An edited version of this article was published in the March/April 1998 issue ofEarly Childhood News magazine
Children's Outdoor Play & Learning Environments: Returning to Nature
By Randy White & Vicki Stoecklin
It is unfortunate that children can't design their outdoor play environments. Research on children's preferences shows that if children had the design skills to do so, their creations would be completely different from the areas called playgrounds that most adults design for them. Outdoor spaces designed by children would not only be fully naturalized with plants, trees, flowers, water, dirt, sand, mud, animals and insects, but also would be rich with a wide variety of play opportunities of every imaginable type. If children could design their outdoor play spaces, they would be rich developmentally appropriate learning environments where children would want to stay all day.
Playground Paradigm Paralysis
We are all creatures of our experience, and our common experiences usually shape the conventional wisdom, or paradigms, by which we operate. When most adults were children, playgrounds were asphalt areas with gross motor play equipment such as swings, jungle gyms and slides where they went for recess. Most adults see this as their model for a children's playground.
So when it comes time to plan and design a playground, the paradigm is to search through the catalogues of playground equipment, pick a piece or two that looks good to the adult and place it in an outdoor space which resembles their childhood memories of playgrounds. This is easy and doesn't take a whole lot of effort. Then once or twice a day, teachers let children go outside for a recess from their classroom activities to play on the equipment.
Today, fortunately, most playground equipment is becoming much safer than when adults grew up. National standards encourage the installation of safety fall surfaces and ADA is making the equipment more accessible. However, limiting outdoor playgrounds to gross motor activities and manufactured equipment falls way short of the potential of outdoor areas to be rich play and learning environments for children. This playground design paradigm paralysis also denies children their birthright to experience the entire natural outdoors which includes vegetation, animals, insects water and sand, not just the sun and air that manufactured playgrounds offer.
It is a well accepted principal in early childhood education that children learn best through free play and discovery. Children's free play is a complex concept that eludes precise definition, but children's play typically is pleasurable, self-motivated, imaginative, non-goal directed, spontaneous, active, and free of adult-imposed rules1,2. Quality play involves the whole child: gross motor, fine motor, senses, emotion, intellect, individual growth and social interaction.3
Childhood of Imprisonment
The world once offered thousands of delights of free play to children. Children used to have access to the world at large, whether it was the sidewalks, streets, alleys, vacant lots and parks of the inner city or the fields, forests, streams and yards of suburbia and the rural countryside. Children could play, explore and interact with the natural world with little or no restriction or supervision.
The lives of children today are much more structured and supervised, with few opportunities for free play. Their physical boundaries have shrunk.4 A number of factors have led to this. Parents are afraid for their children's safety when they leave the house alone; many children are no longer free to roam their neighborhoods or even their own yards unless accompanied by adults. Some working families can't supervise their children after school, giving rise to latchkey children who stay indoors or attend supervised after-school activities. Furthermore, children's lives have become structured and scheduled by adults, who hold the mistaken belief that this sport or that lesson will make their children more successful as adults.
Children have little time for free play any more. And when children do have free time, it's often spent inside in front of the television or computers. For some children, that's because their neighborhood, apartment complex or house has no outdoor play spaces. With budgets for city and state governments slashed, public parks and outdoor playgrounds have deteriorated and been abandoned. Children's opportunities to interact in a naturalized outdoor setting is greatly diminished today.
Childhood and outdoor play are no longer synonymous. Today, many children live what one play authority has referred to as a childhood of imprisonment.5 Child care facility playgrounds are often the only outdoor activities that many young children experience anymore.
Our company first became interested in the opportunities that outdoor play offers children's development when, in 1993, we conducted extensive focus group research with children and parents for a children's center we were designing. We were fascinated when the research consistently showed that children had a strong preference to play outdoors in natural landscapes, and that parents generally supported this kind of play.
Biophilia: The Love of Outdoors
Two new disciplines, eco-psychology6 and evolutionary psychology, are now suggesting that humans are genetically programmed by evolution with an affinity for the natural outdoors. Evolutionary psychologists use the term biophilia7 to refer to this innate, hereditary emotional attraction of humans to nature and other living organisms. Biophilia is the biologically based human need to affiliate with nature and the genetic basis for human's positive responses to nature.8 Why? Researchers say that for more than 99 percent of human history, people lived in hunter-gatherer bands totally and intimately involved in nature. So in relative terms, urban societies have existed for scarcely more than a blink of time.9 Our original nature-based evolutionary genetic coding and instincts are still an essential part of us and continue to shape our behavior and responses to nature.10
Well over 100 studies of outdoor experiences in the wilderness and natural areas show that natural outdoor environments produce positive physiological and psychological responses in humans, including reduced stress and a general feeling of well-being.11,12 It is also a clear-cut finding that people, and especially young children who have not yet adapted to the man-made world, consistently prefer the natural landscape to built environments. Children's instinctive feelings of continuity with nature are demonstrated by the attraction children have for fairy tales set in nature and populated with animal characters.13 Additional anecdotal evidence is that more children and adults visit zoos and aquariums than attend all major professional sports combined.14
Biophobia: The Aversion to Nature
However, if this human natural attraction to nature is not given opportunities to be exercised and flourish during the early years of life, the opposite, biophobia, an aversion to nature, may develop. Biophobia ranges from discomfort in natural places to active scorn for whatever is not man-made, managed or air conditioned. Biophobia is also manifest in the tendency to regard nature as nothing more than a disposable resource.15
Environmental education needs to start at any early age with hands-on experience with nature.16 There is considerable evidence that concern for the environment is based on an affection for nature that only develops with autonomous, unmediated contact with it. In their early years, children's developmental tendency towards empathy with the natural world needs to be supported with free access to an area of limited size over an extended period of time. It is only by intimately knowing the wonder of nature's complexity in a particular place that leads to a full appreciation of the immense beauty of the planet as a whole. In todays society, environmental education requires that in schools, children have regular personal interaction with as diverse a natural setting as possible.17,18
The Importance of Nature to Children
Studies have provided convincing evidence that the way people feel in pleasing natural environments improves recall of information, creative problem solving, and creativity.19 Early experiences with the natural world have been positively linked with the development of imagination and the sense of wonder.20 Wonder is important as it a motivator for life long learning.21 There is also strong evidence that young children respond more positively to experiences in the outdoors than adults as they have not yet adapted to unnatural, man-made, indoor environments.
The natural world is essential to the emotional health of children.22 Just as children need positive adult contact and a sense of connection to the wider human community, they also need positive contact with nature and the chance for solitude and the sense of wonder that nature offers.23 When children play in nature they are more likely to have positive feelings about each other and their surroundings.24
Outdoor environments are also important to children's development of independence and autonomy. Outdoor space allows children to gradually experiment with increasing distance from their caretaker. While the development of greater independence from toddlerhood to middle childhood can happen within the confines of indoor spaces, safe space outdoors greatly adds to the ability of children to naturally experiment with independence and separation, and the adult's willingness to trust the child's competence which is essential for separation to happen. This is particularly important for children who live in small and crowded homes.25
Children's Experience with the Natural World
Children's outdoor play is different from time spent indoors. The sensory experiences are different, and different standards of play apply. Activities which may be frowned on indoors can be safely tolerated outdoors. Children have greater freedom not only to run and shout, but also to interact with and manipulate the environment. Children are free to do 'messy' activities outdoors that won't be tolerated indoors.
Natural outdoor environments have three qualities that are unique and appealing to children as play environments - their unending diversity; the fact that they are not created by adults; and their feeling of timelessness - the landscapes, trees, rivers described in fairy tales and myths still exist today.26
Children experience the natural environment differently than adults. Adults typically see nature as background for what they are doing. Children experience nature, not as background for events, but rather as a stimulator and experiential component of their activities.27 The world of nature is not a scene or even a landscape. Nature for the child is sheer sensory experience.28 Children judge the natural setting not by its aesthetics, but rather by how they can interact with the environment.
Children have a unique, direct and experiential way of knowing the natural world as a place of beauty, mystery and wonder. Children's special affinity for the natural environment is connected to the child's development and his or her way of knowing.29
Plants, together with soil, sand, and water, provide settings that can be manipulated. You can build a trench in the sand and dirt or a rock dam over a stream, but there's not much you can do to a jungle gym except climb, hang, or fall off. Natural elements provide for open-ended play that emphasize unstructured creative exploration with diverse materials. The high levels of complexity and variety nature offers invites longer and more complex play. Because of their interactive properties, plants stimulate discovery, dramatic pretend play, and imagination. Plants speak to all of the senses, so it's not surprising that children are closely attuned to environments with vegetation. Plants, in a pleasant environment with a mix of sun, shade, color, texture, fragrance, and softness of enclosure also encourage a sense of peacefulness.30 Natural settings offer qualities of openness, diversity, manipulation, exploration, anonymity and wildness.31
All the manufactured equipment and all the indoor instructional materials produced by the best educators in the world cannot substitute for the primary experience of hands-on engagement with nature. They cannot replace the sensory moment where a child's attention is captured by the phenomena and materials of nature: the dappled sparkle of sunlight through leaves, the sound and motion of plants in the wind, the sight of butterflies or a colony of ants, the imaginative worlds of a square yard of dirt or sand, the endless sensory experience of water, the infinite space in an iris flower.32
Designing Outdoor Spaces for Children
The goal of designing children's outdoor environments is to use the landscape and vegetation as the play setting and nature as much as possible as the play materials.33 The natural environment needs to read as a children's place; as a world separate from adults that responds to a child's own sense of place and time.
Our company calls well designed outdoor children's play spaces discovery play gardens to differential them from the current design paradigm for children's playgrounds. Some authorities call them naturalized outdoor classrooms or naturalized playgrounds.
There is a sense of wildness about an discovery play garden. Conventional play design focuses on manufactured and tightly designed play equipment. Conversely in a discovery play garden, although there may be some conventional play equipment, many of the spaces are informal and naturalistic so they will stimulate high quality free play and discovery learning.
Children's idea of beauty is wild rather than ordered.34 A discovery play garden that plans for wildness, and provides openness, diversity, and opportunities for manipulation, exploration and experimentation, allows children to become totally immersed in play.35 Children's discovery play gardens are very different than landscaped areas designed for adults, who prefer manicured lawns and tidy, neat, orderly uncluttered landscapes. Discovery play gardens are much looser in design because children value unmanicured places and the adventure and mystery of hiding places and wild, spacious, uneven areas broken by clusters of plants.36
Physical attractiveness and innovativeness are not what is important for quality outdoor play space design. Children need tools, open space, challenge and opportunities to control and manipulate the environment. Suransky calls this "history making power"37 - the power for the child to imprint themselves upon the landscape, endow the landscape with significance and experience their own actions as transforming the environment.38
Outdoor play requires a lot of gear to make a go of it. Loose parts, sand, water, manipulatives, props and naturally found objects are essential tools for children's play. Loose parts have infinite play possibilities, and their total lack of structure and script allows children to make of them whatever their imaginations desire. Simon Nicholson first offered the theory of loose parts in children's play when he wrote in 1971, "In any environment, both the degree of inventiveness and creativity and the possibilities of discovery are directly proportional to the number and kind of variables in it."39
Through children's handling, manipulation and physical interaction with materials and the natural environment, they learn the rules and principles that make the world operate.
Outdoor play areas should flow from one area to the next, be as open-ended and simple as possible, encourage children to use their imaginations, have continuity and be perceived by the children as children's, not adult, spaces. They should be designed to stimulate children's senses and to nurture the child's curiosity, allow for interaction with other children, with adults and with the resources in the play space.
It is also desirable to integrate the outdoors with the indoor classroom with one sense of place and identity, so the transition between the two will be almost seamless. Design that allows children to go freely back and forth between inside and outside encourages children to experiment with autonomy from adults, both physically and symbolically.40 It also allows the outdoor space to become part of the classroom, rather than just a retreat from it.
Things children like in their outdoor environments include:
- vegetation, including trees, bushes, flowers and long grasses,
- animals, creatures in ponds, and other living things
- sand, best if it can be mixed with water
- natural color, diversity and change
- places and features to sit in, on, under, lean against, and provide shelter and shade
- different levels and nooks and crannies, places that offer privacy and views
- structures, equipment and materials that can be changed, actually or in their imaginations, including plentiful loose parts.
The structures and equipment do not all need to be manufactured. As much as possible, they should be made of natural materials such as logs, stumps and boulders and use the landscape in natural ways with berms and mounds.
Outdoor areas lend themselves to meeting children's individual needs. Natural environments allow for investigation and discovery by children with different learning styles.41 Using universal design principals, play areas and events can be designed as accessible to children with special needs without accessibility features being obvious.
Plants are vital. In fact, the identity of many of the play areas can be created through ecological theming with vegetation. For example, an interactive water play can be set in a bog or stream habitat. It is also important to incorporate ecological areas that utilize indigenous vegetation and settings so children can experience, learn about and develop an appreciation of their local environment.
Naturalized outdoor play spaces are rich learning environments for all age children. They contain a hidden curriculum that speaks to children through their special way of knowing nature. Every learning center and activity that can be created in the indoor classroom can be created in the outdoors. Specialized areas can even be designed to meet the developmental needs of infants and toddlers.
Discovery play gardens do not cost more to build than conventional playgrounds. Rather than spend most of the budget on conventional manufactured playground equipment, moneys are shifted to landscaping and creating play areas using natural materials. Discovery play gardens do, however, require specialized design skills to create a holistic and integrated child's world. To accomplish this, a much higher percentage of the budget must be allocated for professional design services than with a dominantly equipment-based playground.
Participatory design - having children, teachers, parents and maintenance staff participate in the design process - is essential to the success of any discovery play garden. Children's input assures that they will feel it is a special place for them. Teachers input is needed so they will take ownership of the discovery play garden as an outdoor classroom and utilize it to support their curriculum goals. Parents need to be involved so they will be supportive of the concept and learn how the naturalized space and often messy play greatly supports their children's development. Maintenance staff need to participate to assure that they will support the space and provide the maintenance required. User participation in the design process also helps to assure that the design will be culturally respectful.
Discovery play gardens offer children chances to manipulate the environment and explore, to wonder and experiment, to pretend, to understand themselves, and to interact with nature, animals and interesting insects and with other children. They are environments that encourage children's rich and complex play and greatly expand the learning opportunities of just conventional playgrounds. Children's discovery play gardens are places where children can reclaim the magic that is their birthright - the ability to learn in a natural environment through exploration, discovery and the power of their own imaginations.
1 Hughes, Fergus P., Children Play & Development, Massachusetts, Allyn & Bacon, 1991.
2 Frost, Joe L. & Paul J. Jacobs, "Play Deprivation, A Factor in Juvenile Violence", Dimensions of Early Childhood, v23, n3, 1995, Southern Early Childhood Association.
3 Haas, Malka, "Children In The Junkyard", Childhood Education, v72, n6, 1996, Association for Childhood Education International, Wheaton, MD.
4 Francis, Mark (interview), au Kathryn Devereaux, "Children of Nature", U. C. Davis Magazine, v9, n2, 1991, University of California, Davis.
5 Francis, 1991, op. cit.
6 Roszak, T, M. E. Gomes & A. D. Kanner, Ecopsychology, San Francisco, Sierra Club Book, 1995.
7 Wilson, Edward O., Biophilia, Massachusetts, Harvard University Press, 1984.
8 Wilson, Edward O., "Biophilia and the Conservative Ethic", The Biophilia Hypothesis, Kellert & Wilson (eds), Washington D.C., Island Press, 1993.
9 Nelson, R., "Searching for the Lost Arrow: Physical and Spiritual Ecology in the Hunter's World", The Biophilia Hypothesis, Kellert & Wilson (eds), Washington D.C., Island Press, 1993.
10 Lewis, Charles A., Green Nature, Human Nature: The Meaning of Plants in Our Lives, Chicago, University of Illinois Press, 1996.
11 Lewis, 1996, op. cit.
12 Lewis, 1996, op. cit.
13 Barrows, Anita, "The Ecopsychology of Child Development", Ecopsychology, Roszak, Gomes & Kanner (eds), San Francisco, Sierra Club Books, 1995.
15 Wilson, Ruth A., Ph.D., "The Wonders of Nature - Honoring Children's Ways of Knowing", Early Childhood News, March/April 1997.
16 Moore, Robin C. & Herb H. Hong, Natural Learning: Creating Environments for Rediscovering Nature's Way of Teaching, Berkeley, California, MIG Communications, 1997.
17 Sobel, David, Beyond Ecophobia: Reclaiming the Heart in Natural Education, Great Barrington, MA, The Orion Society, 1996
18 Hart, Roger. A., Children's Participation: The Theory and Practice of Involving Young Children in Community Development and Environmental Care, London, UK, Earthscan Publications, 1997.
19 Ulrich, R. S., "Biophilia, Biophobia, and Natural Landscapes", The Biophilia Hypothesis, Kellert & Wilson (eds), Washington D.C., Island Press, 1993.
20 Cobb, E., The Ecology of Imagination in Childhood, New York, Columbia University Press, 1977.
21 Wilson, 1997, op. cit.
22 Louv, Richard, Childhood's Future, New York, Doubleday, 1991. (Louv coined the term "Bogeyman Syndrome" to describe parents' paranoia of their children being abducted, kidnapped or physically harmed in the outdoors and public places.)
23 Louv, 1991, op. cit.
24 Moore, Robin, "Compact Nature: The Role of Playing and Learning Gardens on Children's Lives", Journal of Therapeutic Horticulture, Vol. VIII, 1996.
25 Bartlett, Sheridan, "Access to Outdoor Play and Its Implications for Healthy Attachments", unpublished article, Putney, Vermont, 1996.
26 Prescott, E., "Environment as Organizer in Child-Care Settings", Spaces For Children: The Built Environment and Child Development, C. S. Weinstein & I. G. David (eds), New York, Plenum, 1987.
27 Sebba, Rachel, "The Landscape of Childhood, The Reflection of Childhood's Environment in Adult Memories and in Children's Attitudes", Environment and Behavior, v23, n4, p 395-422, 1991.
28 Sebba, 1991, op. cit.
29 Wilson, 1997, op. cit.
30 Moore, Robin C., Plants for Play: A Plant Guide for Children's Outdoor Environments, Berkeley, CA, MIG Communications, 1993.
31 Wilson, 1997, op. cit.
32 Moore, 1996, op. cit.
33 Francis, Mark (interview), au Shell, Ellen Ruppel, "Kids Don't Need Equipment, They Need Opportunity", Smithsonian Magazine, v25, n4, p78-87, July 1994.
34 Francis, 1991, op. cit.
35 Moore, Robin C., Childhood's Domain: Play and Place in Child Development, Berkeley, CA, MIG Communications, 1986.
36 Francis, 1991, op. cit.
37 Suransky, V. R., The Erosion of Childhood, Chicago, University of Chicago Press, 1982.
38 Haas, 1996, op. cit.
39 Nicholson, S., "How Not To Cheat Children: The Theory of Loose Parts", Landscape Architecture, v62, p30-35, 1971.
40 Bartlett, 1996, op. cit.
41 Moore, 1997, op. cit
Of the 33 included publications, 16 reported on experimental studies and 17 on observational studies. Their study characteristics are summarized in Table 1 and Table 2, respectively.
Characteristics of experimental studies (n = 13) and effects of (pre)school playgrounds on children’s health
Characteristics of observational studies (n = 17) and associations of (pre)school playgrounds with children’s health
The 16 publications reporting on experimental studies were based on 13 studies, since some publications were based on the same study sample. This was true for three publications of Ridgers and colleagues, in which the effects of incentives to change playgrounds on physical activity during recess were reported for different follow-up periods, i.e., six weeks, six months, and 12 months [28-30]. Further, Stratton and colleagues reported twice on the effects of painting markings on playgrounds of two early primary schools; once in 2000 and once again in a more recent publication in 2005. In the most recent publication, the sample size had increased through the inclusion of two additional late primary schools [31,32].
Seven of the 13 experimental studies (54%) included a control group in the study design, and only three of them used a randomization procedure to allocate schools and/or children to an intervention or control condition [33-35]. Six of the 13 experimental studies were conducted in the United States [36-41]. The remaining seven studies were conducted in Belgium [33,35,42], United Kingdom [28-32], Cyprus , and Australia . All study samples contained approximately 50 percent girls. Sample sizes ranged from one to 40 schools and from five to 5488 children. Five studies specifically targeted preschools with children from three to six years old [33,36-38,42], whereas eight studies described the effects of playground interventions on primary schools, with children aged four to 11 years old [28-32,34,35,39-41,43].
In preschools, the five interventions included the provision of play equipment [33,36], the promotion of physical activity on playgrounds by teachers , variations in recess duration , and variations in playground density (m2/child) . Two of the five experimental studies included a control group, i.e., Cardon et al.  and Brown et al. . Cardon and colleagues  compared the effects of two intervention conditions (i.e., the provision of play equipment only and the provision of play equipment and painting of playground markings) with a no-intervention condition. Brown et al.  also tested two intervention conditions (i.e., the promotion of physical activity by teachers against the addition of group discussions and the provision of stickers when children showed sufficient physical activity) against a no-intervention condition.
In primary schools, all eight interventions included the provision of play equipment and/or the application of playground markings. Six of the eight experimental studies in primary schools included a no-intervention condition [28-32,34,35,40,44]. One study examined the isolated effect of playground markings with a no-intervention condition [31,32]. Similarly, Bundy et al.  evaluated the effect of the provision of play materials, only without a control condition. In the other studies the interventions contained multiple components. Of these six studies, Hyberty et al.  was the only study without a control condition. Interventions in two studies contained both the provision of play equipment and the application of playground markings, complemented with the creation of space for team games, and staff training respectively [34,39]. Ridgers et al. [28-30], Colabianchi et al. , and Brink et al.  combined the provision of play equipment and playground markings with playground improvements. Verstraete and colleagues provided both play equipment and activity cards which informed children on the activities that could be performed with different pieces of play equipment .
As is shown in Table 1, nearly all experimental studies (n = 11) evaluated intervention effects on time spent in different levels of physical activity during recess. The proportion of time spent in light, moderate and vigorous physical activity and sedentary activity was mostly assessed with the use of accelerometers, with the exception of three studies in which physical activity was assessed by observations [37,40,41], one study that used pedometers , and one study that used heart rate telemetry [31,32]. Other observed outcomes were playfulness  and post-recess attention . Intervention effects were assessed mostly directly during recess. In case of the provision of playground markings and structural playground improvements, outcomes were assessed with follow-up periods ranging from one month to 12 months.
Of the 33 included publications, 17 reported on observational studies. Their study characteristics are summarized in Table 2. Five of the 17 observational studies were conducted in the United States [44-48] and the remaining studies were conducted in the United Kingdom [49,50], Australia [51,52], Norway [53,54], New Zealand [55,56], Belgium , Sweden , Denmark , and the Netherlands . Forty-three to 61 percent of the study samples were girls, and sample sizes ranged from seven to 130 schools and from 128 to 36.955 children. Five studies specifically reported on associations in preschools (age range: 2.6 to six years old), nine studies in primary schools (age range: five to 12 years old), and three in other types of schools, such as in secondary schools, middle schools or mixed schools (age range: eight to 15 years old). The study of Nielsen et al.  was exceptive, as in this study the association of playground characteristics with physical activity was examined at two time points: at preschool, and at primary school.
Many different playground characteristics were examined and categorized as either hardware, software, or orgware. With regard to software, some studies made a distinction between fixed and loose equipment [45,49,51,60]. Playground characteristics were assessed through self-reports [53,55] or observations [44-52,54,56-60], mostly with the use of validated instruments, such as the Environmental Assessment of Public Recreation Spaces Tool (EAPRS), Children Activity Scanning Tool (CAST), and Environment and Policy Assessment and Observation instrument (EPAO) [46,52,60]. In accordance with the experimental studies, the outcomes reported in observational studies were mainly in terms of time spent in different levels of physical activity during recess. Physical activity was mostly assessed through observational instruments [44,46-49,51,52,60], four studies used an accelerometer [45,50,56,59], three studies used a self-report instrument [53-55], and two studies used a pedometer [57,58]. In addition to physical activity, Colabianchi et al.  also assessed the utilization of the playground through observations.
The methodological quality scores of the experimental studies ranged from 4.5 to 10 points on a 14-point scale. Two of the 13 studies (15%) demonstrated high methodological quality [33,41], eight studies (62%) were of moderate quality [28-32,34-36,39,40,43], and three studies (23%) were of low quality [37,38,42]. None of the experimental studies reported on the use of blinding of study participants, personnel who implemented the intervention or outcome assessors. A summary of the methodological quality analysis per study can be found in Additional file 2. The methodological quality of the observational studies ranged from 5.0 to 9.5 points on an 11-point scale. Seven of the 17 observational studies (41%) demonstrated high methodological quality [46,50,52,53,55,58,60], nine studies (53%) were of moderate quality [30,37,44,48,51,54,56,57],, and one study (6%) was of low quality . Information on the reliability and validity of instruments for the measurement of playground characteristics was reported by only one and five studies respectively. Detailed information on the quality scores of the observational studies can be found in Additional file 2.
Although it was our intention to summarize the effects of (pre)school playgrounds on children’s health in terms of physical activity, cognitive, and social outcomes, nearly all experimental studies (n = 11) focused on physical activity as the primary outcome. Only one study examined the effect a playground-based intervention on cognitive outcomes (i.e., post-recess attention, Holmes et al. ) and social outcomes (i.e., playfulness, Bundy et al. ).
In preschools, the effect of decreased playground density was investigated by one study that showed a significant increase of physical activity levels (see Tables 1 and 3) . The provision of play equipment showed mixed effects on physical activity levels of preschool children [33,36]. No effects were found in the two studies that investigated the provision of playground markings and promotion of physical activity by teachers [33,37]. Increase of recess duration was investigated by one single study that showed positive effects on post-recess attention of preschool children . Further, two studies investigated the effects of a multi-component intervention, including a combination of promotion of physical activity by teachers and guided discussions in one study, and the provision of play equipment and playground markings in another study [33,37]. Both studies showed no beneficial effects on preschool children. In summary, taking into account the study design, the sample size and the methodological quality of the experimental studies, there is inconclusive evidence for an effect of the provision of play equipment, playground markings or for multi-component interventions at preschools on children’s health (Table 3). No evidence was generated for the other intervention strategies at preschools.
Summary of level of evidence for the intervention strategies used in the included experimental studiesa
In primary schools, the two interventions that targeted the provision of play equipment were effective with regard to physical activity levels during recess and playfulness (see Tables 1 and 3) [35,43]. The two studies that investigated the effects of the provision of playground markings found mixed effects [31,32]. Allocating play space for team games was investigated by one study that found a significant beneficial effect . Seven studies investigated multi-component playground interventions, of which the majority showed beneficial effects on physical activity levels [29,34,39,41]. Multi-component interventions that showed beneficial effects contained combinations of staff training, play equipment and playground markings , the allocation of play space, play equipment and playground markings , the provision of playground markings and physical structures [28-30] and the installation of play equipment and asphalt areas  respectively. When summarizing the effects into levels of evidence, one can conclude from Table 3 that there is inconclusive evidence for an effect of the provision of play equipment, playground markings and allocating play space in primary schools on children’s physical activity level. Moderate evidence is found for an effect of the use of multi-component intervention strategies. For the remaining separate intervention strategies no evidence is found.
When the evidence for different playground-based intervention strategies is summarized regardless of school type (preschool or primary school), studies generate moderate evidence for an effect of the provision of play equipment, inconclusive evidence for an effect of the use of playground markings, allocating play space or multi-component interventions, and no evidence for an effect of decreasing playground density, the promotion of physical activity by staff and increasing recess duration on children’s health in terms of physical activity, cognitive, and social outcomes.
With regard to subgroups in which playground interventions are most effective, four experimental studies found stronger effects on the physical activity level and on post-recess attention of girls compared to boys. These interventions entailed the decrease of playground density , the presentation and provision of game equipment and activity cards , and variations in recess duration . Brink et al.  indicated that girls were more active on soft structured surfaces (with play equipment, fall zones and play fields with grass), in contrast to boys who were more active on hard unstructured surfaces (unprogrammed creative play, educational marking areas, sitting and social gathering areas, shade areas). Subgroup analyses in observational studies indicated a more pronounced association between less supervising teachers and physical activity levels in girls . In boys, associations of physical activity levels with the presence of hopscotch and skipping rope areas , decreased playground density , and the presence of soccer fields  were more prominent.
In Table 4 the associations with physical activity are summarized per type of playground characteristic: hardware, software, and orgware. In preschools, summary codes of association indicate that there is evidence for a positive association between physical activity and playground size (++), the presence of an open field with no markings (++), a structured track (+++), decreased playground density (++), and increased recess duration (++). Negative associations were found for the presence of slides (- - -), a sandbox (- - -), swinging equipment (- - -), and supervision (−−).
Summary of associations between playground characteristics and physical activity in (pre)school children, according to included observational studies
In primary, secondary or mixed schools, summary codes of association indicate no associations of hardware playground characteristics with physical activity. Of the software characteristics, the provision of play equipment (++), balls (+++), portable (++) and fixed equipment (+++), and hopscotch/skipping rope area (+++) were positively associated with physical levels. With regard to the orgware playground characteristics, no organized activities (++) and decreased playground density (++) were positively associated with recess activities. Negative associations were found with physical activity for trash cans (- - -) and water and swimming facilities (- -).
When the associations between playground characteristics and physical activity are summarized regardless of school type (preschool or primary, secondary, or mixed school), there is no longer an association between hardware playground characteristics and children’s physical activity level. Decreased playground density and no organized activities are positively associated with children’s physical activity, as well as the provision of portable and fixed play equipment (including balls), the presence of a hopscotch/skipping rope area, an open field with no markings, and a structured track.