Acceptability of Dance PREEMIE (a Dance PaRticipation intervention for Extremely prEterm children with Motor Impairment at prEschool age) from the perspectives of families and dancer teachers: a reflexive thematic analysis

ABSTRACT Background Physical activity (PA) participation is important for children born extremely preterm or extremely low birthweight as it provides opportunities to improve motor skills and cardiovascular fitness; however there is little evidence on interventions promoting PA participation for this group, particularly at preschool age. Objective This study aims to explore the acceptability, a critical component of intervention feasibility, of a novel dance participation intervention. Methods Semi-structured interviews explored the acceptability of Dance PREEMIE, a Dance PaRticipation intervention for Extremely prEterm children with Motor Impairment at prEschool age (trial registration ACTRN12619001266156), from the perspectives of dance teachers delivering the intervention (n = 6), and parents of participating children (n = 6). Data were analyzed using reflexive thematic analysis. Results Five themes were developed: 1) placing the child center-stage: a shared motivation to promote child wellbeing and development; 2) Dance PREEMIE as a catalyst for participation; 3) child development takes time, practice and exposure; 4) the value of being informed; and 5) dance teachers as architects of the learning environment. Conclusion Dance PREEMIE was acceptable to both parents of participating children and dance teachers. Findings from this study may inform future interventions aiming to improve PA participation for children with motor impairment at preschool age.


Introduction
At preschool age, participation in physical activity (PA) provides opportunities for children to develop motor, cognitive and social skills, as well as improve cardiovascular fitness and muscle strength (Timmons, Naylor, and Pfeiffer, 2007). Providing children with positive PA participation experiences may promote ongoing PA throughout the lifespan (Imms et al., 2017). This is particularly important for children with motor impairment, who are less likely to engage in PA throughout childhood (Cairney et al., 2005). Recent studies have explored a range of interventions to promote PA participation for children with disabilities (Clutterbuck, Auld, and Johnston, 2020;Reedman et al., 2019;Willis et al., 2018) recognizing that interventions need to target participation directly, rather than assuming improving motor skills will lead to increased participation (Wright et al., 2008). However, there are few interventions targeted specifically at preschool age children (Adair et al., 2015), and little to no evidence available on promoting PA participation for children born extremely preterm (EP: < 28 weeks' gestation) and/or extremely low birthweight (ELBW: < 1000 g birthweight) (Cameron et al., 2020a). Children born EP/ELBW are at greater risk of motor impairment compared with children born at term, with approximately 15% diagnosed with cerebral palsy (CP) (Himpens et al., 2008) and 50% with non-CP motor-impairment (Williams, Lee, and Anderson, 2010). Considering the high rates of motor impairment in this cohort, and the known associations between motor impairment and PA participation, further evidence of how best to promote PA is warranted.
Participation is a complex construct, and can be defined as "involvement in a life situation" (World Health Organization, 2007). It is comprised of both attendance (being there) and involvement, or the experience of participating (Imms et al., 2017). A life situation is considered to be "sets of organized sequences of activities directed toward a personally or socially meaningful goal" (Coster and Khetani, 2008), for example a dance class, family mealtime, or reading time at the community library. By focusing on meaningful goals, participation becomes achievable for individuals with diverse capabilities (Coster and Khetani, 2008). The family of participation related constructs (fPRC) is a framework that explores the complex interactions between the individual and the environment; an intersection that creates the context in which participation takes place (Imms et al., 2017). The relationships between the environment and person-related factors (i.e. sense of self, activity competence, and preferences) are transactional and bidirectional, with both ultimately influencing, and being influenced by, participation (Imms et al., 2017). It is therefore possible to conceptualize participation as: 1) a means for children to learn and engage; and 2) an outcome or objective (Imms et al., 2017).
A Dance PaRticipation intervention for Extremely prEterm children with Motor Impairment at prEschool age, known as Dance PREEMIE (Cameron et al., 2020b), is a novel community-based dance intervention aiming to promote PA participation using participation as both a means and an outcome. Dance PREEMIE used an ecological approach, with physiotherapy expertise employed to train and support community providers of PA. There is increasing recognition that for interventions to be successful, they need to be acceptable to all involved (Sekhon, Cartwright, and Francis, 2017). Acceptability is concerned with the appropriateness of an intervention, based on "the experiential, cognitive and emotional responses" of those receiving and delivering the intervention (Sekhon, Cartwright, and Francis, 2017). The aim of this study was to understand the acceptability of the Dance PREEMIE intervention from the perspectives of the dance teachers and the parents of the participating children. A detailed understanding of how Dance PREEMIE was perceived will enable meaningful improvements to the intervention, and inform future intervention designs for similar cohorts of children.

Study design
This qualitative study used semi-structured interviews to explore the acceptability of Dance PREEMIE from the perspectives of dance teachers and families. It is reported in accordance with the consolidated criteria for reporting qualitative research (COREQ) guidance (Tong, Sainsbury, and Craig, 2007). Interviews were analyzed using reflexive thematic analysis, a form of qualitative analysis used for identifying meaningful patterns or themes within data (Braun and Clarke, 2006). According to Braun and Clarke (2006) a theme is a pattern of common meaning, supported by shared concepts or ideas, which can be used to inform actionable outcomes. This study was granted ethics approval from the Royal Children's Hospital (38133), The Royal Women's Hospital (Project 18/38) and The University of Melbourne Human Research and Ethics Committees, Melbourne, Australia. Written informed consent for was obtained from all participants.

Intervention description
Full details of the Dance PREEMIE protocol have been previously published (Cameron et al., 2020b). In summary, Dance PREEMIE consisted of 8 weeks of dance classes, with children allocated to an existing preschool dance class (30-60 min duration) in their local community. Classes were taught by dance teachers who had attended physiotherapy-led, study-specific dance teacher training. The training session was designed to support dance teachers in creating an environment which facilitated the participation of children with motor impairment. The training included relevant information on preterm birth and common diagnoses of motor impairment, including CP and developmental coordination disorder, as well as evidence-based strategies for motor learning, such as adjusting task or environment to allow each child to achieve (Pollock, Missiuna, McKechnie, and Embrey, 2012). Each dance class was observed for implementation fidelity at least once during the study by investigator (KLC). Participating children attended pre and post-intervention assessments with a physiotherapist for motor skills assessment and goal setting/scoring. Parents were provided with a written report on their child's motor skills after the pre-intervention assessment. The study coordinator maintained regular communication with both dance teachers and parents throughout the study period to ensure the intervention ran smoothly, and to address any questions or concerns. Further details about the intervention design, including details of the studyspecific training, are published in the study protocol (Cameron et al., 2020b). The feasibility results of Dance PREEMIE are published elsewhere (Cameron et al., 2021).

Participants and study procedure
The Dance PREEMIE intervention was completed twice: Program 1 occurred between October and December 2019 with two dance teachers and two children; and Program 2 between January and March 2020 with four teachers and eight children. Children participating in 2020 were offered fewer classes, and could not complete their post-intervention motor assessment, due to the introduction of physical distancing regulations related to the COVID-19 pandemic (Cucinotta and Vanelli, 2020). Two groups of participants were invited using purposive sampling, as part of the consent process of the feasibility study, to complete semi-structured interviews: the primary caregivers of all children participating in Dance PREEMIE, and all participating dance teachers.
Children were recruited into the Dance PREEMIE study from the Victorian Infant Collaborative Study 2016/2017 cohort at 3 years' corrected age. As Dance PREEMIE was a feasibility study, all children were allocated to the intervention. Children were eligible if they had motor impairment at their 2-year developmental assessment defined as scoring > 1 standard deviation below the mean established by a contemporary cohort of children born at term on the Bayley III Scale of Infant and Toddler Development (Bayley, 2005). Children were excluded if at 3 years of age they: 1) were unable to follow simple commands; 2) were non-ambulant; or 3) had a medical condition precluding participation in PA (Cameron et al., 2020b).
Dance teachers were recruited from across metropolitan Melbourne, with targeted recruitment of teachers conducting classes at a convenient distance from participating families' homes. To be included in the feasibility study, dance teachers were required to be teaching a current preschool dance class, attend the Dance PREEMIE training day, and have a valid Working with Children check and a designated first aider present at each dance class (Cameron et al., 2020b).

Interview procedure
Interviews focused on the experience of being involved in Dance PREEMIE, were completed at the conclusion of the intervention, offered in-person or via telephone, and were audio-recorded. The interview schedule consisted of open-ended questions to facilitate the collection of rich data exploring participants' feelings and viewpoints on the intervention (Supplementary Table 1). In order to encourage in-depth discussion on the complex construct of acceptability, the interview schedule development was guided by the Theoretical Framework of Acceptability (TFA), which proposes seven constructs of acceptability: affective attitude, burden, ethicality, intervention coherence, opportunity costs, perceived effectiveness and self-efficacy (Sekhon, Cartwright, and Francis, 2017). Strategies to avoid misleading data were employed, such as repeating back the interviewees language, rather than interpreting meaning through the use of synonyms, and avoiding leading questions (DiCicco-Bloom and Crabtree, 2006).

Research team, reflexivity, and trustworthiness
Reflexivity is essential throughout the analysis process to promote trustworthy and quality results (Braun and Clarke, 2006;Korstjens and Moser, 2018). Strategies used to enhance the rigor and trustworthiness of the research are outlined in Table 1 (Lincoln and Guba, 1986). Our all-female research team consisted of five physiotherapists (KLC, CEM, KA, JLM, AJS), all with qualitative research experience, and a neonatologist (JLC). Reflexive thematic analysis was undertaken by KLC and CEM. The semi-structured interviews were conducted by KLC, a physiotherapist and PhD student who had undertaken training in qualitative interview techniques. In addition, KLC designed and developed the Dance PREEMIE intervention as part of her PhD, delivered the teacher training and acted as the study coordinator throughout the project. KLC's involvement in all aspects of the study allowed an in-depth understanding of interview material and for rapport building with participants, which are associated with better interview data (DiCicco-Bloom and Crabtree, 2006). However, research participants were aware of KLC's role within the study, and as such, it is possible that participants may have been reluctant to talk about negative aspects of the study. To address this, throughout interviews KLC used several strategies to prompt participants to discuss negative aspects of the study including: reassuring participants that anything discussed would not affect their relationship with the study team, encouraging open and honest communication and prompting for negative experiences or barriers to involvement. CEM is a physiotherapist and PhD student with previous experience in qualitative methods. She was only involved in the data analysis process for Dance PREEMIE and does not work in the field of early childhood development, thus she was naïve to the literature regarding participation and less likely to search deductively for known theories and ideas in the data. KLC and CEM completed weekly reflexivity memos throughout the analysis process to engage with and identify possible subjectivities brought to the analysis (Korstjens and Moser, 2018). For example, both CEM and KLC participated in dance classes throughout their youth and had previously worked as dance teachers and therefore were aware of their personal preference for dance when interpreting the data. To further enhance trustworthiness of the work, the analytical process and final themes developed by KLC and CEM were overseen and checked by KA a senior qualitative researcher. This oversight enhanced the rigor of the analytical process and the credibility of the thematic results.

Data analysis
Our approach to reflexive thematic analysis was developed to align with the research aim of understanding the acceptability of the Dance PREEMIE intervention from the perspectives of both teachers and parents. We chose to analyze data inductively, thus developing themes from the data rather than using a preconceived list of themes (Willig and Stainton-Rogers, 2017). This approach allows for the identification of unanticipated themes; particularly important as Dance PREEMIE was a novel intervention, and there are few studies considering the acceptability of participation-focused interventions for preschool age children. Furthermore, we chose to interpret data empathically, aiming to elucidate and amplify the experiences of the research participants (Willig and Stainton-Rogers, 2017). To perform the analysis, audiorecorded semi-structured interviews were transcribed verbatim by an external provider and checked against the recordings by a member of the research team (KLC). Data were analyzed using NVivo 12 software (QSR International Pty. Ltd.). Two members of the research team (KLC and CEM) familiarized themselves with the data through repeated listening of the interviews, to remain close to the original data and pay attention to verbal cues; and reading of the transcripts. After familiarization, each researcher: 1) identified and labeled all meaningful segments within each individual interview to develop the initial codes; 2) revised the initial codes and began formulating candidate themes; 3) compared codes and candidate themes between interviews from the parent perspective and the teacher perspective, to understand concepts and candidate themes from different viewpoints, thus triangulating the data; 4) reviewed and revised themes within and across groups (i.e. teachers and parents); 5) defined and further refined final themes, which were discussed and agreed upon by all members of the research team (Braun and Clarke, 2006). This process was fluid and recursive, codes and themes established throughout this process were continuously evolved and reviewed against the data, while writing occurred across all the phases (Braun and Clarke, 2006). Data saturation was not used to determine sample sizes of the number of participants completing interviews or the number of interviews analyzed. The concept of data saturation implies that themes or codes exist within the data and that the researcher's role is to find them (Braun and Clarke, 2021). This assumption is incongruous with reflexive thematic analysis, which understands themes to be generated through reflexive interpretation of the data (Braun and Clarke, 2006, (Lincoln and Guba, 1986).

Criteria
Strategy Details Credibility Prolonged engagement Interviews were conducted by KLC, who was familiar with all aspects of the study, and had established rapport with all participants through her role as study coordinator. Interviews were of sufficient length to explore the research topic. Researchers undertaking reflexive thematic analysis spent several hours familiarizing and engaging deeply with the data set. Triangulation Both parents and teachers were interviewed, allowing for perspectives from two different groups involved in the intervention; those delivering and those receiving. Two members of the research team analyzed the data, while a third member of the team reviewed the data analysis process and results to confirm credibility. Transferability Thick description The use of detailed demographic and context descriptions will enable readers to decide if results are applicable within their own situations or settings. Dependability and confirmability Audit trail A detailed description of the data collection and analysis process. Weekly memos were completed by each of KLC and CEM during the analysis process. Confirmability was achieved through discussion of the analysis process at multiple time points throughout the analysis by KLC and CEM, and through a review of the data by KA. Final themes were discussed and agreed upon by all members of the research team. Reflexivity Diary Weekly reflexivity memos were completed by each of KLC and CEM during the analysis process. A reflexivity paragraph is included in the methods section of this paper, discussing possible biases as researchers as well as the researcher relationship to participants. 2021). The results outlined in this paper reflect the final conceptualizations of the themes we developed, as well as sub-themes, illustrated with extractions from the interview transcripts. To maintain participant anonymity, all participants have been allocated a participant number, identifying information has been removed from quotes and gendered pronouns replaced.

Participant characteristics
Semi-structured interviews were completed with six parents of eight participating children (two sets of twins), and six dance teachers. Interviews ranged from 16 to 33 minutes in duration (median of 23.5 minutes). One interview was conducted face-to-face at the Murdoch Children's Research Institute, and the remainder over telephone. One family was approached via e-mail and telephone, but could not be contacted to organize an interview. One family withdrew as they were not satisfied with the pre-assessment report and were not approached for an interview as requested by the family at the time of withdrawal. The dance teacher matched with this child subsequently withdrew and declined an interview at the time of withdrawal. Characteristics of children whose parents did and did not complete an interview are provided in Table 2. Dance classes took place at dance schools in metropolitan Melbourne. Most schools were suburban dance schools operating from a single location. One dance school (two teachers recruited) was a company with multiple studios. All dance schools taught a variety of dance styles; four dance schools catered for children of all ages, while one dance school catered exclusively for preschool age children. Three of the participating teachers owned their own dance school. Most teachers had a tertiary qualification in dance (n = 4) or were undergoing a related tertiary degree (n = 1). The remaining teachers had completed training through their dance company (n = 2). Further details on dance classes are provided in Table 2. Classes offered by school during term, median (range)** 6 (5-8) 6.5 (5-8) Distance from family home, median kilometers (range) 2.0 (0.85-9.0) 2.0 (0.85-9.0) *Two sets of twins and one child whose twin was not enrolled in the study; **data from 9 children, number of classes offered were affected by COVID-19 regulations in Melbourne, Australia; MABC-2 = Movement Assessment Battery for Children-2 nd edition; n = number; SDQ = Strengths and Difficulties Questionnaire.

Themes developed in the analysis
Five themes were identified through reflexive thematic analysis. Each theme is described below, inclusive of illustrative quotes. Further quotes supporting each theme are provided in Supplementary Table 2.

Placing the child center-stage: a shared motivation to promote child wellbeing and development
The promotion of child wellbeing and development was important to both parents and teachers, and emerged as one of the strongest themes in the data. All involved believed this was achieved by prioritizing each individual child's needs and preferences. Both parents and teachers spoke of strategies to promote wellbeing through facilitating a positive mood, fostering confidence, or encouraging a sense of inclusion; therefore cultivating successful participation. Both parents and teachers observed that a positive experience was associated with engagement in the class, while a child's negative mood adversely affected participation. For teachers, this was best articulated when class was not going according to their lesson plan, with children's emotional wellbeing considered the most important aspect in salvaging the situation There were definitely weeks where everything works . . . and there were other weeks where I feel like it just completely went out the window and you've just kind of gotta do your best to get it done and try and give a positive lesson to each of the kids (teacher 2) Similarly, parents spoke of finding ways to promote their child's wellbeing, focusing on enjoyment, confidence, and inclusion I used to have little talks with [my child] . . . I bought them everything that, you know, that they wore at the dance . . . and I've really gotten them motivated and by talking to them and just before classes, after classes. And I have a good chat with them and I was just making them really enjoy it more by our conversations (parent 5) In addition to wellbeing, both parents and teachers were motivated to promote or find opportunities for development. For many families, this was why they decided to participate in Dance PREEMIE, suggesting a good value fit between the perceived intervention aims and participating families: I'm always interested in anything that may potentially help them develop, develop, and succeed in motor skills, social skills, language skills; whatever it is that means that their development is progressing (parent 3) The shared value of a child-centered approach, with a focus on wellbeing and development, facilitated positive intervention coherence and ethicality, with both parents and teachers agreeing on the program aims and outcomes. Ultimately, the child's emotional response to the dance classes was monitored closely by both parents and teachers; a high value was placed on children enjoying the dance sessions, and feeling confident in the dance classroom: [my child] has lots of therapy, they have lots of OT and speech therapy and psychology and while . . . they make all those things fun for children, this is just something fun for them to do with no expectation and no pressure on them to perform or to do anything, so it was good to take them to do something that was just for fun, simply for fun, and I think they really enjoyed that (parent 1)

Dance PREEMIE as a catalyst for participation
Not only did families view dance classes as enjoyable, but for many families, Dance PREEMIE acted as a catalyst to change PA participation. The opportunity for their child to try something "that they wouldn't have otherwise had the opportunity to do" (parent 2) was described as a new and exciting experience for many families: Parents observed their child's success and enjoyment in participating in PA and were motivated to enroll their child in dance for the following term, or to try a new activity.
The act of participating itself was enough for parents to feel more confident in their child's ability to participate further. This sentiment was often referred to in the context of having a child born preterm, or a child with a disability: because they were born preemie [preterm] I was very curious as a mother, as a parent, to see if . . . just to see how they will [cope] with that [Dance PREEMIE classes]. So that did give me a chance to see that they were coping, it made me as a parent happy (parent 5) Parents also observed changes in their child's PA participation and associated these changes with children experiencing a "taste of physical activity" (parent 2). Children were keen to show their parents what they had learnt in class, were dancing to music at home or were "more willing to try new things now" (parent 1). This suggested that parents perceived the intervention as effective as it had increased their child's participation in PA within and outside of dance classes.

Child development takes time, practice, and exposure
Perceived effectiveness was explored further though a common understanding of childhood development shared by parents and teachers, in which time, practice and exposure were all essential components for change. This was described in the abstract sense, with a shared belief in what was required for change to occur: the main thing that I notice with [children's] development is being more confident within themselves and relaxed within that environment . . . and that can happen in the first week or two. Sometimes it takes a little bit longer (teacher4) Improvements observed in the participating children by both parents and teachers provide more concrete examples, with changes attributed to trying dance, spending time practicing or repeating dance steps, gaining experience of the classroom environment, or attending multiple classes. Through this theme we gained further indication that the intervention was perceived as effective by parents and teachers, with developmental improvements observed in domains as diverse as motor skills, confidence, endurance, mood, attention, behavior or communication skills: towards the end of the term I've noticed that they know the words to some of the songs and they know the actions and they know sort of what's going to happen next versus the first couple of weeks where they would just outright refuse to . . . clap their hands or even look at the teacher, so yeah I feel like they've come a long way (parent 1) For teachers, the understanding that change took time, practice and experience was closely connected to their expectations of the child's performance or behavior, and the strategies they chose to use in the classroom: repetition is really important for kids this age . . . they need to do the same exercise each week, repeating it, otherwise they're not going to get it. I think that really helped, like I saw, [participating child], specifically, they were starting to finally get some of the techniques . . . by like the last two weeks that I had them because we constantly repeated it all the time (teacher 5) In particular, teachers felt that starting classes at the final term of the year, arriving late, or missing the first class, was a disadvantage as other students had spent time becoming familiar with the classroom environment: I think that it's really hard to have them start after the term's started . . ..what it does is it puts pressure on the child because they're coming into an environment where they're the new thing instead of everyone and everything all being new at the same time (teacher 3) For both teachers and families, putting in time and effort to promote development was considered both necessary and acceptable. Dance PREEMIE fell short in not providing enough time, with both parents and teachers feeling that more time would have produced a greater effect, with one parent commenting; "I'm sure that if [they] had continued more in following weeks [they] would've been even better" (parent 5). This sentiment was particularly pronounced for teachers and families who participated in the second program, which was shortened due to canceled classes because of government imposed restrictions introduced during the COVID-19 pandemic: I was a little bit frustrated because of course, what we've experienced as far as this global situation, so I think I could've given more. I think I could've achieved more (teacher 4).
However, the sense of needing longer was also articulated by those involved in the first, full length, program; "there's only 10 weeks in the term and . . .. 10 weeks isn't a long time" (teacher 1). This suggests parents and teachers experienced low burden with the current model and implies that a longer intervention would have been acceptable, perhaps more acceptable, to both parents and teachers.

The value of being informed
For both parents and teachers, having or gaining information was considered valuable. Information, whether the pre-intervention assessment results or training material, acted as a strong incentive to participation in the study, while learning something new was often perceived as a key benefit of involvement. Useful or adequate information was associated with feeling involved, prepared and confident: In terms of the involvement, the communication has been excellent. There's always updates, emails and follow-up phone calls, and I felt like I've been very much involved and in the loop the entire time (parent 3) In contrast, not having enough information was associated with feelings of nervousness or uncertainty: I would've really liked to have met [participating child] before the class but I don't know if that's like, um, practical, I guess. But I think it would just calm my nerves a little bit, like I was quite nervous before I met them because I didn't know -when I was planning my class, I wasn't sure what they were able to do (teacher 5) Within this theme, there are two clear sub-themes, illustrating the ways in which teachers and parents required different types of information, and value these differently.
Advancing my teaching practice. This subtheme focuses on the connection between the knowledge teachers feel they have gained or consolidated through participation in this intervention, and its relationship to self-efficacy and perception of the study as valuable to their teaching practice. Teachers explicitly articulated the value of the information gained through the physiotherapy-led training day; through the training itself and from interacting with other participating teachers.
Many teachers felt what they had learnt was applicable to all their dance students, suggesting the training had merit beyond its applicability to the study: I really got a lot out of it [the training] and even in all of my dance classes, whether that's for young children or into the older ones, it actually helped me develop the way that I've been teaching. It's given me more tools to work with (teacher 4) Teachers commented on increased awareness of their teaching strategies, of introducing new strategies or better understanding the 'why' behind their already existing teaching practice. Both learning something new and consolidating existing knowledge was linked with feeling confident (self-efficacy) to deliver the intervention: the training day really, um, probably made me just confident enough to go and do it. Like before the training day I was a little bit nervous, like questioning whether like I'd be, you know, up to the task kind of thing. But I think the training day and just talking to [the physiotherapist delivering the training] and the other like ladies in the room was like just a really good reminder that we all are capable of doing our job, like, and that we're ready to help out all the kids no matter what their disability is, like we're all catering to help the kids no matter what (teacher 5) The training day could also produce conflicting feelings, with nervousness or uncertainty associated with increased awareness of the challenges faced by children born EP/ELBW, alongside more positive feelings of excitement about welcoming new students and offering an inclusive experience. This compelled many teachers to plan more thoroughly for classes: I became more prepared because of the training day which is needed so, I wasn't nervous per se or anxious, I was just, understood that okay I do need to prepare for this class actually I can't just walk in and do my normal thing or just do it absentmindedly, I do need to put some thought into it and, achieve some goals (teacher 1) Teachers also valued receiving specific information about each child's abilities; some teachers expressed a wish to meet with participating children, or to know more about their motor function, prior to the intervention period.
Finally, teachers expressed gratitude at the opportunity to challenge their teaching skills and put their knowledge, new or consolidated, into practice. This suggests that the opportunity to learn was considered a means of advancing teaching practice, and that the training was acceptable to dance teachers: I'm really grateful for the experience. It gave me time to reflect on actually what I was doing as a teacher (teacher 4) Learning about my child's development. Parents articulated a desire to learn more about their child and attributed significant worth to this information. The value of the pre-intervention assessment came across strongly, with parents finding this useful (for seeking further therapy), informative (for understanding how their child's motor skills compared with others of the same age) and often reassuring: I thought it was fantastic how [the research team] did the assessments for [participating child] before they started and then gave me the feedback. I've got that now as a baseline for things. Like that was really helpful . . ..and, just, I don't know how to explain this other than to have the peace of mind that they are being monitored in that way, um, and have this amazing, um, team available to them, watching their development, it gives me a lot of peace of mind (parent 2) The dance classes themselves also provided parents with an opportunity to learn about their child. For some parents, this was a very positive experience; watching their child gain confidence and independence, or learning that their child could do more than they had realized: look I think [my child] got a lot out of it but I think us as parents probably got more, just understanding what they can do and I guess pushing them to sort of um, pushing them a little bit to be more active and to do more stuff (parent 1) For other parents, being unable to watch classes meant they were unable to learn about their child in this environment, and parents felt it "a little bit frustrating" (parent 2) and that they would know more about their child's ability and development had they been able to watch.
Another frustration for families participating in the second program was missing the opportunity for the follow-up assessment: "we're not able to do the reassessment of them, which is a bit disappointing" (parent 3). Many parents cited their desire to know more about their child's development as an incentive for participating in the study. As such, in cases where parents were not allowed to observe dance classes or were not able to attend the post-intervention assessment, parents felt they gave up a benefit of the intervention. Both these examples indicate an opportunity cost of the intervention.

Dance teachers as architects of the learning environment
This final theme highlights the role of the teacher in creating the dance and learning environment. There was a feeling of empowerment and pride from teachers; they were confident in their abilities to improve motor skills (to teach dance steps) but also to create an inclusive and positive learning environment: "my focus for term one for all of my dancers was to get them moving freely and confidently within the space" (teacher 4). Furthermore, parents recognized the role teachers played in ensuring that they felt welcomed to class, and in helping families transition into a dance class environment and know what to expect: the school was very supportive and when I asked them which class should I enrol my child in they could pin point which teacher they thought would be best for this. I thought the school was right on top of it (parent 2) The sense of ownership and pride teachers took in creating their welcoming classroom environments speaks of the ethicality of the interventions for this group; with teachers willing and enthusiastic to provide an inclusive program.
Although teachers were asked to incorporate some new teaching strategies and to accept additional students into their classes, the dance classroom was still very much their domain, the study was "un-intrusive" (teacher 4): I think what [Dance PREEMIE] is doing is great because, you know, you could send out lesson plans and all of that stuff but I think that defeats the purpose of the study because, you know, I feel like the study is about the kids getting involved like any other kid and joining in activities that all the other kids get to do. So if it was a, you know, if the teachers were given strict instructions from [the research team], I think it would defeat the purpose (teacher 6) Teachers offered examples of how they were able to engage and adapt both dance steps and the classroom environment to promote involvement for all children in the class. These were stories of success, reflecting confidence and self-efficacy in their abilities as teachers, and pride in their role in the intervention (Supplementary Table 2). This sense of self-efficacy and ownership over their classroom environment was maintained when being observed by a member of the research team, or in challenging fluctuating classroom environments, with teachers taking responsibility for improving difficult situations: "To be honest the way we tried to adapt for this particular student it didn't quite work" (teacher 6). Furthermore, teachers were enthusiastic to take on a challenge, to be seen at their best and create a dance class environment that suited children in the study, as well as children already enrolled in their schools: I was really excited to be able to take on . . . I thought it would be a little bit different, a little bit more challenging I guess but I was really excited to see if I could you know, help impact this little person's life-so I was really looking forward to having them in class (teacher 1) The excitement expressed by teachers, and their view of involvement in the intervention as a welcome challenge suggests that the program was a good value fit for the teachers that enrolled in Dance PREEMIE.

Discussion
This study explored the acceptability of Dance PREEMIE to parents of participating children and dance teachers delivering the classes. Five themes were developed through reflexive thematic analysis, with findings supporting the acceptability of the study to both groups. Parents and teachers shared a motivation to promote a positive participation experience for each child, with an emphasis on fun, confidence, and development. Furthermore, both groups understood that child development takes time and effort, and thus felt that a longer intervention duration would have been beneficial. Throughout the intervention process, teachers maintained a sense of self-efficacy in teaching children with disabilities, suggesting that the level of support provided was acceptable, but importantly that the study did not intrude on their usual teaching practice. Encouragingly, both parents and teachers noted improvements across a range of developmental domains. In particular, parents noted a change in their child's approach to, or willingness to try, a range of physical activities. Throughout the study, both parents and teachers valued information, emphasizing the importance of good communication. Parents found the initial assessment particularly beneficial to their understanding of their child's development, highlighting the importance of ongoing follow-up for children born preterm for their families. Teachers valued the information provided in the physiotherapy-led training day, but felt more information on the participating children in their class would have been helpful.
Previous research has considered the acceptability of PA participation interventions for school age children with motor impairments, exploring the perspective of children, parents (Fernandez et al., 2018;Willis et al., 2018), and, less frequently, physiotherapists (Clutterbuck, Auld, and Johnston, 2020). Like Dance PREEMIE, these interventions aimed to improve PA participation in children's local communities, however most interventions were facilitated by physiotherapists (Clutterbuck, Auld, and Johnston, 2020;Willis et al., 2018). Dance PREEMIE is distinctive firstly, in targeting the intervention to preschool age children and secondly, in assessing acceptability from the perspective of community providers of PA. By prioritizing the acceptability of the study, we recognize the central role of the family as experts in their child's health and wellbeing (Palisano et al., 2012). Furthermore, as the goal of many PA interventions for children with disabilities is participation in the local community, understanding the community provider perspective is essential, especially as community attitudes and knowledge of disabilities are considered important facilitators of PA participation for children with disabilities (Wright, Roberts, Bowman, and Crettenden, 2019).
Our results support previous research in finding that adequate time was considered necessary for children to adapt to a new environment and allow for skill development (Fernandez et al., 2018;Willis et al., 2018). This suggests that intervention duration has implications for intervention effect as well as acceptability, providing support for longer-term participation-focused interventions for children with motor impairment. Furthermore, both parents and teachers felt skill development was facilitated through exposure to the dance class environment, reflecting research advocating for skill development to occur in context (Palisano et al., 2012). Moreover, knowledge of their child's capabilities in a participation-focused environment gave parents confidence to enroll their child in further PA activities mirroring Willis et al.'s (2019) finding that knowledge empowered families to take on a facilitator role in the local community.
The findings of this study support the assertion that PA participation opportunities should align with children's preferences and be adapted to allow each child to achieve (Palisano et al., 2012;Pollock, Missiuna, McKechnie, and Embrey, 2012). Integral to the placing the child center-stage theme were the concepts of fun, enjoyment, and child preference. Fun was identified as a key mechanism for participation through a realist evaluation of an intensive, participation-focused camp for children with disabilities (Willis et al., 2018). Likewise, enjoyment and having fun were identified by children and parents as a key motivational aspect of a circus intervention (Fernandez et al., 2018), while enjoyment was the most frequently mentioned positive outcome identified by parents participating in a sports transition program for children with CP (Clutterbuck, Auld, and Johnston, 2020). These findings highlight the value placed on child enjoyment by a range of stakeholders, not only as an outcome or marker of success, but also as a motivator for ongoing PA participation. Positive experiences with Dance PREEMIE were associated with improved motivation for PA participation in the theme Dance PREEMIE as a catalyst for participation. This reciprocal relationship between child preference and participation is theorized by the fPRC, in which a child's preferences are established through, and precursor to, participation (Imms et al., 2017). Similar to Dance PREEMIE, participation in a circus intervention created a 'flow on effect,' with children more willing to try new activities, engaging in PA more frequently in their home environments, and planning on continuing with circus or a different activity as a result of participating (Fernandez et al., 2018). Comparable results were found in a study evaluating adaptive dynamic cycling for children with CP; the experience of succeeding in, and enjoying cycling prompted children to engage more readily in PA opportunities (Pickering, Horrocks, Visser, and Todd, 2013). Furthermore, positive PA participation experiences have been identified by physiotherapists as a motivator for children with disabilities, underlining the importance of providing fun and enjoyable experiences for children with motor impairments to promote ongoing PA participation (Wright, Roberts, Bowman, and Crettenden, 2019).
As participation occurs at the intersection between an individual and their environment (Imms et al., 2017), it is essential to consider the dance class environment in which Dance PREEMIE took place. The fPRC theorizes reciprocal transactions between preference and sense of self (e.g. confidence) and the environment (e.g. the dance class) (Imms et al., 2017). In this study, these connections were discussed by parents and teachers, who associated a welcoming classroom environment with improved confidence and enjoyment, and in a negative sense when teachers observed that an upset child could influence the classroom environment, and negatively affect engagement. Most teachers in our study felt empowered and confident adapting their classes and creating an inclusive environment, suggesting an ecological approach was acceptable to teachers (Palisano et al., 2012). While it is likely that only teachers with an interest in inclusive dance enrolled in the study, the positive responses from teachers at the opportunity to learn about teaching children with disabilities (i.e. advancing my teaching practice) suggests that this information is sought and valued by the dance community. However, this wasn't the case for all schools and teachers. One teacher in this study felt that children should fit the class, suggesting children should not join until they were older or had learnt classroom behaviors. Other schools declined to be included in the study, although reasons were not recorded. The strong connection between confidence and environment experienced by families of children with disabilities (Willis et al., 2018) highlights the importance of appropriate recruitment strategies. Most dance teachers in this study displayed a strong sense of self-efficacy when interviewed post-intervention, suggesting that community providers of PA can be supported to feel confident in including children with motor impairment in their classes. This provides preliminary support to an alternative model of participation intervention, in which community providers are supported by physiotherapists in delivering interventions.
Previous PA participation interventions for children with motor impairment have been delivered in groups of children with similar disabilities, with both parents and children feeling that peer support was a positive aspect of these programs, and facilitated feeling included (Clutterbuck, Auld, and Johnston, 2020;Fernandez et al., 2018;Willis et al., 2018). Willis et al. (2018) explored this concept in depth, finding that for children with disabilities participation is facilitated by contexts that are emotionally safe, social, and accommodate support from, followed by independence from parents. In contrast to other intervention models, Dance PREEMIE was integrated into community dance classes open to the public, with one or two Dance PREEMIE participants allocated per class. This was an intentional decision, to use participation as a means in a real-world context (Imms et al., 2017). Although not a disability-specific group, parents found the dance classes welcoming, and this contributed to their perception of the study as successful for their child. While individual preference is likely to influence preference for groups composition, it is possible that PA participation interventions may be acceptable to families in both disability-specific and all abilities formats. It is also possible, however, that the age differences of the children participating may influence the value of a peer-support. Dance PREEMIE was directed at preschool age children, when children are not able to accurately compare their motor performance to other children (Pless, Carlsson, Sundelin, and Persson, 2001). As children get older, and they are more aware of their abilities, the opportunity to participate with other children with similar disabilities may become more important (Stodden et al., 2008). It is also essential to consider the role of the environment in creating a context where participation is facilitated (Imms et al., 2017). While a child can learn motor skills through participating, this relationship is bidirectional, and a level of activity competence may be required to successfully engage in a particular context, and thus participate (Imms et al., 2017;Wall, 2004). Our results suggest that the Dance PREEMIE model was acceptable for preschool age children, however, as children age and dance classes become more demanding, different approaches may need to be trialed (Wall, 2004). As yet, there is little information on whether or not a disability-specific model would be equally acceptable to families of preschool age children, and further research is required to determine if a there is a preferred approach.

Strengths and limitations
This study was able to gain the perspectives of parents and teachers through semi-structured interviews, however we chose not to interview the participating children due to their age. While parents are considered experts on their own children, it is possible that interviewing children may have provided additional insights into the study. Furthermore, member and participant checking, which was not completed for this study, may have improved credibility, and thus confidence in our results (Korstjens and Moser, 2018). While interviews were completed with most families, we were not able to interview the family that withdrew from the study. Likewise, the family unable to be contacted for an interview only attended a single class. These families may have provided valuable information on the acceptability of the intervention from their perspective, and may have offered insight into study burden, opportunity costs to participation, or how the intervention could be adapted or improved. It is also possible that parents and teachers did not openly discuss all the negative aspects of the study during the interview as all were aware that the study was part of KLC's PhD. For teachers, it may have been difficult to discuss negative opinions of the study training as this was delivered by KLC. For this reason, we may have missed valuable information which may have been useful to inform future interventions. Furthermore, the results of this paper only represent the views of the participants who engaged with the Dance PREEMIE feasibility study. Future iterations of Dance PREEMIE, whether in a different context or in a larger efficacy trial, should further evaluate the acceptability of the intervention, to provide further insight or different viewpoints. Strengths of this study include; use of a rigorous analytical process, including extensive reflexivity, and fully developed themes as indicated by Braun and Clarke (Braun and Clarke, 2006).

Conclusion
Dance PREEMIE provides insight into a participation intervention in which dance teachers are supported through physiotherapy-led training to include children with motor impairment in their preschool dance classes. The acceptability of this approach to both parents and teachers is encouraging, as this model facilitates 'real world' participation which is likely to be easily accessible to families on an ongoing basis. Further research is needed to explore the efficacy of Dance PREEMIE at improving ongoing PA participation, as well as its suitability for implementation on a broader scale.