Numerous physiological health and psychological well-being benefits associated with regular physical activity have been documented [1, 2]. Current United Kingdom (UK) and international physical activity guidelines recommend that children undertake health-enhancing moderate-to-vigorous physical activity (MVPA) for at least 60 minutes over the course of each day [3, 4]. Despite this, low levels of children's physical activity are commonly reported, with a recent large scale study observing that only 5.1% of boys and 0.4% of girls met current recommendations when measured using accelerometry [5].

Although the prevalence of childhood obesity is thought to have 'levelled-off' in recent years, previous stable phases have been followed by further increases, and the current prevalence of obesity remains extremely high [6]. Reversing the prevalence of childhood overweight and obesity, therefore, is still an important public health priority, since childhood obesity tracks through adolescence [7] and into adulthood [8], and increases the risk of adult premature mortality [9]. Numerous strategies and school-based interventions to tackle obesity through enhanced physical activity have been implemented, though few studies have demonstrated sustained behavioural change (i.e., more than one year) or positive impacts on children's health and well-being [10].

Despite large-scale quantitative studies being able to assess the direction and strength of trends in participation of physical activity, they are unable to explain the reasons why children and significant others (i.e., parents and teachers) maintain or cease to participate in life-long physical activity [11]. Intervention and strategy development, therefore, have largely overlooked the views of potential participants [12] even though, according to Potvin et al. [10], the need to consult and engage intervention participants (e.g., children, parents, and teachers) within the context of their community has been advocated for some time. Furthermore, a comprehensive understanding of the perceived benefits and barriers to physical activity, afforded by qualitative research, is deemed imperative in the design of successful interventions [11, 13]. Such approaches have effectively informed previous physical activity interventions (e.g., [14–16]). Although focus group studies have examined physical activity correlates in children, there is a paucity of research directly comparing the views of children, parents and teachers on the issues relevant to any proposed intervention [17]. There is therefore a need to use informed methods based on extracting such views to design and develop a school-based physical activity intervention, which aim to encourage health-promoting behaviour change.

Behaviour change is complex to both achieve and maintain. In order to develop a successful physical activity-based intervention, an appropriate conceptual health promotion model should be utilised to prioritise the key assets of the target group [13]. One such model is PRECEDE-PROCEED [18], which provides the target population with a comprehensive and structured assessment of their own needs and barriers to a healthy lifestyle. When applied to a tailored intervention programme it is suggested that this approach promotes successful and sustained participant compliance to the intervention protocol [19]. Effective physical activity promotion strategies and interventions are based on known correlates of youth physical activity [20, 21], and increases in physical activity have been linked to a range of social, behavioural, physical and social environmental correlates [22]. Inter-relationships between these correlates have been proposed the Youth Physical Activity Promotion Model (YPAPM) [23], which is based on the PRECEDE-PROCEED health promotion model [18]. This hierarchical model is specifically relevant to children's physical activity, and has been previously used in correlates research [24]. The model is underpinned by four categories of correlates termed personal demographic, predisposing, enabling, and reinforcing factors.

Whilst research has generally documented children's physical activity levels, there are less comprehensive data examining underlying reasons and choices for different behaviours. Moreover, research into antecedents and determinants of regular physical activity has predominantly used quantitative methods to identify cross-sectional views to individual's knowledge, attitudes and beliefs towards physical activity in predetermined categories [20]. The aims of this study were to (i) elicit the views of primary (called elementary internationally) school children aged 9-10 years old, their parents, and teachers in relation to their own knowledge, behaviours and perceptions towards childhood physical activity, and to examine perceived benefits and barriers to participation; and (ii) use these data to subsequently inform the design of a tailored physical activity intervention programme, CHANGE! (Children's Health, Activity, and Nutrition: Get Educated!).

The first aim of this study was to elicit the views of primary school children, their parents, and teachers in relation to their knowledge, behaviours, and perceptions towards physical activity, and to examine the perceived benefits and barriers to participation. This builds on previous research by using a new qualitative technique to inform the development of an intervention that will largely be delivered through schools, but will require family support to deliver on the objectives. The use of an emerging qualitative methodology enabled a comprehensive review of a large data set in conjunction with an established theoretical model. Pen profiles allowed a 'reader-friendly' representation of a quantitatively based analysis procedure therefore eliminating the likelihood of data, and hence key emergent themes, being skewed by dominating participants, whose views may be of the minority. This analysis technique therefore advances previous qualitative research studies by providing a basis of organising and representing key emergent themes. The second aim of the study was to use these formative data to inform the design of a tailored population-based physical activity intervention programme with the aim of enabling primary school children to develop healthy physical activity behaviours and make more informed lifestyle choices.

The data revealed a range of health knowledge in children and adults [17, 37], but also identified lack of physical activity identification, which is contradictory to previous research [12]. Participants had a good understanding of the relationship between physical activity and health, although contrary to a previous study [12], some children demonstrated a limited understanding of what constitutes physical activity. Parents' views indicated that they understood the importance of their child being physically active, regardless of SES, which is consistent with more recent research [17]. Despite high levels of child and parent knowledge about the importance of physical activity engagement, this knowledge did not appear to always translate into actual physical activity behaviours. These results suggest that enhancing family based education on what constitutes physical activity, and how it can be incorporated into familial daily lifestyles, should be the focus of tailored interventions.

Fun, enjoyment, and social support were important predictors of physical activity participation and non-participation. Children see enjoyment and peer interaction as reasons to be physically active [11], prompting the development of interventions that maximise the fun and enjoyable aspects of physical activity [14]. Children from lower SES schools demonstrated autonomy over their physical activity through activities such as devising 'made up' games. When autonomy is conceptualised as choice as in Self-Determination Theory [38], then increased choice in active behaviours might have been expected to be demonstrated more by children from the higher SES schools. These children may, theoretically, have more opportunities for participation related to parental income, leisure time, and the value placed on active lifestyles [8]. Nonetheless, it could be drawn from these data that the lower socioeconomic areas are linked with more opportunity due to unsupervised play; parents from higher SES backgrounds advocated more sense of accompanying their children to the park, for example. Further, it may also be proposed that those children with lesser access to organized physical activity may have to rely on their imagination to devise games. Interestingly, teachers in a high SES school suggested that lack of structure during school playtime (i.e., recess) compromised participation in physical activity, contrary to reported literature conveying that the interventionist approach may have limited effects on physical activity and play behaviour [39, 40]. Perhaps this reinforces why low SES school children expressed a sense of choice at playtime. Given that children with a sense of autonomy participated in regular physical activity, children's physical activity could be facilitated with a greater choice and variety of activities/opportunities [41, 42]; thus, part of the intervention could provide suggestions for inexpensive and fun activities to do alongside family members.

Children reported participating in a variety of structured sports, such as organized football and swimming lessons. This supports previous research [12, 17, 43], though several barriers to physical activity engagement were also identified across all group interviews. The barriers elicited by children and parents were generally consistent with those presented in previous studies [17, 44, 45], with parents perceived by the children to be the biggest barriers to their physical activity participation (37%), regardless of SES or gender. Teachers also conveyed experiences of parents acting as barriers to their children's health and physical activity participation. For both parents and teachers, safety concerns were a significant perceived barrier to children's physical activity participation, particularly in relation to adverse weather and proximity of activity to busy roads, both of which were associated with restrictions on children's play [46–48]. While some children and adults reported weather preventing them from engaging in physical activity, supporting previous research [49, 50], others noted that weather was not only perceived as less restrictive, but provided extra opportunities for participation. For example, both children and parents conveyed that snow provided opportunity for family physical activity:

"cause it snowed a lot over the winter, me and my friend [name], we were playing out like every single day making snow forts and having snowball fights." (B15)

However, it is noteworthy that snow in this north-west England Borough is infrequent; therefore it is more likely that it is the novelty which increases physical activity. Children and parents both identified that high levels of sedentary screen time (i.e., television and video-games) negatively impacted on physical activity [17], suggesting that the range of sedentary behaviours available may be more reinforcing than physical activity even when physically active alternatives are available [51]. Nonetheless, teachers did not advocate the negative association between screen time and physical activity, perhaps because they associate it with positive learning outcomes. Other barriers identified mainly by lower SES families included lack of money and transportation, both of which are consistent with previous research [12, 17].

Despite similarities between enabling factors identified by adults and children, parents in particular perceived holidays as an opportunity for family based physical activity, perhaps as a result of overcoming time barriers associated with work and school commitments [37], thus allowing focus on leisure. Children, parents and teachers all reported that peers as well as families were major influences on children's physical activity participation [12, 52], and dog ownership often led to increased frequencies of family walks [53]. Parental influences were thought to operate primarily through providing support and encouragement [11, 54], but also through role modelling and providing opportunities for activity, which together influence children's learning, how children respond to the external environment, and what children expect of themselves [55]. Peer influences were seen as supportive by children, but as role models by teachers. Paradoxically, parents were both significant barriers (i.e., 'grounding') and enablers (i.e., encouraging) to children's physical activity participation, indicating that parents effectively have the greatest influence over their children's involvement in physical activity with the ability to both facilitate and impede participation [56]. Families, therefore, play a powerful and important role in promoting health-enhancing behaviours, thus involving parents and the whole family appears fundamental to approaches attempting to increase children's physical activity levels. Moreover, this approach should help overcome any potential conflicting messages between school and home-life.

In agreement with Power et al. [17] parents and teachers believed that schools were influential contexts for children's physical activity participation by offering various structured and unstructured opportunities for physically active pursuits. It is therefore important that the key features of the intervention are structured around both parents and schools. Further, within the intervention children need to receive support from teachers and parents in order to increase their perceptions of competence, self-efficacy and enjoyment [57].

The use of comprehensive formative research enabled depth of data to be gathered in a relatively short period of time. These findings will specifically be used to devise and implement an intervention for this population. A major strength of the study is not only supporting new methodologies within qualitative research, but advancing previous research utilising pen profiles [31] through the use of triangulating data between groups (i.e., children, parents and teachers). This research advances previous qualitative formative studies through the use of a large sample size. Other methodological strengths are the inclusion of both high and low socioeconomic backgrounds and the triangulation consensus of data between authors providing credibility, transferability, and dependability. Indeed, group interviews with children allowed an insight into their thoughts, beliefs and experiences towards physical activity, respecting the expert knowledge of the participant [58]. Moreover, triangulation between children's and parents and teachers decreased the risk of misinterpreted views and therefore potentially inaccurate data. There may be a risk that the data were influenced by sampling bias, though it is noteworthy that the majority of children (63%) in every school consented to take part.

Group interviews revealed consistent themes between the socioeconomic groups, and gender for knowledge, behaviours, and perceptions towards physical activity. Aspects of the intervention can be modified depending on local need and resources, based on these findings. The results of this formative research will be used to inform the content and delivery of the physical activity component of the CHANGE! (Children's Health, Activity, and Nutrition: Get Educated!) health education intervention.