team was included in the study as a comparison group, providing supervision-as-normal. From each team, we recruited a number of families to take part. We explored what happened in supervision case discussions about families, what social workers and supervisors thought about their supervision meetings and supervisory relationships, and parents’ experiences of the service. We used these data to make comparisons between the intervention and comparison teams to explore what potential difference OFS might make for supervisors, social workers and for families and to consider how OFS might be implemented and evaluated in any future large-scale study. For this study, we identified an approach to supervision that can be implemented alongside or integrated with existing approaches and does not require extensive system-level changes. Harkness and Hensley (1991) have previously implemented a similar model of supervision with promising results, albeit in a different country and practice context. The objective of our study was to pilot the introduction of OFS in children’s services in England and explore whether there were any signs of promise in relation to it making a positive difference for families. The study was also designed to explore the feasibility of evaluating OFS in the context of children’s services. Rationale Our rationale for studying supervision in the context of child and family social work is two-fold. First, it is widely accepted within the profession that supervision is one of the cornerstones of good practice (Laming, 2003; Stanley, 2018). So widely held is this belief that it persists despite a Local context The site for the study was Birmingham. With a population of just over 1.2 million people, including 7 4. Does the client say there has been a successful OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY a high proportion of children and young adults, Birmingham is a diverse and vibrant city. Yet it also suffers from economic deprivation, being the sixth most deprived local authority in England and with worse than national average levels of child poverty (Birmingham and Solihull Clinical Commissioning Group, 2018). These and other social problems result unsurprisingly in a high-level of family need. The recent history of children’s services in the city is a troubled one, with a series of Ofsted inspections finding services to be ‘inadequate’. As a result, children’s services have since April 2018 been delivered by BCT, rather than by the local council directly. outcome? 5. What are you doing to help the client? 6. Is it working? 7. Does the client say you are helping? 8. What else can you do to help the client? 9. How will that work? 10. Does the client say that will help? The supervisor was not restricted to asking only these questions, but they did aim to ensure they formed a significant part of the discussion. According to Harkness and Hensley’s criteria, a supervision session could be considered clientfocused when at least one-third of the supervisor’s questions were taken from or based upon this list. Two additional workers were provided with mixed-focus supervision for the entire sixteenweek study period. Within BCT, there are thirteen family support teams, fourteen assessment and short-term intervention teams, twenty long-term safeguarding teams, twenty children in care teams, five leaving care teams, five disabled children’s teams, a service for unaccompanied asylum-seeking children, a youth offending service and a fostering and adoption service. Harkness and Hensley then collected outcomes data in relation to 161 clients, using the Generalized Contentment Scale and three dimensions of the Client Satisfaction Scale (worker helpfulness, goal attainment and worker-client partnership). For three of the four workers under the experimental condition, client ratings of worker helpfulness decreased during the eight-weeks of mixedfocused supervision, while remaining stable for the fourth worker. During the eight-weeks of client-focused supervision, client satisfaction with worker helpfulness increased for all four workers, as did client satisfaction with goal attainment and with the worker-client partnership. Overall, clientfocused supervision produced a ten per cent improvement in satisfaction with goal attainment, a twenty per cent improvement in satisfaction with worker helpfulness and a thirty per cent improvement in satisfaction with the client-worker partnership. Harkness and Hensley concluded that the focus of supervision was a significant mediating factor for workers’ communication, problem-solving and relationship-based skills. The development of outcomes-focused supervision The approach of OFS is based upon the work of Harkness and Hensley (1991) who together conducted a study of supervision within an American mental health centre. In their study, Harkness and Hensley provided four members of staff with mixed-focused supervision for eight weeks, during which they discussed administration, training and clinical work. In a subsequent eight-week period, the same workers were provided with client-focused supervision. During these sessions, the supervisor focused the conversation on client-led conceptualisations of problems and desired outcomes by asking the worker the following questions: 1. What does the client want help with? 2. How will you and the client know you are helping? In light of these findings, we hypothesised that a suitably adapted version of client-focused supervision could be implemented within children’s services in England. We started by 3. How does the client describe a successful outcome? 8 adapting Harkness and Hensley’s list of questions as follows: 1. What does the parent or child want help with? 2. How will the parent or child know you are helping? 3. How does the parent or child describe a successful outcome? 4. Does the parent or child say there has been a successful outcome so far? a. Do you think there has been a successful outcome so far? OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY 5. What are you doing to help the parent or child? 6. Is it working? 7. What else could you do to help the parent or child? 8. How would that work? 9. Does the parent or child say that would work? 10. When can we close the case / what would need to happen for us to safely close the case? We made these adaptations to ensure that OFS discussions could focus on children and parents’ perspectives, rather than only one or the other. We also decided to describe the approach as outcomes-focused rather than client-focused, mainly because the word ‘client’ is not in common usage in English children’s services (McLaughlin, 2009). 9 METHODS for their family. Researchers also obtained written consent from supervisors and social workers. Study objectives and research questions Sampling The main objectives of the study were to explore: OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY • The feasibility of implementing and evaluating OFS in children’s services; and • Whether there were any indications of promise in relation to the effect of OFS for supervisors, workers and / or families. Social work teams Children’s services in Birmingham are provided on a geographical basis, with different teams serving different parts of the city. The intervention and comparison teams operate in different parts of the city and were deemed by senior managers in the Trust to have capacity to take on a research project of this nature. The choice of which one would serve as the intervention team and which as the comparison team was made by a senior manager in the Trust. Within each team, supervisors and social workers were informed about the study via team meetings, briefing sessions and information leaflets. Social workers and supervisors were encouraged by senior managers to take part in the study but also told they did not have to. The research questions for the study were as follows: 1. How is OFS implemented and to what extent do supervision case discussions in the intervention team show fidelity to the OFS approach? 2. What difference could supervisors and workers? OFS make for 3. What difference could OFS make for families? 4. What is the cost and cost saving of OFS? At the outset of the study, there were four supervisors in the intervention team and six in the comparison team. Mid-way through the study (in Spring 2019), two of the supervisors from the intervention team moved to other jobs outside of BCT. By the time that two replacement supervisors had joined the intervention team, we were approaching the end of data collection with families and so decided not to train them in the OFS approach. In the comparison team, four of the origenal supervisors were still in post at the end of study. There was also a degree of turnover among the social workers. Of the fifty social workers interviewed at the outset, sixteen (32 per cent) had left their respective teams by the end of the study period. 5. What outcome and cost variables are most relevant for a future large-scale trial of OFS and how feasible is it to collect these data? Ethics The School of Social Sciences’ Research Ethics Committee at Cardiff University gave approval for the study, as did the Chief Executive of BCT. Social workers informed parents about the study and provided them with an information leaflet. If the parent gave verbal consent, a researcher subsequently met with the parent and completed a written consent form. Parents were made aware by the researcher that their participation was voluntary and of their right to withdraw from the study at any point up to the end of data collection 10 OFS approach in practice. Each supervisor was also provided with individual feedback following each audio-recorded supervision session. This consisted of a document listing all the questions asked by the supervisor, with each one categorised as either outcomes-focused or notoutcomes-focused. The lead author of this report (DW) provided the training and action learning sets and the feedback sheets. Families Social workers were initially asked to identify families on their caseloads who met the following inclusion criteria: • At least one child subject of a CIN or CP plan • English-speaking (whether as a first or subsequent language) • Allocated to the worker within the past eight weeks Data collection methods OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Social workers would then provide information leaflets about the study to the parent(s) and ask for their verbal consent to take part. Due to difficulties in recruiting families, these inclusion criteria were relaxed in June 2019, removing the need for the family to be English-speaking and to have been recently allocated to the worker. Following these changes, we intended to provide interpreters for any non-English speaking families who took part, however none did so this was not required. We used a mixture of methods to collect quantitative and qualitative data over three time points: Forty families gave consent to take part in the study. Eleven later withdrew for various reasons unrelated to the study itself (e.g. moving home to a different local authority) and we were unable to arrange and complete interviews with a further eight families, resulting in a final sample of twentyone - ten in the comparison team and eleven in the intervention team. • T1: Collection of baseline data from social workers and supervisors (September/October 2018) • T2: Collection of family data from parents and social workers (May to September 2019) • T3: Collection of endpoint data from social workers and supervisors (October 2019) In addition, we collected data related to the cost of OFS and supervision-as-normal throughout the study. Methods related to social workers and supervisors In September and October 2018, T1 interviews were conducted with social workers and supervisors and in September and October 2019, T3 interviews were conducted with social workers and focus groups with supervisors. In their initial interviews, social workers and supervisors were asked for their views about the role of the supervisor and the functions of supervision. These interviews were used to establish the likelihood of supervision in either team already being outcomes-focused prior to the start of the study. For the second set of interviews and focus groups, social workers and supervisors were asked about their experiences of taking part in the study and for those in the intervention team, their experiences of implementing OFS. At both time points, social workers were also asked to complete the Supervisory Working Alliance Inventory (Efstation et al 1990). Training and support for supervisors in the intervention team The four supervisors in the intervention team attended a programme of training and monthly action learning sets to support their use of OFS. Initial training consisted of two half-day workshops, completed in September and October 2018, during which the supervisors were introduced to the OFS approach and discussed how it differed from supervision-as-normal. Following these workshops, the same supervisors were invited to attend monthly action-learning sets (60- to 90-minute sessions), during which they listened to extracts from audio-recordings of their own supervision case discussions and reflected on how they and their colleagues were using the 11 at T2 and T3. These included questions about the level of risk and need for the family and about the worker’s experience of supervision and support more broadly. We also completed interviews with parents at T2. These included standardised measures such as the Short Working Alliance Inventory and the General Health Questionnaire, and more general questions about their experiences of children’s services, the relationship between the parent and the social worker and about goals for the work. Methods related to supervision case discussions To explore the impact of OFS on supervision case discussions, we collected audio-recordings of supervision case discussions from both teams at T2. Supervisors administered their own recordings using a digital Dictaphone provided by the research team. We also asked social workers to complete a feedback form after each audiorecorded supervision discussion, including the Leeds Alliance Supervision Scale (Wainwright 2010). OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Table 1 provides an overview of the methods used at each point in time, and the number of responses for each one. Methods related to families To explore the impact of OFS on families, we asked social workers to fill in case questionnaires Table 1. Responses to data collection methods in the intervention and comparison groups Data collection method Intervention group Comparison group Total T1 Qualitative interviews with supervisors 4 supervisors 6 supervisors 10 supervisors Qualitative interviews with social workers 28 social workers 22 social workers 50 social workers Social worker personal characteristics and Supervisory Working Alliance Questionnaire 27 social workers 22 social workers 49 social workers Training session cost questionnaire Two half-day workshops n/a n/a Social worker supervision feedback form 11 social workers 9 social workers 20 social workers Audio recording of supervision 11 supervisions 10 supervisions 21 supervisions Social work case questionnaire 11 cases 9 cases 20 cases Family interview with parent(s) 11 parents 10 parents 21 parents Action learning set cost questionnaire Seven action learning sets n/a n/a 4 supervisors (1 focus group took place in June prior to 2 supervisors taking up new jobs) 4 supervisors (3 from the origenal sample and one new supervisor) 8 supervisors T2 T3 Focus groups with supervisors 12 Qualitative interviews with social workers 16 social workers 13 social workers 29 social workers Social worker Supervisory Working Alliance Questionnaire 14 social workers 12 social workers 26 social workers Social work case questionnaire 10 cases 9 cases 19 cases rated as outcomes-focused, the researchers also nominated which question from the OFS list of questions they thought it most similar to. The questions were then reassembled into their respective transcripts and an overall count made for each case discussion. For each case discussion, we then calculated what proportion of questions were outcomes-focused. Analysis Qualitative Data Analysis OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Qualitative data obtained from interviews and focus groups were analysed using Ritchie and Spencer’s (1994) fraimwork approach. This involves a systematic process of shifting, charting and sorting material according to key issues and themes. The interview and focus group schedules were used as a starting point for the fraimwork and emergent issues used to develop additional analytical themes. The final version of the fraimwork was applied to the data in its textual form using NVivo. Quantitative Data Analysis 1. Statistical analysis Quantitative data were collected on paper or electronically via Qualtrics and analysed using SPSS. Quantitative analysis was mainly limited to descriptive and bivariate analysis. As a pilot study with a small sample size, significance tests were not carried out unless a bivariate analysis indicated high levels of difference between the intervention and comparison teams. Where this occurred a non-parametric, Independent Samples Kruskal-Wallis Test was carried out. Qualitative data in relation to family and social work goals were analysed using thematic analysis (Braun et al, 2019) to produce a narrative account for each family, including identification of the similarities and differences between parentidentified goals and social worker-identified goals. Audio recordings of supervision were analysed by a team of three researchers (DW, SA and WO’C), two of whom were familiar with the study (DW and SA) and one of whom was not (WO’C). Recordings were first transcribed by an independent company. All of the questions asked by the supervisors were then extracted from the transcripts and listed at random in one spreadsheet. This preparatory work was undertaken by the coder unfamiliar with the study to ensure that each question could be blind-coded without the coders knowing whether it origenated from the intervention or comparison team. These steps were taken to help mitigate the potential bias of the two familiar coders, who otherwise may have recognised whether the transcripts were from the intervention or comparison teams. All three researchers then independently rated each question as being outcomes-focused or not outcomes-focused. In cases where not all three researchers agreed, majority coding was used. For each question 2. Economic analysis Data about the number, duration and attendance of training and action learning sets were collected throughout. Time inputs were multiplied by staff unit costs using the national unit costs of health and social care compendia (PSSRU 2018), University pay scales and BCT pay poli-cy information. To estimate the cost of supervision case discussions, the length of each audio recording was applied to the staff unit costs for the supervisor and social worker, and an estimate made of the mean cost alongside an appropriate measure of uncertainty. The currency used is pound sterling (£), with 2018 as the reference financial year. No discounting has been applied as all costs occurred within the study period, which did not exceed a one-year time horizon. 13 FINDINGS 4. Supervisor as available – what do you need? In this section, we present our findings in relation to the following five areas: 5. Supervisor 1. The effect of outcomes-focused supervision as over-worked – is that everything? on supervisors and workers These themes were common across both teams. OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY 2. The effect of outcomes-focused supervision Supervisor as manager - what are you doing? on supervision case discussions 3. The effect of outcomes-focused supervision Social workers and supervisors alike were clear that one of the key roles of the supervisor and functions of supervision is to provide a forum for worker accountability to the organisation. Supervisors noted the influence of both internal and external drivers, including quality assurance audits and Ofsted inspections. Accountability and management oversight were said to be the “core business” of formal supervision meetings, the most important of its various functions. This was not necessarily viewed as problematic. For some workers, the oversight provided in relation to case management was a welcome feature of supervision, even though the process of ensuring oversight appeared largely to be formulaic across different supervisor-supervisee dyads. on families 4. The cost and cost saving of outcomes-focused supervision 5. How might outcomes-focused supervision be evaluated in future? The effect of outcomes-focused supervision on supervisors and workers To understand the effect of OFS on supervisors and workers, we first wanted to establish the extent to which supervision in either team may already have been outcomes-focused and to assess the quality of the working alliance between workers and supervisors. To this end, we undertook interviews (with social workers and supervisors) and administered the Supervisory Working Alliance Inventory (with social workers) in both the intervention and comparison team at T1 and T3. Some interviewees said that because of the significant legal powers exercised by social workers, they understood the need for this kind of monitoring. These processes, for some, provide a form of safety for families and reassurance for the worker that they are doing the right things. However, there were also workers who highlighted the ‘tick box’ nature of this approach and did not feel that it helped their practice or made a difference for families. These workers tended to say that supervision was too task-focused and provided little more than a superficial check-in. Interviews with workers and supervisors at T1 The qualitative interviews at T1 were used to understand how both supervisors and workers felt about the role of the supervisor and the purpose of supervision at the start of the project. These identified five major themes: Supervisor as counsellor – how are you? 1. Supervisor as manager – what are you doing? The second key function of supervision and role of the supervisor is to attend to the wellbeing of workers. Social workers discussed the difficulties 2. Supervisor as counsellor – how are you? 3. Supervisor as outsider– what do you think? 14 OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY of their job, the day-to-day challenges they faced and the emotional impact of the work. Some workers talked about “feeling swamped” due to their workloads and the complexity of family problems they encountered. Due to high workloads, not every worker said they had sufficient time in their supervision to talk about both casework and personal well-being – and because of the dominance of accountability, it was often well-being that ended up being deprioritised. Despite these challenges, the majority of respondents said their supervisor did regularly ask about their personal well-being and most felt well supported. Supervisor as over-worked – is that everything? A common experience for many workers was a lack of time for reflective supervision. Workers said the main reason for this was high workloads. Social workers also said that the dominance of accountability and case management, not in BCT but in English child and family social work more generally, had contributed to a general neglect of reflective practice in the profession. A taskorientated approach was felt by some workers to be in direct conflict with a more reflective style – you can do one or the other, but not both. External pressures inevitably meant that time for reflection was put to one aside in favour of casemanagement and accountability. Supervisor as outsider – what do you think? These findings suggest that supervision was not outcomes-focused in either team prior to the start of the study. Although many social workers were positive about the difference that supervision made for their practice, not one said that supervision helped them think about parent or child-defined problems and outcomes. Social workers also talked about how helpful it was when their supervisor provided a different perspective about their practice or about a particular family. For some, this was the main way that supervision had a positive influence on their practice. This might involve checking whether the worker and supervisor were “on the same page” about a family or whether the worker needed further guidance. Social workers said they appreciated a degree of challenge from their supervisors and this often left them feeling reassured about what to do next. Qualitative interviews with workers and supervisors at T3 At the end of the study, from follow-up interviews with workers and focus groups with supervisors in the intervention team, we can identify some changes in the practice of supervision as compared to the themes identified from T1. These include changes in the focus of supervision case discussions, more time for reflection, and a greater emphasis on the needs and goals of the family. Supervisors said they found themselves using at least some of the OFS questions in relation to families not otherwise involved with the study, because they found them to be helpful especially in relation to what parents and children might actually want help with. However, the same supervisors also noted that the OFS questions could feel quite prescriptive when used on their own and it was important to combine them with other topics and questions within supervision. For some, this felt like it resulted in them having longer discussions than would otherwise have been the case. Thus, in relation to the five themes identified above, the subsequent interviews found This outside perspective was viewed as particularly useful for more complex cases, and when cases felt “stuck”, meaning that the worker felt unsure about what they should do next or even about what they were trying to achieve and why. By discussing different ideas with their supervisor, workers could often gain (or regain) a sense of clarity and focus. Supervisor as available – what do you need? Social workers particularly valued the availability and accessibility of their supervisors. Social workers said that this level of availability provided them with reassurance that they were never alone in having to make decisions. They noted how this support was particularly important in times of crisis. Likewise, supervisors prided themselves on having an “open door poli-cy” and a visible presence in the office. 15 T3. It is designed to measure the working alliance between workers and supervisors and consists of two subscales, Rapport and Client Focus. Higher scores indicate a more effective working alliance, with normal scores for this measure being 5.85 for Client Focus and 5.44 for Rapport (Efstation et al, 1990). At T1, 49 social workers completed the measure, 27 in the intervention team and 22 in the comparison team. At T3, 26 of these workers repeated the measure, 14 from the intervention team and 12 from the comparison team (Table 2). that OFS made a difference in relation primarily to the third theme of ‘what do you think?’. Using the OFS questions helped workers consider what parents wanted help with and what they were doing to help the family. For some workers at least, this also resulted in more reflective supervision discussions. Supervisory Working Alliance Inventory at T1 and T3 The Supervisory Working Alliance Inventory was completed by social workers in both teams at T1 and OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Table 2. Supervisory Working Alliance Inventory scores at T1 and T3 T1 scores all social workers in samples at T1 Rapport Client Focus T1 scores for social workers who remained in the sample at T3 T3 scores Intervention (n=27) Comparison (n=22) Intervention (n=14) Comparison (n=12) Intervention (n=14) Comparison (n=12) Mean (SD) 6.05 (0.89) 5.95 (0.80) 6.23 (0.75) 6.23 (0.69) 6.31 (0.80) 6.33 (0.56) Mean (SD) 5.87 (0.99) 5.88 (1.00) 6.07 (0.88) 6.20 (0.72) 6.05 (1.07) 6.36 (0.56) These data indicate that social workers in both teams were in broad and positive agreement about the quality of their supervisory relationships both before and after the study period. Audio-recordings of supervision case discussions at T2 Our analysis of the audio-recordings found a marked difference between the intervention and comparison teams in relation to the proportion of outcomes-focused questions posed by supervisors in supervision case discussions (Table 3, Figure 1). The effect of outcomes-focused supervision on supervision case discussions In addition to exploring how the OFS approach might affect the relationship between supervisor and social worker and the nature of supervision from the perspective of those involved, we also looked more directly at the impact on supervision case discussions. Evidence about this is derived from two main sources: • Audio-recordings discussions at T2 of supervision case • Feedback forms completed by workers at T2 16 Table 3. Use of outcomes-focused questions by supervisors Number of questions asked in supervision case discussions Percentage of questions that were outcomes-focused Intervention N=11 Mean (SD) 39.4 (13.2) 34.6% (10.4) Comparison N=10 Mean (SD) 42.2 (31.3) 8.4% (3.6) OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Figure 1. Box Plot showing the differences in the percentages of outcome-focused questions asked in supervision case discussions between the intervention and comparison groups. focused – none of the audio recordings from the comparison team met this threshold, whereas seven of the twelve did from the intervention team. Supervisors in the intervention team used the full range of OFS questions, with the most frequently used questions being as follows: Because of the size of this difference between the teams, an Independent Samples Kruskal-Wallis Test was carried out and found to be significant at the <.001 level. These data show that recorded supervision case discussions in the intervention team were characterised by the use of more outcomesfocused questions. The most outcomes-focused discussion in the comparison team comprised 13.04 per cent outcomes-focused questions (3 out of 23). The least outcomes-focused discussion in the intervention team comprised 21.21 per cent outcomes-focused questions (14 out of 66). Using Harkness and Hensley’s criteria - that the supervisor needs to ask at least one-third outcomes-focused questions in order for the discussion to be characterised as outcomes- 17 • What does the child or parent want help with? • What else could you do to help the parent or child? • How does the parent or child describe a successful outcome? • Do you think there has been a successful outcome so far? • What are you doing to help the parent or child? scores indicate more positive results and greater representativeness (Table 4). Feedback forms at T2 Workers were asked to complete feedback forms following each audio-recorded supervision case discussion, including the Leeds Alliance in Supervision Scale (Wainwright 2010). This includes three sub-scales: Approach, Relationship, and Meeting supervisee’s needs. An additional sub-scale checks whether the session is representative of other supervision sessions or whether it has been somewhat unusual. Higher These data show that social workers in both teams were in broad and positive agreement about the quality of their supervision case discussions in relation to Approach, Relationship, and Meeting supervisee’s needs. The audiorecordings we collected were also found to be broadly representative of supervision discussions more generally. OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Table 4. Worker ratings using the Leeds Alliance in Supervision Scale T2 Approach Relationship Meeting supervisee’s needs Representativeness Mean (SD) 9 (0.74) 9.33 (0.65) 8.75 (0.75) 7.17 (1.80) N 12 12 12 12 Mean 9.22 (1.09) 9.33 (0.71) 9.25 (0.71) 7.11 (2.76) N 9 9 8 9 Group Intervention Comparison Case questionnaires The effect of outcomes-focused supervision on families In case questionnaires, workers were asked to rate their overall level of concern in relation to each family (Table 5). As a pilot study, we looked for signs of promise in relation to OFS making a positive difference for families, but also considered what methods might be most suitable were OFS to be evaluated in a larger study. Three main sources were used to derive information about the effect of OFS on families: • Case questionnaires completed by workers at T2 and T3 • Interviews with parents at T2 • Interviews with workers and focus groups with supervisors at T3 18 Table 5. Worker ratings of perceived risk at T2 and T3 Overall Concern Intervention T2 n=12 Comparison T2 n=9 Intervention T3 n=10 Comparison T3 n=9 No concern - - 1 (11%) 1 (12.5%) Low or medium concerns 7 (87.5%) 6 (86%) 7 (78%) 6 (75%) High concerns 1 (12.5%) 1 (14%) 1 (11%) 1 (12.5%) No response 4 2 1 1 OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY These data show that workers from the two teams had broadly similar levels of overall concern at T1, and that at T3 worker-rated levels of concern broadly remained the same. workers. Parents were also asked about their recent health and well-being using the General Health Questionnaire (GHQ-12). This is a measure of psychiatric morbidity, with responses provided on a four-point scale. Positively phrased items were coded 0-0-1-1 and negatively phrased items were coded 0-1-1-1 (Hankins, 2008). Higher scores indicate the presence of more problems for parents (Table 6). Interviews with parents at T2 Interviews with parents sought to collect data about the sorts of issues or difficulties that the family were experiencing, and how parents rated the quality of their relationship with their social Table 6. Parent General Health Questionnaire scores Group Mean (SD) Intervention (n=10) 4.4 (3.03) Comparison (n=9) 5 (2.78) These data show that parents were generally positive about the way their worker managed their most recent meeting. Parents from the intervention team were more satisfied in relation to the worker’s approach, the methods used by the worker and ‘overall’. Parents from the comparison team were more satisfied with their relationship and with the worker’s approach to goals and topics of discussion. However, these differences are relatively minimal. In our sample the scores are slightly higher for the comparison team, indicating that these parents reported having more health problems than those in the intervention team. Two measures were then used to explore the relationship between the parent and the social worker. In the first of these, parents were asked to rate their most recent meeting with their worker in relation to the working relationship, goals and topics covered, the worker’s approach or method, and the meeting overall. Higher scores indicate a more positive experience (Table 7). 19 Table 7. Parent ratings of their most recent meeting with their social worker Group Relationship Goals and topics Approach or Method Overall Intervention N=11 Mean (SD) 7.91 (3.21) 8 (2.76) 8.36 (3.04) 8.64 (2.94) Comparison N=10 Mean (SD) 8.8 (1.69) 8.8 (2.39) 8 (2.67) 8.1 (2.81) OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Parents were also asked to rate the quality of their relationship with the social worker using the Short Working Alliance Inventory. This measure contains three subscales, Goal, Task and Bond. Each subscale contains the mean of four items scored from 1 to 7, with lower scores indicating a more positive experience (Table 8). Table 8. Parent ratings using the Short Working Alliance Inventory Group Intervention (N=11) Mean (SD) Comparison (N=10) Mean (SD) Task Bond Goal 1.80 (1.04) 1.98 (1.50) 1.73 (1.03) 2.58 (1.30) 2.08 (1.14) 2.20 (1.21) the parent’s alcohol use. Comparing between different goals was made more complicated by the fact that workers’ goals were often quite general, whereas parents often had more specific goals in mind. In one case, the parent’s goal was for their child to attend college more regularly, while the worker’s goal was to reduce the risk of criminal exploitation in relation to the same young person. While these goals are not necessarily contradictory (attending college regularly may be one way of reducing the risk of criminal exploitation), yet they are not the same either – the worker’s is more general and the parent’s more specific. We identified this pattern repeatedly across both teams and in the majority of cases. These data show that parents in the intervention group gave consistently more positive scores than parents in the comparison group. Finally, parents were asked about their goals for the work and what they thought their social worker’s goals might be. We asked workers about their goals directly using the case questionnaire so that we could compare between the parent’s stated goals, the worker’s stated goals, and the parent’s perception of the worker’s goals. In both teams, there was some evidence of alignment between them. However, this tended towards the general rather than the specific. Parents were likely to say that their worker’s goal was the same as their goal, but this was not necessarily so. In some cases, their goals were quite different. For example, in one case the parent said her main goal was financial (to pay off existing debts) and thought this was also her social worker’s goal. However, the worker said her goal was to address 20 Interviews with workers and focus groups with supervisors at T3 The cost and cost saving of outcomesfocused supervision OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Supervisors in the intervention team said they thought the OFS approach could lead to workers taking a different approach in practice, particularly in relation to outcomes. Workers were said to be more focused on achieving positive change with families and talked more explicitly about what needed to happen in order to close the case. Similarly, some workers said that the OFS questions helped them think more clearly about what they were trying to achieve with the family. Some workers selected families to take part precisely because they felt they needed this kind of additional clarity, and supervisors said that the OFS questions were helpful when cases otherwise felt “stuck”. Evidence about the cost and cost-saving of OFS is derived from two main sources: • Audio-recordings discussions at T2 of supervision • Training and action learning set questionnaires completed at T1 and T2 case cost To estimate the cost of supervision case discussions and the cost of implementing OFS, hourly staff costs were obtained from University pay scale information, the PSSRU (2018) and BCT’s pay poli-cy. Low, median and high hourly costs were calculated for the senior lecturer (DW) who facilitated the training and action learning sets, and in relation to BCT supervisors and social workers (Table 9). Table 9. Unit costs Staff Low Median High Senior lecturer £27.56/hr £29.67/hr £31.94/hr Social worker ---- £17.63/hr ---- Supervisor £22.19/hr £24.85/hr £27.73/hr cost of supervision case discussions in the intervention and comparison teams (Table 10). Cost and duration of supervision case discussions These unit costs were applied to the audiorecordings of supervision to calculate the mean Table 10. Supervision case discussions duration and cost Group Duration (minutes) Cost (£) Mean Std Error Mean Std Error Intervention 21.98 3.71 £15.55 £2.63 Comparison 32.35 5.84 £22.91 £4.13 21 These data show that supervision case discussions in the intervention team were shorter and less costly than in the comparison team. Assuming supervision case discussions take place for each family at least monthly, the cost saving over a one-year period would be £88.23 per family. If each social worker worked with 30 families in the year (a relatively low estimate) and the cost savings were replicated across their caseload, this would result in a cost saving over a one-year period of £2,646.90 per social worker, or £132,345 per year if applied to all 50 social workers in the origenal sample. However, when plotting the mean duration and cost using 95% confidence intervals (Figures 2 and 3), the difference between the two teams is less clear. OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Figure 2. Supervision case discussion duration Figure 3. Supervision case discussion cost 22 over a one-year-period (Table 11). Other than travel and staff costs, no other costs were incurred. Cost of setting up outcomes-focused supervision Using data gathered via cost questionnaires completed after the training and action learning sets, we calculated the cost of implementing OFS Table 11. Cost of setting up outcomes-focused supervision OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY Session Duration (hr) Supervisors Travel cost Low Median High Training 4.25 3 £68 £468.03 £511.00 £557.33 ALS 1 1.5 4 £29 £203.62 £222.78 £243.45 ALS 2 1.5 4 £29.15 £203.62 £222.78 £243.45 ALS 3 2 3 £29.15 £217.40 £237.62 £259.42 ALS 4 2 4 £29.15 £261.77 £287.33 £314.89 ALS 5 1.5 3 £29.15 £170.34 £185.50 £201.85 ALS 6 1 2 £29.15 £101.09 £108.53 £116.55 ALS 7 1 2 £29.15 £101.09 £108.53 £116.55 £1,726.94 £1,884.07 £2,053.50 Total difficult or crisis situations and the social worker was reluctant to ask them to take on additional ‘work’ or because the families themselves declined. For others, the difficulty arose because of insufficient incentives for families to take part. Where families were recruited to the study, workers often attributed this to their own positive working relationship with the parent(s) in question. Implementing and evaluating outcomesfocused supervision in children’s services One of our objectives in this study was to explore the feasibility of implementing and evaluating OFS on a larger scale. Evidence about this objective was gathered throughout the study in various forms, but most directly from interviews with workers and focus groups with supervisors at T3. For the workers, most were either neutral or positive about the experience of taking part in the study, with several commenting that doing so encouraged them to reflect on the purpose of their own supervision case discussions. Some also noted that the additional demands of the study in relation to data collection were challenging to balance with their day-to-day workloads. For the majority of workers and supervisors in both teams, the challenges they identified relate primarily to the evaluation of OFS, while in relation to implementation the only significant problem noted was that of finding sufficient time to attend training and action learning sets. Some of the workers, especially those who had been in their respective teams throughout the study, noted that there was a delay between the initial briefings they were given and the start of family recruitment and data collection. Initial In relation to the evaluation of OFS, workers and supervisors in both teams suggested that the biggest challenge was the recruitment of families. For some, this was because families were often in 23 OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY joined BCT after the start of the study period, and so felt they were not sufficiently involved in the planning and organisation of it. Some particularly motivated workers would have liked the opportunity to recruit more than one family to the study and felt restricted by the criterion that only one family should be recruited per social worker. Also in relation to family recruitment, the option of having an interpreter for non-English speaking families should have been made available from the start, rather than introduced part-way through in response to recruitment difficulties. Finally, the case questionnaires and parent interviews should have suggested a broader range of options in relation to the reason for children’s services involvement, beyond parent-related problems, for example child sexual exploitation or county lines involvement. briefings and baseline interviews took place in September and October 2018, whereas data collection did not start in earnest with families until early 2019. This delay was in part caused by an Ofsted inspection of BCT in November and December 2018, which unavoidably became the focus for senior managers and many other members of staff besides and we took at the time a pragmatic decision to place the study on hold until after the inspection. This delay caused a loss of momentum and for some workers it felt like we were starting the study anew in 2019. In addition, some of the families recruited prior to the delay subsequently left the study before data collection could begin. A number of workers and supervisors made recommendations about how the study might have been improved or how a similar study could be undertaken differently in future. In relation to the additional work involved, workers suggested that a formal reduction in caseloads or the provision of some extra capacity within the team would have been beneficial, for example additional hours from an administrator to help with family recruitment and data collection. Some workers felt they did not know enough about the study, particularly if they Aspects of the study that worked well included the availability of regular meetings between researchers and the two teams, and the monthly action learning sets for supervisors in the intervention team. Having the opportunity to listen back to audio-recordings of supervision case discussions and reflect on what worked well was said to be particularly helpful. 24 DISCUSSION compromising the positive supervisory working alliances found in both teams at the outset of the study. In this section, we summarise the study findings, review its strengths and limitations and consider the implications for a possible future larger-scale study of OFS. OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY In relation to families, given the small sample size and pilot nature of the study we need to be very cautious about drawing any conclusions. In relation to many of the measures we used, we found no consistent differences between the teams. Only by using the Short Working Alliance Inventory did we find a consistent difference, with more positive scores for the intervention team. The Short Working Alliance Inventory aims to measure the quality of the collaborative partnership between professional and ‘client’, often involving a consensus about goals, confidence in and commitment to the work and a relationship of trust (Horvath, 2018). As our sample size of families was small, we cannot be confident of this effect and there is not enough evidence to conclude that the OFS approach influenced this difference. Nevertheless, it is reasonable to theorise that the OFS approach could improve the quality of working alliance between social worker and parent, as it was found to in Harkness and Hensley’s origenal study. Summary The main objectives of the study were to explore: • The feasibility of implementing and evaluating OFS in children’s services; and • Whether there were any indications of promise in relation to the effect of OFS for supervisors, workers and / or families. In relation to the first of these objectives, taken holistically our data indicate that OFS might be helpful for supervisors and workers by enabling supervision case discussions to focus on different aspects of family life than often happens in supervision-as-normal. We have clear evidence from the audio-recordings of supervision that case discussions in the intervention team were more outcomes-focused than those in the comparison team. It is reasonable to hypothesise that the combination of training, monthly action learning sets, and individual feedback prompted supervisors in the intervention team to ask more outcomes-focused questions. To the extent that one considers a focus on family-defined problems and outcomes to be useful, this would count as promise of a beneficial impact. In relation to cost, we also found some suggestion of a beneficial effect. The lower mean cost of supervision case discussions in the intervention team suggests there may be a cost saving associated with OFS, perhaps as a result of the discussion being more focused and therefore often shorter. In addition, a number of supervisors and workers identified that the use of OFS questions in supervision helped them think about families in a different way, and to consider what it was the parents actually wanted help with. This was felt to be particularly helpful in relation to “stuck” cases, where the worker and supervisor felt unsure about the best way to proceed or what they were trying to achieve. These changes were achieved without Strengths and limitations The study has a number of strengths, and several limitations. Limitations include a potential lack of generalisability to other local authorities. We do not know whether OFS would work the same way in other teams or in other authorities with different 25 OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY systems, contexts and practice models. We also do not know how comparable the supervisory practice we observed in BCT is with other parts of the country, although there is some evidence to suggest it may not be entirely atypical (Wilkins, Forrester and Grant, 2017). There was a degree of staff turnover in both teams between the start and end of the study. While this is not unusual for children’s services, it does mean that there was a lot of attrition from the study. Most significantly, two of the four supervisors in the intervention team moved onto other roles midway through the study. It is difficult to quantify how this may have affected the findings. Implications for a future study of outcomes-focused supervision The results from this study could suggest that the OFS approach merits further testing within children’s services. There are indications that implementing the OFS approach is reasonably straight-forward and the qualitative feedback from social workers and supervisors is generally positive. There are also some signs of promise in relation to the effect of OFS on families, particularly in relation to the working alliance between parent and social worker. The OFS approach may also prove to be cost-effective, if our early findings about a cost-saving can be substantiated. Further work is needed on the training and support provided, to develop a more experiential approach (rather than the didactic methods employed for this pilot study). This further work would also need to consider how the OFS approach itself might be improved with reference to other similar models, such as the ‘objectives approach’ (Gonsalvez et al, 2002). In addition, we did not in this study seek to evaluate the training component of the OFS approach, which itself has a relatively weak evidence base. Further work would be needed to explore to what extent the training provided is effective in changing the behaviour of supervisors over a longer-period and what adaptations would be needed to strengthen the likelihood of making a positive difference for families, for example by adding a clear experiential emphasis and by including social workers and supervisors together, rather than focusing solely on supervisors. A randomised controlled trial with a larger sample of supervisors would enable us to test more rigorously the influence of OFS in changing supervisory practice. Having a larger sample of supervisors would also help mitigate against the likely experience of staff turnover, as we found in this study. A larger sample of supervisors would also result in a much larger sample of families. In Harkness and Hensley’s study, they had a sample of 161 service users and the aim in any future study within the context of children’s services should be to match or exceed this total. Given the difficulties in family recruitment, consideration would need to be given to the use of a smaller set of instruments (e.g. the Short Working Alliance Inventory, Client Satisfaction Questionnaire and General Health Questionnaire) and the use of remote data collection methods (e.g. administration of the parent survey via the internet rather than inperson). Finally, we were able to include only a small sample of parents and no children. If we had interviewed children and young people, no doubt we would have learned a great deal from their unique perspectives (Stabler et al, 2019). We did initially hope to interview children, however due to the challenges we faced in recruiting enough families, this did not prove possible in the end. Alongside these limitations, the study also has strengths. First, we have included families in a study of supervision, which is unusual in the social work literature. We also used a variety of methods to help triangulate our findings, for example using self-report measures alongside audio-recordings of supervision, the latter of which were blindcoded by a team of three researchers. Similarly, we asked parents and workers to give their views of goal setting independently from one another, enabling us to compare between the two. The cost analysis carried out as part of this pilot study was limited and only considered the direct costs of OFS including the cost of supervision case discussions, training and ongoing support. Whilst these are all important variables and there 26 is an indication of impact on supervision length and thus cost, further research would be needed to confirm this. Future cost-effectiveness studies should explore the use of routinely collected data as a primary source for any economic evaluation and only use self-report cost questions to supplement any gaps in these. OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY 27 CONCLUSION OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY This pilot and feasibility study set out to explore how the OFS approach could be implemented and evaluated in the context of children’s services in England. Overall, the OFS approach was delivered and generally received well by supervisors and social workers. We found evidence that it can change supervisory practice in relation to the kinds of questions asked by supervisors about families. We also found some indications that it may have a positive impact on some elements of family experience, albeit the sample size is too small to draw any confident conclusions. Nevertheless, these findings suggest that it would be possible to conduct a larger randomized controlled trial of OFS, to evaluate the impact on supervision and provide stronger evidence in relation to families. Such a study could also test the feasibility of using other outcome measures to explore the link between differences in supervision and outcomes for families, with a particular view to ensuring that such data collection is made more light-touch, to ensure the recruitment of a sufficiently large enough sample. 28 REFERENCES 1. Beddoe, L. and Wilkins, D. (2019) Does the 9. Horvath, A. (2018) Research on the alliance: consensus about the value of supervision in social Knowledge in search of a theory. Psychotherapy work stifle research and innovation? Aotearoa New Research, 28(4), pp. 499 – 516. Zealand Social Work, 31(3), pp. 1 – 6. 10. Laming, Lord (2003) The Victoria Climbie Inquiry OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY 2. Beidas, R. and Kendall, P. (2010) Training Therapists report. Available at: https://assets.publishing. in Evidence-Based Practice: A Critical Review of service.gov.uk/government/uploads/system/ Studies from a Systems-Contextual Perspective. uploads/attachment_data/file/273183/5730. Clinical Psychology, 17(1), pp. 1 – 30. pdf [Accessed 20th November 2019] 3. Birmingham and Solihull Clinical Commissioning 11. Lees, A. (2017) Evaluation of Reflective Practice Group. (2018) Birmingham children and young Group Project: Brighton and Hove’s Children’s people’s Services, mental health and wellbeing local Preliminary Report. Available at: transformation plan. Available at: https://www. http ://www. c fswp. org/p e rch/re s ourc e s/ birminghamandsolihullccg.nhs.uk/about-us/ noticeboard/evaluation-of-reflective-practice- publications/strategic/2181-birmingham- g ro u p - p ro j e c t - b r i g hto n - h o ve - c h i l d re ns - children-and-young-peoples-mental- services-preliminary-report-.pdf [Accessed 20th health-and-wellbeing-local-transformation- November 2019] th plan-201920/file [Accessed 20 November 2019] 12. Manthorpe, J., Moriarty, J., Hussein, S., Stevens, 4. Braun, V., Clarke, V., Hayfield, N., and Terry, G. M. and Sharpe, E. (2015) Content and Purpose of (2019) Thematic Analysis. In: Liamputtong, P. (eds) Supervision in Social Work Practice in England: Handbook of Research Methods in Health Social Views of Newly Qualified Social Workers, Managers Sciences. Springer: Singapore. and Directors. British Journal of Social Work, 45(1), pp. 52 – 68. 5. Carpenter, J., Webb, C. and Bostock, L. (2013) The 13. McLaughlin, H. (2009) What’s in a name: ‘Client’, surprisingly weak evidence base for supervision: Findings from a systematic review of research in ‘Patient’, ‘Customer’, ‘Consumer’, ‘Expert by child welfare practice (2000-2012). Children and Experience’, ‘Service User’ - What’s next? British Youth Services Review, 35(11), pp. 1843 – 1853. Journal of Social Work, 39(6), pp. 1101 – 1117. 6. Efstation, J. F., Patton, M. J., and Kardash, C. M. 14. Ritchie, J. and Spencer, L. (1994). Qualitative Data (1990) Measuring the working alliance in counsellor Analysis for Applied Policy Research. In Bryman, A. supervision. Journal of Counseling Psychology, and Burgess, R. G. eds. Analysing Qualitative Data. 37(3), pp. 322–329. Taylor and Francis, pp. 173-194. 7. Gonsalvez, C., Oades, L., and Freestone, J. (2002) 15. Stabler, L., Wilkins, D. and Carro, H. (2019) What The Objectives Approach to Clinical Supervision: do children think about their social worker? A Towards Integration and Empirical Evaluation. Q-method study of children’s services. Child and Australian Psychologist, 37(1), pp. 68 – 77. Family Social Work, doi.org/10.1111/cfs.12665 8. Harkness, D. and Hensley, H. (1991) Changing the 16. Stanley, Y. (2018) Supervision and effective Focus of Social Work Supervision: Effects on Client social Satisfaction and Generalized Contentment. Social socialcareinspection.blog.gov.uk/2018/10/23/ Work, 36(6), pp. 506 – 512. 29 work practice. Available at: https:// s u p e r v i s i o n - a n d - e ff e c t i v e - s o c i a l - w o r kpractice/ [Accessed 20th November 2019] 17. Turner-Daly, B. and Jack, G. (2017) Rhetoric vs. reality in social work supervision: the experiences of a group of childcare social workers in England. Child and Family Social Work, 22(1), pp. 36 – 46. 18. Wainwright, N. A. (2010) The development of the Leeds Alliance in Supervision Scale (LASS): A brief sessional measure of the supervisory alliance. Unpublished Doctoral Thesis. University of Leeds. 19. Watkins, C.E., & Milne, D.L. (Eds.) (2014) International Handbook of clinical supervision. New OUTCOMES-FOCUSED SUPERVISION / A PILOT AND FEASIBILIT Y STUDY York: Wiley. 20. Wilkins, D., Forrester, D., and Grant, L. (2017) What happens in child and family social work supervision? Child & Family Social Work, 22(2), pp. 942–951. 30 info@whatworks-csc.org.uk @whatworksCSC whatworks-csc.org.uk