Challenging behaviour around challenging behaviour

Introduction: The United Kingdom's Department for Education's advice on behaviour focuses on the power of staff and the strength of the policy in challenging behaviour, via rules, sanctions and rewards. We designed a video- feedback intervention for staff teams in a special educational setting who were working with children with intellectual disability and challenging behaviour. The intervention aimed to raise reflective capacity on relational mechanisms that offer new response possibilities in everyday practices within trans- disciplinary teams. Method: We conducted research with three teams (between five and seven participants in each). We report findings from two teams who were working with children (aged between 10 and 14) who staff identified as having behaviour that challenged. The intervention consisted of two video- feedback intervention sessions, using clips of good interactions between themselves and the child and a review. These sessions took place over three or four months. Qualitative analysis was conducted to analyse changes to the language and depictions of the children. Changes to the participants’ goals during the intervention were also analysed.


| INTRODUC TI ON
Bronfenbrenner's ecological theory of child development showed how interactions outside and around the family had an impact on child development inside the family (Bronfenbrenner, 1979). He drew on tangible events in wider society to show child development as a function of systemic interactions. Shotter (2010Shotter ( , 2011 Researching cultures of practice around children and adults whose behaviour is challenging is a priority area for service improvement in the United Kingdom (Department of Health, 2014). A systemic perspective places equal emphasis to what is going on around the child as to what is going on inside the child. The latter has received more attention, to date, in research around children with intellectual disability and challenging behaviour (Hutchins et al., 2017;Losinski et al., 2016;Weston et al., 2017). In this paper, we present exploratory findings of our systemically orientated intervention for teams working in a special school around children whose behaviour was challenging sometimes.
The interactional web around the child contains more than verbal and non-verbal messages. According to Shotter's conceptualisation of living encounters, the interactional web contains clues, often partial and intangible, to the culturally specific concepts that shape beliefs about what it means to be good. The individual's personal biography fuels the imaginative resources used to vision and motivates movement towards being-being a good teacher, therapist, parent, etc. (Brockmeier, 2015). Researchers who use biographical narrative methods suggest that clues to what is of existential importance to individuals are found in the content and structure of their narratives. How the story is being told (i.e., the structure) yields as much insight into the social and material context of a person's perspective as what is being told (i.e., the content). Narrative analysis can add rigour to thematic analysis of group data gathered through conventional qualitative research methods such as focus groups (Ahmed et al., 2017). Narratologist, Michael Bamberg, focuses on the fleeting incidental telling of stories that occur inside longer sequences of dialogue or monologue (Bamberg, 2004). Drawing on the idea that identity is constantly re-negotiated through the narration of oneself; these 'small stories' provide a window on identities as they are coming into being and reveal how the individual is managing the interplay between personal and social identities (Bamberg & Georgakopoulou, 2008).
In prior longitudinal exploratory intervention research, James et al. (2016) triangulated findings from traditional thematic qualitative analysis with micro stories embedded in transcripts from workers in a health/education and social care charity for children and adults with intellectual disabilities. Thematic analysis of the participants' perspectives over the course of the intervention showed that they began to include the perspective of the children and adults in their descriptions of working life (James et al., 2015). A separate analysis of their micro narratives contained vivid depictions from their practice and their pictures also changed to include the service users. Small stories provide clues, often in the form of pictures expressed through metaphors, of the way that participants see the world. Changing pictures in the minds of teaching professionals has been suggested as a way to increase inclusion and participation of children in services (Bijleveld et al., 2015). Language is not used to create a carbon copy of meaning in another's mind (Shotter, 2011); rather our interactional dialogue, which includes the pictures we draw with words and colour in with our body, face and voice, plays a fundamental role in fixing and unfixing identities. Taking a dialogic perspective, looking beyond communication as message coding and decoding system, we sought to explore the interactional web around children whose behaviour was challenging, asking, 'What language (and pictures) is the child going on inside of?'. This systemic orientation can complement research rooted in the behavioural model (Hastings & Remmington, 1994) where the focus on behaviour, its contigencies and reinforcers shapes a large body of research in mainstream education (Maggin et al., 2017;Mrachko et al., 2017). The attribution theory (Weiner, 1972) continues to frame studies in special educational contexts. Surveys of staff's attributions about children's behaviour, such as Erbas et al. (2010) are retrospective in nature and often only conducted with one group (usually staff). The focus on staff is found even in research motivated to explore the role of culture in preventing challenging behaviour (Arthur-Kelly et al., 2017) staff's individual skill and behaviour were the primary outcomes measured. Studies restricted to staff have obvious limitations to apply the research to the interactional realm where cultural concepts and structural positions interact to influence practice.
A recent study explored the culturally specific understanding of violence (Pihl et al. 2018). They looked at the conceptual underpinnings of challenging behaviour. Their study was rooted in the lived experience of workers in special schools and they used narrative analysis to highlight how the concept of violence infused participants' understandings of their work role, work place and work practices, including supervision and leadership. Their conceptual analysis provided a fresh orientation on the concept of violence and showed how it shaped everyday interactional encounters. Research in adult services on team culture and people with intellectual disability and challenging behaviour has long been theorised (Hastings & Remington, 1994) and also found (Knotter et al., 2013) to be a significant explanatory correlate of restrictive practices, (far larger than individual staff characteristics such as experience, educational level or age). The reduction in restrictive practices remains a current relevant policy priority for adults and children's services (Department for Health, 2014). More research on how cultural concepts shape expectations and responses in collective and individual action is needed to inform systems' transformation. Whilst inductive thematic analysis with groups tell us something about themes in the practice context (as Pihl et al., above) and taking more than one perspective may expand understanding of the system, these methods cannot capture the uniqueness of the interactional encounter, which is where change is needed if systems are going to be transformed (Dougall et al., 2018). It is also where the potential for change is located (Shotter, 2010).
If we want to ask what the child is going on inside of without problematising or objectifying the people in the system, research needs to focus on the dialogue around children and their families.
One way to do this whilst retaining a legitimate focus on staff's individual development is to use video recording of in-situ practice.
Video review has been used in the field of intellectual disability to create space to consider resonance between micro moments of interaction and the meaning of the relationship and identities of the participants (Zijlmans, 2014). It can also provoke a dilemma between what is seen to be the case and what is believed to be the case and thus open new possibilities for the future, for the self and the other Mezirow, 1991;Pilnick & James, 2013). Transforming perspective of the child is a key enabler or barrier to any child's development, but re-seeing the self may be a prerequisite to unfixing the picture of the other. Video review in teams provides the opportunity for individuals to explore their personal beliefs about what it means to be a good teacher. Their colleagues' responses provide clues to the micro-cultural interpretation of 'good teacher'. Rooting the review in interaction practices they recognise as evidence of being a good teacher provides opportunity for social and individual reflexivity. This has the potential to create or amplify the team's theory of practice.
The aim of the study was to explore the impact of a new video feedback intervention for staff teams from multiple disciplines who work with children with intellectual disability and behaviour that was challenging. The intervention consisted of two half-day video coaching sessions and a one-hour team review session. The intervention gave the teams the opportunity to see themselves in relation to each other and to a child they all worked with. This created the conditions for them to re-appraise their perspectives of themselves, each other, the child and family as well as their collective identity and the beliefs (or theories) shaping their practices. The exploratory nature of the research and the dialogic theoretical premise shaped the design of this study and the analytical methods. The intervention was designed to create the conditions for moments of living responsiveness that would prompt this re-appraisal. Our research goal was to capture these moments and explore the participants' responses within the intervention sessions, thus generating research findings from the teams' unfolding dialogue (not from retrospective interviews or focus groups after the intervention). We conducted a thematic analysis to provide a rich description of the entire dataset. This theory-free approach to analysis, was especially relevant given the exploratory nature of the research and the aim to explore the participant's perspectives of this new intervention (Braun & Clarke, 2006). We also conducted an analysis of the participant's micro stories which contained the actual or imagined dialogue they used to evidence the reason for their perspectives or beliefs. This combination of analytical approaches coheres with our epistemological premise that the construction of meaning through dialogic interactions with others and with the artefacts in the milieu such as policies and practice conventions, constitute the social domain where narratives, and the concepts contained therein capable of generating renewed perspectives about the self and other, are created. The goals of this initial paper are to provide a rich description of the intervention and address the broad research question asking, how do staff's perspectives of themselves, the child and family change during the course of it.

| ME THOD
We present data from a small-scale exploratory intervention research study drawing on datasets from two multi-disciplinary teams in a special school in the United Kingdom (UK).

| The school context
The school had a trans-disciplinary model where education and therapy staff worked together in the classrooms on shared childcentred goals. The value of the team approach was captured in the mantra there's no 'I' in team. The individuality of each child, however, was central to the school's mission to support children's wellbeing, functional communication and independence. Development was regularly monitored and the findings used as a basis for team-based inquiry and child-centred planning. The collective ownership of children's outcomes and the systemic focus on improvement through a culture of inquiry meant that working on challenging behaviour by challenging the team's perspective of themselves, each other, and the child was conceptually permissible at the school.

| Ethics
The study was reviewed and approved by the Faculty Research Ethics Committee at Northumbria University in the United Kingdom.
Participant inclusion in the study was on the basis that the employees of the school worked in a team where a child was presenting a significant challenge. The senior leaders at the school determined the identification of the teams for inclusion. All staff who were eligible for inclusion received written information about the study. Consent was conducted by a member of the school's Senior Leadership Team.
A senior leader also managed informed consent of parents and carers which was necessary in order for the child to be video-recorded by staff participants. The staff participants managed the child's assent for participation; they were able to explain the research to the child and gauge responses with validity.

| Participants
Eight staff participated in this study. They were working in two teams across Key Stage Three (children aged 11 to 14 years) and Key Stage Four (14 to 16 years). (The intervention was piloted with a team from Key Stage Two (8 and 11 years) for the purposes of intervention refinement. No research data from this team were used in the construction of the findings for this report.) The team from Key Stage Three is referred to as Team A and the team from Key Stage Four is Team B. All participants were employed by the school either in teaching roles (teachers or teaching assistant roles) or in professions allied to health (occupational, physio or speech and language therapists). Further detail on participants and the teams A and B are in Table 1. We did not set out to explore the potential impact of years of experience or professional qualification or educational level in this study, so we did not collect data on the participants' years of experience in role or in employment in the school. However, for descriptive purposes we can report that there was a wide range of years of experience in role (from newly qualified to near retirement) and in length of service in the school (newly employed to over 30 years of employment). Not all staff who were eligible for inclusion in the research chose to participate. One participant in Team B left the school's employment during the study.
It was not necessary nor beneficial to collect personal details about the child or the family at the outset of the study (see further justification of this position in the intervention description). The researchers had the child's first name, sex and age in years at the start of the study, and a minimal description of the behaviour that was challenging.

| Procedure
The research was conducted in the school during 2015 and 2016.
The findings presented in this paper are from the video-review sessions (Session 1 and Session 2) and the final review session (Session 3). The video-review sessions were around 2.5 hours, the review session was an hour and the participating staff members attended all sessions together. All sessions took place in the afternoon and straddled the end of the school day. These elements constitute the intervention. The intervention took place over three months and sometimes spanned two school terms. The first author, who was an accredited practitioner, supervisor and trainer in Video Interaction Guidance (VIG) (Kennedy., 2011) led all intervention sessions. A separate review with parents/carers and the child was offered with the first author, but it was not recorded for research purposes.
The intervention sessions were audio recorded and orthographically transcribed to provide a verbatim transcript of all the verbal utterances. Field notes were taken by the third named author of this paper who attended all intervention sessions for Team A and the Review Session for Team B. Reflective notes were made by the first author after all the sessions. These notes were part of the research record and provided additional context for the analysis and interpretation of the findings. The findings presented in this paper are from the intervention, that is the two video-review sessions and the final review, referred to as Session 1, Session 2 and Session 3 below.

| The intervention
The intervention was based on the principles of an established family intervention, Video Interaction Guidance (Kennedy et al., 2011) and its sister intervention, Video Enhanced Reflective Practice (VERP), which focuses on workforce development (Kennedy et al., 2015).
Like VIG, the adapted intervention works through a strengths-based relational lens. The main organising structure in VIG and VERP is the establishment of a desired goal for a relationship or set of relationships. In VIG, the interventionist, or guider, establishes that goal with a single person or couple. The guider then takes video footage of the relationship within the situated practice (i.e., the family home) and edits the footage to find moments where the desired goal is achieved, partly achieved, or has elements that indicate development towards the goal. In summary, there were two main departures from the VIG/VERP model.

Number of participants
Participants' roles Key stage

Child gender
Team A Unlike VERP, where practitioners reflect on their individual and often unrelated work. This approach was designed for teams who worked closely with each other on a daily basis. In the adapted intervention, we worked with the team's goal for the child's development and focused on individual participant's goals.
2. VIG has a set of attunement principles that are used to guide the editing of video recordings from situated practice. These attunement principles were not used in this intervention. Instead, the team's understanding of the child and his ways of communicating and/or expressing happiness, learning or wellbeing was the basis for video editing. This led to much greater use of the team's expertise by the facilitator who had to rely on their understanding of the child to edit video footage.

| Session 1
The main aim for Session 1 was to provide a positive and developmental experience of the intervention. We used video recording, video editing and team-based video review. This was achieved in the following activities led by the facilitator (the first author).
1. Identification of a personal goal for each team member and the team goal for the child.  At Session 2, each participant came with a recording of themselves with the child. As in Session 1, the editing was conducted during a break in the middle of session. The facilitator worked in a separate room and during this time, participants discussed their action learning objectives for the child (activity 3 above). Recordings varied in length from 2 to 30 min, but most were between 5 and 8 min long (see James, et al., 2012). Video editing was necessarily rapid. After the break, each participant saw a clip of themselves having a positive impact on the child (using the team's framework for successful interactions). Where possible, this strength was linked to participant's personal goals.

| Review session
Session 3 reviewed the team's perspective of the intervention. We explored participant's development towards personal goals, their team goal and the effects of their changes on the child. Video was not used.
The mother of the child in Team B attended a parent review with the first author prior to Session 3. Several video clips of the staff working with her son were shared and we discussed her view of the intervention and its effects on her child. She said which parts of our discussion could be shared with the team. Session 3 differed, therefore, with respect to the inclusion of the mother's perspective in Team B.

| Analysis
All

| Findings
The themes from inductive coding of the participant's verbal utterances during the intervention sessions are in Table 2

Changes to team's perspectives
Themes that arose from the analysis of the micro stories were triangulated with findings from the whole dataset and this guided the selection of data extracts to exemplify the thematic findings of change in Team A (Table 3) and Team B (Table 4). An interpretation of these basic themes is provided in the final column of each table.

| Organising themes
We now turn to consider how change, our global theme, was entwined with the intervention process.
Individual goal setting and monitoring was a key component of the intervention. The main mechanism for developing goals was through the use of video editing and review of interactional strengths. So first, it is important to note that at the beginning participants found it challenging to see and hear themselves on video. However, they quickly became comfortable enough to see beyond their appearance or behaviours, to see and think about the connections between themselves and the child. The participants distinguished the intervention from other ways of working. One participant described the process as 'surreal'. It was like being 'naked almost' and 'egoless', and he concluded that 'leaving your ego at the door is hard'. The individual goals were a way of making the idealised future concrete and possible. They were used to guide the selection of video edits for review, but they were also a starting point for participants to articulate their understanding of themselves and the enactment of their idealised self. The proposition was that the processes of self-renewal would open the possibility for renewed perspective of the other.

| Individual goals-Team A
One participant's goal at the outset had been to be less of a people pleaser. After watching video footage in Session 1, she concluded lack of confidence might be a problem, not that she was a people pleaser. The video review drew her attention to the interaction between context, self-narratives and her own beliefs. A similar reframing occurred for another participant whose initial goal was to be a better communicator. After video review of himself and with reflection between sessions, he concluded that he was not as bad a communicator as he had previously thought. He recognised that he often paused to give himself time to think of what to say, and that he was more thoughtful in how to communicate and explain things. A third team member changed her communication, saying that she no longer talked over people as much in meetings, and for this participant, the video review sessions also brought about a new way of framing herself and the child; she realised that she and the child were in the same team. This individual's shift is indicative of what appeared to be a core underlying perspective transformation across the team as indicated in Table 3. The participants were beginning to see themselves and the child as inhabitants of the same world. One participant commented on a video of her colleague interacting with the child, saying that it was 'nice' because it almost made him an equal with the child 'tending with him rather than shadowing over him'. In this quote, the sports metaphor (a facet of Team A's way of seeing the world) was replaced by a horticultural one. The extension of this ecological perspective into understanding of their role is resonant with the changes seen in Table 3. Not only do they develop much greater awareness of the environment and how it impinges on the child (see Physical Environment), but also they see the child through those conditions (e.g., Perspective on Child) and enact changes to their practices, accordingly (e.g., Rewards and Sanctions).
The quotes in column one of in the team, their initial framing of the child was restricted to his negative traits, which centred on his inconsistency and fear of failure. He was a child who needed micro-management. As can be seen in Table 4, their view of the child changed and they began to enjoy working with him and working him out.

Summary
The articulation of individual goals with video review of situated practice was critical in provoking changed perspectives in both teams' individual and team identities. Both teams developed a theory for change rooted in practice, a priority for inclusive education (see Nilholm, 2020). Seeing themselves and the child as inhabitants of the same world and seeing their development entwined with the child emerged as fundamental themes. In the next section, we demonstrate the most prominent outcomes in their changed perspectives, explore how these changes were narrated by participants and consider the intervention elements that fostered them.

| Narratives of the children's behaviour
The teams were not explicitly asked for descriptions of the child's challenging behaviour rather the explicit request was for descriptions of how the child expressed feelings of happiness, engagement etc. Nevertheless, both teams described the child's inconsistency as a challenge in Session 1. The child's underlying anxiety, fear and 1 In a later Ofsted inspection, this specific example of teaching practice was highlighted as nationally outstanding.

Summary
Our analytical approach was informed by the proposition that quoting the speech of others, whether those words were real or imagined, provides interactional evidence for the speaker's beliefs (Clift, 2006). We used that analytical method to highlight the most prominent changes in the teams' perspectives. These were increased empathic perspective on the child and family in Team A and greater self-belief in their ability to overcome the child's history and fear of failure in Team B. These new perspectives were accompanied by new theories of practice. By design, our approach prioritised working with the teams' knowledge to create equality in the mobilisation of expert knowledge between facilitator and team (Heritage, 2012).
This active engagement with the team's knowledge, coupled with the strengths-based video editing of situated practice created team reflexivity on practice and with it the potential for sustained and generalised change.

| DISCUSS ION
In this study, both teams changed their perspectives on children they were finding challenging. At the start of the intervention, the child and family were objects (predominantly negatively depicted). The depictions of the children and their families became more positive over time, making these findings comparable to prior research from a similar service context (James et al., 2015). As in prior research, increasing staff's positive feelings towards the beneficiaries of the service went hand in hand with increased intent to understand them, listen to their 'voice' and make reasonable adjustments for their inclusion in and benefit from services. These elements are so important in quality service provision that they are specifically monitored in the UK by the Care Quality Commission in reviews of health and care services that have a propensity to create closed cultures and associated abuses of power (Quality Care Commission, 2019).
As the intervention progressed, the child was no longer seen as the problem, but neither had the 'problem' disappeared. In fact, the micro-story analysis showed that the problem had become more visible as it was taken out of the child, positioned in a space like the medical room or the family hallway, and placed where it could also be shared. In both teams, the pictures painted of the child at the outset were of isolated figures, but at the review, the children were Did the language around the child change? Whilst we did not directly investigate this outside the intervention, the pictures and perspectives of the children did change and interaction practices that have the potentiality to sustain more nuanced, multi-perspectival understanding of the children were clearly operating in both teams by the end of the intervention.

| Did the intervention challenge behaviour around challenging behaviour?
Although both teams placed greater emphasis on finding inherently motivating rewards, the school's policy using rewards and sanctions remained unchallenged. The lack of critique was surprising.
At the time of the project, the school's behaviour policy was being re-written to reflect a relationally based preventative premise for practice. One of the participants was implementing the new behaviour policy, which focused on the child's perspective of behaviour management. So, questioning the dominance of the behavioural reward/sanction policy was possible at this school. Despite this, neither teams questioned the restrictions that the narrowly defined policy created in structuring their practices. In future, further adjustments in facilitation to support collective critical awareness will be explored (Mejia et al., 2019) because school cultures that afford distributed intra-organisational inquiry into operational structures are likely to create better conditions for development and inclusion (Ainscow et al., 2013).
The intervention challenged the school's senior leadership mantra that there is 'no "I" in team'. Individuals clearly influenced the topics for consideration. Individual participants' knowledge and practices also shaped the teams' narratives around the children.
As outlined in the introduction, individuals have unique interpretations of what it means to be a good teacher, therapist, parent etc. (Brockmeier, 2015). The intervention created an opportunity for front line staff to experience the mutuality of their development with that of children and families. This orientation reduces the probability of problematising individuals whether they be staff or services users and is aligned to contemporary systemic approaches to improving support around children with developmental disabilities and behaviour that challenges (British Psychological Society, 2018).

| CON CLUS ION
The relational context of practice was highlighted using video footage of participants' interactions with each other and the child. The concept of interconnectedness was highlighted by linking participants' individual desires to their actual relational worlds. The relationship with the child became part of their individual and team narratives. Thus, the intervention re-orientated away from a theory of individuality towards a concept of mutuality without dishonouring, displacing or disregarding each individual. The intervention design differs from its sister interventions as it sees participants as a collective, (or even as a single body). There is no evidence that these departures from the VIG protocol limited the change that might be expected using a traditional VIG model. In future, the impact of the intervention on the lived experience of the child and family, their interactions with the system and incidents of challenge should be recorded.
This small-scale research was conducted in a special school that was open to challenge and inquiry. Even in this context, we found traits indicative of closed cultures. Our intervention was designed to draw on the realities of practice, not to refine skills and techniques or impart specialist knowledge, but to stir up tacit knowledge and parallel processes that underpin, and sometimes fix narratives that reduce opportunities to see and act differently. This approach aims to address challenging behaviour through a systemic lens, and we will continue to explore the potentiality of it to support the safety, inclusion and development of children and families within their communities.

ACK N OWLED G EM ENTS
We thank Gill Naylor and Lesley Gallacher for their contribution to the investigation. We thank the senior leadership of staff at the school for supporting this research. We thank all the staff at the school who took part in this investigation, the parents who gave their consent for their child's participation and the children who gave their assent to be included. We acknowledge the role that Barry Ingham and Julie Morrow from a local NHS mental health and learning disability foundation NHS trust played in the development and supervision of similar work that was undertaken under an honorary clinical contract by the corresponding author. This work was funded by the school via a contract research grant awarded to the corresponding author.