Over the past decade, much attention has been paid to school-based recess and the implications of recess on child development. In 2013, the American Academy of Pediatrics released a policy statement citing the crucial role of recess within schools [1]. Authors of the policy statement presented evidence to suggest that recess provides cognitive and academic benefits, social and emotional benefits, and physical benefits to the child. More recently, the United States (U.S.) Centers for Disease Control (CDC) and SHAPE America [2] released evidence-based guidelines for recess strategies citing many of the same academic, social, emotional, and physical benefits recess can have on students. Within these guidelines, strategies to help facilitate positive outcomes included: (a) leadership decisions in which recess plans are developed, space is designated, and adults are properly trained; (b) communicating and enforcing behavioral and safety expectations such as communication and conflict resolution skills; (c) creating an environment that is supportive of physical activity by ensuring, among other things, that proper space and equipment is available; (d) engaging the school community to support recess; and (e) collecting data on the recess environment and potential outcomes that recess may affect (e.g., student and school outcomes). Thus, stakeholders have identified a high-quality recess as one that includes planning, structure, access to space and equipment, positive student behaviors, trained adult staff, and high levels of physical activity (PA). However, to date, there is a lack of assessment tools to measure the context of recess that might support positive child development beyond levels of physical activity.

In considering the possible benefits associated with recess, the majority of research findings, specifically within the public health domain, have focused on recess as instrumental to achieving PA related goals. Research findings within this domain have consistently shown that recess provides a needed contribution to children’s levels of PA during the school day [3–5]. Specifically, a RWJF report [5] stated that recess accounted for 42% of children’s opportunities to be physically active in school, while Erwin et al. [3], reported that recess accounted for up to 44% of step counts during the school day. These data should not be overlooked, as children spend approximately 40% of their waking hours at school [4] and 60% of school districts have no formal recess policy [6]. Moreover, only 22% of school districts in the U.S. require daily recess for elementary school students, with less than half of these requiring at least 20 min of recess per day. Given the PA benefits, recess has also been implicated as a potential contributor to children’s cognitive and academic development. This argument is predominantly made by those showing links between PA and health to cognitive and academic performance. For example, results from various studies and meta-analyses have highlighted that higher-fit children outperformed their lower-fit counterparts in laboratory measures of inhibition [7, 8], working memory [9], and overall cognitive functioning [10, 11]. Additional studies have also demonstrated body mass index and levels of inflammation are inversely related to various measures of cognitive performance [12, 13]. School-specific data have suggested that school-based PA interventions have the ability to positively affect academic performance [14, 15], with additional evidence to support a dose-response relationship [16]. Despite this, very limited evidence specific to increased PA through recess has shown a beneficial impact on academic well-being and classroom behavior [17, 18].

Aside from physical and cognitive benefits, it has been proposed that participation in play can help facilitate the development of social and emotional skills such as cooperative goal setting, teamwork, and emotional regulation [19]. Naturally, these ideas have been transposed into the recess environment. Proponents of these ideas have suggested that participation in physically active games during recess is positively associated with pro-social behaviors such as the ability to develop peer relationships, sharing, problem solving, and conflict resolution [20]. However, recent experimental data show a null effect on social skills competency following recess intervention programs [21, 22]. In considering prosocial behaviors and social skill development through recess, Left and colleagues [23] reported that the presence of organized games was predictive of higher levels of cooperative and intercultural play, and therefore might be an important mechanism for promoting social development through recess experiences. Thus, in considering the potential benefits of recess, the quality of the environment likely shapes how an individual experiences recess, and is likely to affect outcomes associated with participation in school-based recess.

In light of considering the contextual variables associated with recess, available data suggests that aside from potential benefits, recess is a place where violent and anti-social behavior can occur [24], and that some elementary school students view the playground as an unsafe space [25]. In a large survey of third through fifth grade students, Glew et al. [26] reported that the playground is the most likely place for bullying to occur at school, and that those involved (bullies, victims, bully-victims) had higher odds of feeling unsafe and sad at school. Moreover, victims and bully-victims had higher odds of low academic achievement and lack of belonging at school than bystanders. If the playground environment is one in which bullying and anti-social behaviors occur, it becomes difficult to make the claim that this environment inherently contributes to social, emotional, and cognitive health. Altogether, recess has been shown to promote PA, and PA has been linked to cognitive, social, emotional, and physical benefits; yet the context (environmental, social, etc) of recess likely plays a mediating role as to whether or not recess is beneficial to children’s health. Given this collective premise, and in line with recommendations from the CDC and SHAPE America noted above [2], a need exists to measure the recess environment beyond PA levels.

To date, PA remains the primary outcome associated with recess, while ignoring other indicators inherent in the strategies released by the CDC and SHAPE America [2]. There remains a specific need to better understand environmental factors associated with recess, how students interact and communicate during recess, how conflict is managed during recess, the overall safety of the environment, and the role adults play in this process. Limited studies have assessed observed behaviors such as teacher/student interactions and peer conflict [18], or social interactions at recess [27], however no validated instrument exists to examine the recess environment, student behavior, and levels of adult engagement. The most common observational instruments cited in the literature appear to be the system for observing children’s activity and relationships during play (SOCARP) [27] and the system for observing play and leisure activity in youth (SOPLAY) [28]. These tools, however, are reliant on time sampling techniques that observe one child at a time and likely miss out on larger environmental influences. Others have created observational and coding schemes specific to study aims [18, 21, 23], however these do not allow for scalable research efforts. The purpose of the current study was to develop a valid, and reliable, assessment tool intended for use in measurement of the contextual factors and behaviors associated with recess. Within this, there was a specific focus on safety, resources, student engagement, adult engagement, pro-social/anti-social behavior, and student empowerment.

An iterative and multi-step approach was taken to meet the purpose of the current study. Prior to formal data collection, item development took place over an extended period of time using multiple expert working groups. This process resulted in Great Recess Framework -Observational Tool (GRF-OT) to measure the context in which recess takes place and the behaviors that manifest within this context. The measurement model of the GRF-OT was then tested, followed by reliability and stability testing of the tool. The methods and procedures used to accomplish these goals are described below.

The purpose of the current study was to develop a valid, and reliable, assessment tool intended for use in measurement of the contextual factors associated with recess, and the behaviors that manifest within this context. Within this, were was a specific focus on safety, resources, student engagement, adult engagement, pro-social/anti-social behavior, and student empowerment. An evidence-informed, and psychometrically-sound, tool termed The Great Recess Framework – Observational Tool (GRF-OT) resulted from the data collection and analyses described in this paper.

Item development for the GRF-OT resulted in 17 items that each describe in short detail critical aspects of the recess environment with the particular area of focus noted above. Furthermore, we have provided reliability data for all initial items in the event school districts, practitioners, or evaluators find these items as relevant to their own evaluation efforts. Response formats for all items were according to a 1 (low quality) to 4 (high quality) rating of the particular focus and included succinct and distinct descriptions to anchor each possible score on the specific item. Subsequent thorough data collection and analyses revealed a four-factor measurement model and established measurement validity for the subscales including (1) structure and safety; (2) adult engagement and supervision; (3) student behaviors; and (4) transitions. Convergent validity was demonstrated in the significant associations of observed activity levels and the subscales of ‘structure and safety’, ‘adult engagement and supervision’, and ‘student behaviors’. Adequate levels of inter-rater reliability were found for all items, each of the four subscales and the entire GRF-OT. These findings create evidence of strong levels of consistency and agreement among raters across all items, scales and for the total tool. Last, examination of the stability of the GRF-OT revealed results that suggest three-day average of the scores indicate higher levels than two-day averages. Ultimately, these findings indicate that evaluation of a minimum of three days of recess best describe, in terms of stability, the features a specific recess environment.

Previous research shows school-based recess provides significant opportunity for, and accrual of, PA among children; thus contributing to physical health benefits [3–5]. Given this, it is not surprising that evaluations of recess have relied on examining self-reported, observed, or objective PA levels as a proxy measure for recess quality. Notably, to date, ‘success’ at recess has been a measure of how moderate-to-vigorously active children are during this time period, with the assumption that this level of activity can yield physical, cognitive, social, and emotional health benefits beyond the playground. Yet, if children are active in an unsafe environment that is prone to bullying, fighting, and lack of adult engagement, it would be naïve to assume this time period contributes to the holistic development of children, as recent policy articles might suggest [2]. A more holistic form of evaluation of the recess environment is therefore warranted. Furthermore, the acknowledgment that recess has implications for social, emotional, and cognitive health [2], should prompt the need for evaluation that considers contextual factors, student behaviors, and adult interactions during recess. Previous research has shown that access to equipment [37], levels of cooperative play [23], adult engagement and interactions [18], and conflict resolution skills [20, 29] are important to promoting a quality recess. While currently available assessment tools that examine child-level interactions are limited by time sampling methods that hone in on specifically targeted children, the GRF-OT provides researchers and practitioners with an opportunity to evaluate the overall environment and the nature of interactions that take place within it. As school-based recess remains at the forefront of policy discussions and decisions, there is a need to consider the quality of this environment for children in schools. The initiative forwarded by the CDC and SHAPE America [2] to develop evidence-based strategies for recess in schools should be assessed in a consistent manner inclusive of observed behaviors such as teacher/student interactions, peer conflict and safety at recess. The development of the GRF-OT is an important step in filling this evaluation gap currently identified within the extant literature.

As efforts to understand the impact of school-based recess move beyond a focus on the PA-related aspects, children’s social, and emotional, development have emerged as relevant outcomes. Findings from extant research consistently suggest that the recess environment is conducive to facilitating positive growth in children’s emotional control, teamwork, cooperation, goal-setting, peer relationships, sharing, problem solving and conflict resolution [19, 20]. Yet for the benefits of recess to be fully realized contextual variables, student behavior, and adult interactions must be considered alongside PA in the evaluation of future research.