The dynamic relationship between outdoor environments and children’s play

ABSTRACT

Play is a fundamental activity for experiences, learning and development among children in their early years. Early childhood education and care (ECEC) institutions’ play environments, their features and design, are therefore of great importance for the opportunities provided for children to create and engage in a wide range of play. This study examines how children utilise features in the ECEC outdoor environment (spaces and materials) to engage in different types of play. Children (3–6 years, N = 86) were observed in two-minute sequences during periods of the day when they were free to choose what to do. The data consists of 935 randomly recorded two-minute videos, which were coded second by second to register the type of play occurring, the space in which it occurred and the materials children used. The results indicate a dynamic relationship between the outdoor environments and the play in which children engage and point out the complex nature of playground design, where planning for the predictable, and at the same time opening up for the unpredictable, is important.

KEYWORDS:

• Play environments
• outdoor play
• playground design
• early childhood

Introduction

Play is a ubiquitous activity of young children, and within Early Childhood Education and Care (ECEC), play is a fundamental avenue for learning and deeply rooted as the primary way for meeting children’s development requirements (Broström 2017; Garner and Bergen 2006). In children’s lives, play is a key aspect and an activity in which children engage not as a means to an end, but simply because they enjoy it (Sutton-Smith 2001). From the perspective of children, play is voluntary, self-controlled, fun, active, spontaneous, free, unlimited, natural and self-initiated (Wiltz and Fein 2006). Viewing play as a fundamental need and right of all children offers a powerful construct to legitimise and secure the place of play in the lives of young children (Kernan 2007). For play to appear, children need an environment to play in, and in ECEC institutions, competence in how different qualities and features in the play environment could support different kinds of play (Storli and Sandseter 2019).

Free play is most often defined very broadly as play that is dictated, initiated and controlled by the children themselves (Hewes 2014; Zigler and Bishop-Josef 2006). Santer, Griffiths, and Goodall (2007) elaborate this definition as free play being when children choose what they want to do, how they want to do it, and when to stop and try something else. In ECEC settings that are supporting spaces and materials for free play, the children normally will take the lead with the adults as distance observers of the play, ready to get involved and respond to cues from the children. Children’s spontaneous free play is, by nature, complex in the way that it may exhibit multiple forms, types and stages simultaneously.

The dynamic between children’s play and their play environments

The theory of affordances (Gibson 1979) represents an important framework for considering the utility and flexibility of the physical environment since it concerns the individual's perception of the environment surrounding him/her. It states that the physical environment in which we live affords different actions and behaviours. The concept of affordances includes both the environment and the person, meaning that the affordances are unique for each individual (e.g. while one child may perceive a tree as something to climb, another may see it as a place for hiding when playing hide-and-seek). This person-environment relationship is dynamic, immediate and based on functional activity, which means we must perceive in order to move, but we must also move to perceive (Kyttä 2004). Similarly, Fromberg (2006) claims that children demonstrate their power as agents in their own activities and learning through a dynamic process of play (and meaning), where both predictability and unpredictability is important. The predictability in play are the frames in which the play takes place, such as the agreed-upon theme of the play or the more static physical environment where play develops. On the other hand, the unpredictability in play is how the individual child brings his/her own experiences, ideas, perceptions and creativity into the play situation and develops it in unpredictable ways.

In examining criteria for child-friendly environments, Kyttä (2004) suggested two central criteria for an ideal child-friendly environment: diversity of environment resources and access to play and exploration. Fjørtoft (2000) explored the relationship of environmental affordances to children’s play and development and maintained that environmental complexity and diversity are highly associated with increased play opportunities. Miranda et al. (2017) concretised the concept of child-friendly environments by claiming outdoor environments of high quality provide a wide variety of affordances related to spaces, equipment and materials that encourage a variety of interactions and types of play, while also allowing children independent mobility. In her empirical work, Kyttä (2004) highlighted the positive dynamic interrelationship between children’s independent mobility and the actualisation of affordances. The more mobility licences the children have to explore and play, the more likely they will actualise affordances in their surroundings. When children acutalise affordances they first perceive them and then use and/or shape them in their actitivities. Actualised affordances will in turn motivate children to be mobile. From a dynamic approach of play, Fromberg (2006) similarly stated that when children interact with the physical world, other children and adults, they experience phase transitions that lead to fresh perceptions. Nonlinear phase transitions are situations where children explore what Fromberg (2006, 161) referred to as the ‘bridge’ between one state and another, e.g. finding the exact point to stand or sit to tip the see-saw to the other side, or how hard they can jump on the ice on the puddle before it breaks. When educators match children’s need for nonlinear experiences by providing nonlinear and meaningful themes and environments for play, children can engage in activities in functional, intuitive and intrinsically motivated ways.

Children’s engagement in types of play, related to spaces and materials

Throughout childhood, individual, holistic developmental changes encourage and alter the kinds and levels of play in an intertwisted and interactive way (Garner and Bergen 2006). As children acquire motor skills that allow mobility and exploration of the environment, this will afford possibilities for an increase of functional and object play. Play with objects encourages manipulative skills, and pretend play and play with peers and adults both requires and enhances cognitive skills, as well as increases social interactions. As such, engaging in play enhances development and skills, and development and skills enhance play opportunities. Still, what children’s interests and how they want their play environment to be are important for play to evolve.

Several studies have focused on children’s own perspectives and voices when looking into childrens play and play environments. In a Norwegian study on children’s perspective on outdoor play, Bratterud, Sandseter, and Seland (2012) found through interviews that children preferred functional play such as swinging, sliding/gliding, bicycling, jumping, rolling, climbing, rough-and-tumble/chaotic play, and constructive playing with different textures such as sand and water. Similarly, to examine 6–11-year-olds’ perspectives on public playgrounds and their playground use, Jansson (2015) conducted group interviews supplemented with maps and photos with 141 children in two Swedish cities. Among her results, the children described their use of playgrounds in terms of challenges, manipulation and placemaking. The children especially mentioned green elements and nature as a way of engaging in challenging play such as climbing. When children described their preference for manipulation, they described how they liked to alter their play environments by using, for example, stones, sand and various tools. Placemaking, such as making their own spaces and hide-outs, was connected to green spaces (e.g. trees) and playground equipment (e.g. under the slide). Also, Brussoni et al. (2020) found through interviews that children aged 10–13 preferred neighbourhoods that afforded access to alleys, hiding spots, loose parts and nature where they could engage in activities and play. From interviews supplemented with photographs and drawings, Zamani (2016) found that 4–5-year-olds preferred playgrounds where they could engage in functional play, hiding, dramatic play and explorative play in natural environments/zones and with both loose natural materials—sand, leaves, twigs, stones, dirt and flowers, among others—and loose manufactured elements such as tires, ropes and toys. In an analysis of 5–9-year-olds’ drawings of their idealised play spaces, Ward (2018) also found that children predominantly depicted spaces for active play, such as equipment for climbing, monkey bars, spaces for ball games, scooters, rock walls and rocks to climb on. They also draw spaces such as child-sized houses and materials to support symbolic and fantasy play. They also expressed a wish for nature environments and natural materials, as well as different opportunities to play with and in water. Muela et al. (2019) similarly found, from child photographs and drawings, that children between 3–6 years preferred outdoor environments that afforded challenges to their motor skills in different ways, seeking shelter and hiding, and to experiment with nature material. All these studies show that children have preferences for active, challenging and manipulative play but at the same time wish for spaces where they can withdraw for more calm play and to create symbolic and fantacy play.

One thing is what children wish in their play and play environments, but how children actually take advantage of the available affordances in their existing play environments is also interesting. Observations in Zamani’s (2016) study showed that the children engaged most in functional play, followed by constructive play, exploratory play and then dramatic play in their time on the outdoor playground. The manufactured zone (fixed playground equipment, playhouses, play materials and toys, etc.) and the mixed zone (both manufactured elements and natural elements) afforded more functional play than the natural zone (woods and natural loose elements). On the other hand, the natural zone provided more constructive, exploratory and dramatic play than the other zones. They also preferred to use loose or modifiable elements that inspired their constructive, dramatic and rule-based game play. In examining the impact of playground design on children’s play choices and behaviours in pre-school centres, Dyment and O'Connell (2013) found that children engaged most in functional play followed by constructive play and to some extent symbolic play. Features in the play environment that especially proved to promote functional play were manufactured functional equipment, paved expanses, grassy areas and soft fall surfaces, while constructive play, and to some extent symbolic play, were promoted in areas such as sand features (Dyment and O'Connell 2013). An observational study of children’s preferences on a playground in New Zealand found that swinging, spinning and climbing were the most popular activities, and playground equipment that afforded these activities (e.g. roundabout, eagle’s nest, climbing frame, large-basket swing) were the most used spaces (Bourke and Sargisson 2014). Studies looking at functional and physical active play more specifically have shown that fixed playground equipment and portable equipment such as balls and wheeled toys increases physical activity (Bower et al. 2008; Smith et al. 2014), while other objects or play materials reduce the level of physical activity among 3‒5-year olds (Smith et al. 2014). These studies show that children engage more in functional, exploratory and constructive play in outdoor environements, and that they to a lesser extent have affordances that promote symbolic play in their outdoor environments. This is supported by research showing that children engage far more in symbolic play indoors (20.5%) compared to outdoors (3.1%), and much more in functional play outdoors (36.5%) than indoors (10.3%) (Storli and Sandseter 2019).

The Norwegian ECEC context

In Norway, the ECEC institution (barnehage) is an early-year setting working to meet and ensure educational and care needs of children from birth to six years of age. The Norwegian Framework Plan for the Kindergartens (NMER 2017) applies to all ECEC settings in Norway and guides their pedagogical work with children. Children’s right to play is regarded as an important element of the content of the Norwegian ECEC, and play is considered as having an intrinsic value and being a part of children’s culture. Outdoor and nature play is particularly emphasised in the Framework Plan: ‘Staff shall introduce the children to varied and challenging movement environments, sensory experiences and physical play both indoors and out, within and outside the kindergarten grounds’ (NMER 2017, 50). This includes a focus on the importance of diverse and challenging play environments that offer a wide range of play and activities for all children. Due to the focus on being outdoors and close to nature as promoting important possibilities for play, children attending Norwegian ECEC institutions usually spend a great amount of time outdoors, more than 70% of the total time in ECEC in the summer semester and more than 30% of the total time in the winter semester (Moser and Martinsen 2010). Children usually have access to a rather large outdoor playground connected to the ECEC setting (Lysklett, Emilsen, and Hagen 2003). In general, outdoor environments in Norwegian ECEC provide high-quality, child-friendly environments for multiple play and nonlinear learning situations as described by Fromberg (2006), Kyttä (2004) and Miranda et al. (2017).

Aim of study

The aim of this study was to examine how features in children’s ECEC outdoor environment (spaces and materials) were actualised by children for different types of play in children’s time for free play. In this study, free play implied that children could decide what they wanted to do, where they wanted to be and with whom they wanted to interact.

Method

The sub-study reported in this article, is part of a larger study, Competence for Developing Early Childhood Education and Care (ECEC) Institutions’ Indoor and Outdoor Environments (EnCompetence), funded by the Research Council of Norway and approved by the Norwegian Social Science Data Services. EnCompetence was designed as a mixed-methods research approach (Creswell 2015) and included systematic and randomised video observations of children in outdoor environments during free play at two data points (fall 2017 and fall 2018).

Participants

The participating ECEC institutions in the study were selected from facilities operated by three partnering ECEC owners. The owners made at least twice as many ECEC institutions available as were required for the study and provided relevant information about each of them, including their size, location, age, spatial qualities, number of departments and number of children in attendance. An important criterion for selected institutions was having at least 20 children aged 3‒4 years old who could be recruited as participants. The researchers selected eight ECEC institutions based on a strategic choice to include different types of institutions in terms of the size, quality and age of the spaces therein. On average, Norwegian ECEC institutions have a fenced playground of 2,619.5 square metres, or 47.1 square metres per child (Moser and Martinsen 2010). In this study, the largest playground was 13,000 square metres. While there were large variations in the sizes of playgrounds in the study, the design and fixed playground equipment in the ECEC institutions were quite similar in terms of type, e.g. swings, slides, sandpits and climbing equipment. All institutions provided play materials like tricycles, buckets, toy trucks, cups and spades. In one of the urban playgrounds, the children had no access to nature, e.g. trees, flowers, grass, or natural elements like stones, gravel, logs or sticks.

The strategy for sampling children to participate was to seek informed consent from all the children’s parents, then randomly draw 10 children who consented to participate—five boys and five girls—from each institution. As a result, the first period of data collection (T1) included 80 children. Because the second period of data collection (T2) occurred a year after T1, some amendments were made to the sample at T2. In particular, six of the 80 participants no longer attended the institutions at T2, and one child was not included at T2 for ethical reasons. Following the likelihood of dropout anticipated at T1, a list of other children who consented to participate was used to randomly select seven additional children for T2 to replace the dropouts. However, one of the children was sick on the day of observation, which left only six children as replacement participants. Ultimately, the sample consisted of 86 children: 80 at T1 and 79 at T2. The distribution of gender between T1 and T2 was nearly equal, with 51% of the observations being of boys and 49% being of girls. Children’s mean age was 3.8 years (SD = 0.6) at T1 and 4.7 years (SD = 0.6) at T2.

Procedure and data

All observations were video-recorded and performed in accordance with a strict protocol that ensured a random sampling of observational sequences and identical methods of data collection at each institution. Researchers selected two children to be observed on each day of observation, and each child was observed for six two-minute sequences during free play outdoors. The protocol instructed the data collector to perform each observation by recording Child 1 for two minutes, followed by a six-minute break to locate the next child in the play area. Next, Child 2 was recorded for two minutes as well, followed by another six-minute break to find Child 1 for his or her second round of observation, and so forth. If the data collector encountered a child in a situation that could not be filmed (e.g. using the toilet or changing clothes), then the observation was postponed until filming was permitted. If the child was in such a situation for more than 10 min, then the data collector continued to observe the other child and performed the missing observations at the end of the observation period.

Although a full sample of six observations of 80 children at two data points would have included 960 observations, the final sample included only 935 video observations, meaning that 25 observations were missing (2.6%). Some missing observations occurred because children were sick or picked up early, while others were excluded because the child was hidden from view, the child was preoccupied with the recording equipment, or a technical or human error occurred. The final sample reflected a fairly equal distribution of observations at the two data points, with 471 observations at T1 and 464 at T2.

Ethical considerations

There are special ethical issues in research involving young children (Fine and Sandstrom 1988). One of these issues is the need to gain informed consent from both the parents and the children (for the children also in situ before each observation). Informed consent from parents was gathered through their signature after presenting them with written information about the project and how data was to be handled and disseminated. Concerning children’s own consent, it is important to ensure that the they understand both their own and the researcher’s role during the data collection, and that they can withdraw from the project at any time (Grieg, Taylor, and MacKay 2007). The fellow researchers in this study, who knew the children well, explained to each child in an understandable way the observations that would be conducted, and informed them of their right to withdraw at any time. The researchers were also very conscious to refrain from recording children in sensitive situations such as toileting, change of clothes, etc.

The study was approved by the Data Protection Official for Research in Norway, under the premise that the data would not be analysed or published at group level due to the relatively low number of children in each institution.

Measures

The play types utilised in this study were inspired by Dyment and O’Connell’s play categories (Dyment and O'Connell 2013). For the purpose of this study, the two last categories (self-focused/looking on, and talking) were merged and re-named ‘non-play’ since this was not considered to be play activity. This was, for instance, children just looking at other children playing or looking around for something to do, children and practitioners talking about things other than play-related themes, or merely practical activities such as changing mittens or finding some water to drink. One additional category called ‘mixed play’ (inspired by Luchs and Fikus 2013) was added to capture sequences where it was difficult to assign one play category because the child engaged in a mix of more than one category. The following play categories were used in the coding process:

• Functional (physical play activities, e.g. running, riding bikes, tumbling, climbing rocks, sliding slopes, climbing trees, playful skiing).

• Constructive (building play activities, e.g. building sandcastles, creating huts and shelters, playing with loose parts like sticks, cones, pebbles).

• Symbolic (creative/imaginative play, e.g. role-play, dramatic play, social play like house and pirates).

• Non-play (self-focused/looking on; no interaction with others, not engaged in play, e.g. daydreaming, empty staring, watching activities; or talking, not engaged in active play but talking with another child.)

• Mixed play (when children combine several types of play without any type being dominant).

The two latter categories are not in focus in this article. Non-play was not the focus of this article, and mixed play was found to be very limited in the present data material.

Categories for play spaces and play materials were developed to measure the components of the physical environment. This was done by adjusting categories used in previous research (Cosco, Moore, and Islam 2010; Dyment and O'Connell 2013; Lerstrup and van den Bosch 2017) to the context and theoretical framework for this study. The categories for play spaces included sandbox, pathways, nature, open area, fixed functional play equipment (swings, climbing towers, slides, etc.), fixed role-play equipment (playhouses, boats, huts, stores, etc.), fixed equipment other (tables, storage, etc.) and indoor (cubbies, huts and semi-heated outdoor rooms). Play spaces were coded continuously, and the categories were mutually exclusive. The use or presence of play materials was coded when a child was holding, using or interacting with a material. To capture the idea that children can use several materials at once, the categories were not mutually exclusive. The categories for materials were sand, water, mud, nature materials, (defined) outdoor toys, open-ended materials and wheeled toys.

The assessors coded every category of play type, space and material second-by-second in each observation. The categories were coded by one assessor, and a random sample of 10% of the video observations was reviewed by a second assessor to ensure consistent coding and interpretation. This procedure resulted in discussions about how specific observations should be interpreted. These discussions resulted in a unified understanding of each category, and some minor revisions to the initial coding. The overall consistency was considered satisfying.

Analysis

Categories of play, the space where the child was, and the use of materials were coded with Noldus Observer XT 12.5, a type of behavioural coding, analysis and management software for observed data (Zimmerman et al. 2009). Observer XT data were imported to Stata (MP 15.1) for statistical analysis. Given the hierarchical structure of the data with nested observations of children in ECEC institutions, researchers also used multilevel regression analysis (Goldstein 1986), which enables researchers to control for contextual factors and increases the accuracy of predictions (Gelman 2006). As such, multilevel regression analysis was applied to analyse the association between the play types and spaces and materials in the outdoor environment. This analysis controls for the nested data structure and the children's age and gender. Random intercept models were used in all multilevel analysis. The data were nested at three levels: observation level (level 1) (N = 935), child level (level 2) (N = 86) and institutional level (level 3) (N = 8). The variance partition coefficient (VPC), with a limit of 5% variance, was used to determine the number of levels in the model (Mehmetoglu and Jakobsen 2017). VPC calculations for functional play indicated that there was a 1% variance at the institutional level and 6% variance at the child level. For constructive play, 4% variance was found at the institution level and 10% variance at the child level. For symbolic play, there was a 0% variance at the institution level and a 4% variance at the child level. Two-level models were selected for further analysis. Functional play, symbolic play and constructive play were used as dependent variables in the analysis to investigate the association with spaces and materials in the outdoor environment. Stepwise inclusion of variables starting at the lowest level in the model (Hox 2010) was performed, implying that the variable describing spaces and materials used in the observation was added first, before children's age and gender. An intercept-only model (a model without any explanatory variables) was run first (M0), followed by a model including a variable describing spaces (M1), before the materials were added (M2). Lastly, the second-level variables describing children's age and gender were added to the model (M3). Deviance, Akaike's Information Criterion (AIC) and Schwarz's Bayesian Information Criterion (BIC) are presented to indicate how well the model fits the data and to compare the final model to the previous models (Hox 2010).

Results

The mean duration of the 935 video observations was 122 s (SD = 5). The average amount of functional play in these observations was 35% (SD = 40). Constructive play was coded in 23% (SD = 39) of the observed time, and the average amount of symbolic play was 7% (SD = 22). Table 1 presents descriptive statistics for the key variables.

Table 1. Descriptive statistics (N = 935 observations).

Variable	Mean	SD	Min	Max
Age	4.2	0.7	2.9	5.8
Functional play	35%	40	0	100
Constructive play	23%	39	0	100
Symbolic play	7%	22	0	100
Spaces
Sandbox	9%	27	0	100
Pathways	4%	15	0	100
Nature	6%	22	0	100
Open area	52%	42	0	100
Fixed functional	15%	33	0	100
Fixed symbolic	7%	21	0	100
Materials
Sand	11%	28	0	100
Water	8%	23	0	100
Natural materials	14%	31	0	100
Outdoor toys	32%	44	0	100
Open-ended	7%	23	0	100
Wheeled toys	13%	32	0	100

Functional play, spaces and materials

The final regression model (M3) for functional play (Table 2) indicates that there is a positive association between the space categories for fixed functional equipment, pathways, open area and nature and children's engagement in functional play. The amount of functional play in the observation is estimated to be 53% higher when children spend 100% of the observed time on fixed functional equipment. Using pathways the entire observation is estimated to increase the amount of functional play by 30%, whereas the effect of using open areas (17% increase) and nature (16% increase) is weaker. Besides wheeled toys (25% estimated increase), all categories describing children's interaction with materials are negatively associated with functional play. There is no significant association between functional play and the child's gender or age. For functional play, M1 and M2 are significantly (p < .001) improved models compared to the previous model using a likelihood-ratio test. M3 is not a significantly improved model compared to M2. In line with this, AIC and BIC measures indicate that M2 is the better model. The considerable reduction in Deviance, AIC and BIC from M0 to M1 and M1 to M2 indicates that spaces and materials have a substantial impact on the amount of functional play.

Table 2. Regression models for functional play (N = 935 observations).

Model	M0 Empty	M1 Spaces	M2 Materials	M3 Final model
Fixed part	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)
Intercept	35 (2)	12 (5)	22 (5)	9 (9)
Sandbox		−.12 (.06)	.02 (.07)	.02 (.07)
Pathways		.53 (.09)***	.30 (.09)**	.30 (.09)**
Nature		.14 (.07)	.17 (.07)*	.16 (.07)*
Open area		.23 (.05)***	.17 (.05)**	.17 (.05)**
Fixed functional		.62 (.06)***	.53 (.06)***	.53 (.06)***
Fixed symbolic		−.02 (.08)	.03 (.07)	.03 (.07)
Sand			−.15 (.06)**	−.15 (.06)**
Water			−.15 (.05)**	−.14 (.05)**
Natural materials			−.18 (.04)***	−.17 (.04)***
Outdoor toys			−.12 (.03)***	−.12 (.03)***
Open-ended			−.10 (.05)*	−.12 (.05)*
Wheeled toys			.24 (.04)***	.25 (.04)***
Age				3.0 (1.6)
Boy				.8 (2.4)
Random part
Level1Var.	1510 (73)	1195 (58)	1032 (50)	1030 (50)
Level2Var.	101 (38)	37 (24)	28 (20)	24 (19)
Deviance	9545	9304	9163	9160
AIC	9551	9322	9194	9194
BIC	9565	9365	9266	9276

*p < .05: **p < .01: ***p < .001.

Constructive play, spaces and materials

The final regression model (M3) for constructive play (Table 3) indicates that there is a positive association between the space categories sandbox and nature, and children's engagement in constructive play. The amount of constructive play in the observation is estimated to be 14% higher when children spend 100% of the observed time in sandboxes. Being in nature for the entire observation is estimated to increase the amount of constructive play by 18%. Besides wheeled toys (no significant association), all categories describing children's interaction with materials are positively associated with constructive play. The estimated effect of using sand (49% increase) and water (42% increase) on constructive play are strong. Also, using open-ended materials (26% increase), outdoor toys (20% increase) and natural materials (18% increase) are positively associated with constructive play. There is no significant association between constructive play and the child's age. There is significantly (p < .05) more constructive play among the girls in the sample. For constructive play, M1 and M2 are significantly (p < .001) improved models compared to the previous model using a likelihood-ratio test. M3 is not a significantly improved model compared to M2, and also the model fit measures indicate that M3 is no improvement compared to M2. The substantial reduction in Deviance, AIC and BIC from M0 to M1 and M1 to M2 demonstrates that the variables describing spaces and materials in the physical environment explain a considerable amount of the variation in constructive play.

Table 3. Regression models for constructive play (N = 935 observations).

Model	M0 Empty	M1 Spaces	M2 Materials	M3 Final model
Fixed part	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)
Intercept	35 (2)	24 (5)	7 (4)	15 (7)
Sandbox		.57 (.06)***	.14 (.06)*	.14 (.06)*
Pathways		−.34 (.08)***	−.17 (.07)*	−.17 (.07)*
Nature		.20 (.07)**	.18 (.07)**	.18 (.06)**
Open area		−.04 (.05)	−.04 (.06)**	−.04 (.04)**
Fixed functional		−.24 (.06)***	−.10 (.05)*	−.10 (.05)*
Fixed symbolic		−.02 (.07)	−.10 (.06)	−.10 (.06)
Sand			.49 (.05)***	.49 (.05)***
Water			.43 (.04)***	.42 (.04)***
Natural materials			.19 (.03)***	.18 (.03)***
Outdoor toys			.20 (.03)***	.20 (.03)***
Open-ended			.25 (.04)***	.26 (.04)***
Wheeled toys			−.03 (.03)	−.04 (.03)
Age				−1.4 (1.4)
Boy				−4.2 (2.1)*
Random part
Level1Var.	1510 (73)	1012 (49)	681 (33)	679 (33)
Level2Var.	101 (38)	85 (30)	30 (15)	27 (15)
Deviance	9454	9179	8786	8781
AIC	9460	9197	8816	8815
BIC	9474	9240	8889	8897

*p < .05: **p < .01: ***p < .001.

Symbolic play, spaces and materials

The final regression model (M3) for symbolic play (Table 4) indicates that there is a positive association between the space category for fixed symbolic equipment and children's engagement in symbolic play. The amount of symbolic play in the observation is estimated to be 14% higher when children spend 100% of the observed time on fixed equipment for symbolic play. There is no significant association between children's interactions with materials and symbolic play in M3. Older children engage significantly more in symbolic play (4.3% per year). Also, boys engage significantly more (3.7%) in symbolic play. For symbolic play, M1 and M3 are significantly (p < .001) improved models compared to the previous model using a likelihood-ratio test. M2 is not a significantly improved model compared to M1. Similarly, AIC and BIC measures indicate that M2 fits the data worse than M1 and thus that the variables describing children’s use of materials do not contribute to explaining the variation in symbolic play. Furthermore, the limited reduction in Deviance, AIC and BIC from M0 to M1 indicates that the spaces measured in the present study are of limited importance for symbolic play.

Table 4. Regression models for symbolic play (N = 935 observations).

Model	M0 Empty	M1 Spaces	M2 Materials	M3 Final model
Fixed part	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)	Coeff.(SD)
Intercept	7 (1)	14 (3)	13 (3)	−8 (6)
Sandbox		−.04 (.04)	−.02 (.05)	−.02 (.05)
Pathways		−.11 (.05)	−.10 (.06)	−.10 (.06)
Nature		−.08 (.05)	−.09 (.05)*	−.10 (.04)*
Open area		−.10 (.03)**	−.10 (.03)**	−.10 (.03)**
Fixed functional		−.09 (.04)*	−.08 (.04)*	−.07 (.04)*
Fixed symbolic		.13 (.05)**	.13 (.05)**	.14 (.04)**
Sand			−.06 (.04)	−.05 (.04)
Water			−.07 (.03)*	−.06 (.03)
Natural materials			.01 (.02)	.02 (.02)
Outdoor toys			.04 (.02)*	.04 (.02)
Open-ended			.06 (.03)	.03 (.03)
Wheeled toys			.01 (.03)	.03 (.03)
Age				4.3 (1.1)***
Boy				3.7 (1.6)*
Random part
Level1Var.	454 (22)	429 (21)	424 (21)	416 (20)
Level2Var.	21 (10)	23 (10)	22 (9)	18 (9)
Deviance	8407	8358	8347	8325
AIC	8413	8376	8377	8359
BIC	8428	8420	8450	8441

*p < .05: **p < .01: ***p < .001.

Discussion

Being the most common type of play (Table 1) in the data material in this study, functional play apparently is a type of play the ECEC outdoor environments afford for children, and children seek to engage in. This is in accordance with earlier studies finding functional play to be the most common type on outdoor playgrounds (Dyment and O'Connell 2013; Zamani 2016). The results (Table 2) show that fixed playground equipment for functional play, as would be expected, is strongly associated with functional play among children. Children use this equipment, among others, to slide, climb, swing, spin and jump. These results are in accordance with earlier research showing that children engage in functional play in equipment specifically designed for this kind of play (Bourke and Sargisson 2014; Bower et al. 2008; Dyment and O'Connell 2013; Smith et al. 2014; Zamani 2016). Knowing that many outdoor playgrounds are dominated by large fixed structures designed for functional play, this is not surprising. Nevertheless, there is reason to ask whether this makes less space for equipment that affords other types of play and as such makes the play space less diverse and inviting for more diverse play. On the other hand, in children’s view, this kind of equipment are what they ideally want to be included in their play environment (Ward 2018).

In addition, pathways and open areas are significantly related to functional play in the present study. Seen in connection to wheeled toys as also promoting functional play, children seem to use tricycles, bicycles, trolleys, etc. for movement play on pathways and open spaces with gravel, concrete, asphalt and similar surfaces. This is supported by earlier research finding that paved expanses, soft fall surfaces (Dyment and O'Connell 2013) and wheeled toys (Smith et al. 2014) afford functional and physical active play. The findings show that nature spaces also afford functional play. In nature spaces, children would have more difficulty using wheeled toys with the natural uneven, grassy or stony surfaces. Rather, nature spaces are predominantly used for gross motor activities such as climbing (e.g. trees), sliding down (e.g. hills), balancing (e.g. on logs), rolling (e.g. down a grassy hill), and children have expressed a wish for nature spaces and nature features in order to engage in active and challenging play (Brussoni et al. 2020; Jansson 2015; Ward 2018). In the present study, some of the participating ECEC institutions had limited or no access to natural spaces within their playground. The results calls for more awareness on how to design play spaces for children where nature is kept instead of removed, or added as an important part of the space when the chosen site does not include nature to start with.

While it seems that a large part of the typical spaces and equipment normally included in playground design (functional equipment, pathways, open areas, nature spaces) afford functional play, it is interesting to see that all other sorts of materials or toys (apart from wheeled toys), similar to Smith et al.’s (2014) findings, have a negative impact on children’s functional play. This means that even though one would think that materials such as sand, water, defined toys, nature materials and open-ended materials have the potential to afford activities that include movement and functional play types (Kyttä 2004), the children in this study do not actualise these affordances. This study is not designed to give reasons for such findings, but one reason could be institutional restrictions on where to use materials. Bringing sand, water and other materials onto the slide or climbing frame would for instance be avoided in many ECEC institution due to the mess it might create.

Constructive play is the second most prevalent play type (Table 1) in this study and includes activities where children build or construct things out of different kinds of play materials. This could be materials of all kinds and in all sizes. Therefore, it is not surprising that all kinds of materials in this study (sand, water, natural materials, defined outdoor toys, open-ended materials) have a positive relation to children’s constructive play (Table 3). Sand and water especially seem to afford constructive play among the children in the present study. Children have a preference for play where they can manipulate, explore and construct with materials such as water, stones, sand, tools, leaves, twigs, dirt, flowers, tires, ropes, toys, etc. (Brussoni et al. 2020; Jansson 2015; Ward 2018; Zamani 2016). Looking at spaces that promote constructive play, the results show that nature spaces and sandboxes afford opportunities for constructive play. Combining the use of diverse play materials in these spaces would give very good conditions for creating constructive play, with a mix of defined toys, open-ended materials and more natural materials. One could expect that this would lead to nonlinear creative exploration of different ways of using the material to construct new and existing play props (Fromberg, 2006; Muela et al. 2019; Ward 2018; Zamani 2016). Earlier studies have, similar to the present study, found nature spaces (Zamani 2016) and sandboxes (Dyment and O'Connell 2013) to afford constructive and exploratory play. On the other hand, the results of this study show that pathways, open areas and fixed functional equipment are negatively associated with constructive play. The aforementioned proposed reason for materials being negatively related to functional play, typically happening on functional equipment, pathways and open areas, could also be an explanation for this finding. Children need an abundance of different play materials to use in their constructive play, and this would then naturally emerge in spaces where materials are already present or where the children are allowed to bring materials for their use. Cultural constraints (Kyttä 2004) on not interfering with a lot of materials, or materials not being easily accessible in some parts of the outdoor environment would restrict constructive play to confined spaces.

Similar to constructive play, symbolic play (also called dramatic play or role-play) would be promoted by access to loose modifiable materials and elements (Zamani 2016). In the present study, the results show, surprisingly enough, that there is no significant association between any of the coded materials and symbolic play. The results of the present study also show that the type of space within the playgrounds significantly affording symbolic play is fixed equipment for symbolic play. Even though the ECEC institutions in this study had a range of play materials available for the children, there is reason to believe that these afforded more constructive play rather than symbolic play, and they were most likely not available or used in connection with children spending time on equipment for symbolic play. Materials that especially support symbolic play such as cups, cutlery, plates, mugs, clothes for dressing up, dolls, etc. might not have been available close to these features. There is also reason to reflect upon the result that the other spaces (open areas, functional equipment and nature) are negatively associated with symbolic play (Table 4). Looking at earlier research finding that children wanted natural zones and green spaces with loose materials to use for dramatic and symbolic play (Jansson 2015; Zamani 2016) and that sand features and nature spaces afford symbolic play (Smith et al. 2014; Zamani 2016), the present findings are puzzling. Again, sandboxes and nature spaces, with the available materials there, seem to afford constructive play rather than symbolic play. It must also be noted that the amount of time children spend engaging in symbolic play in the outdoor environment was not very high (7%), and that this indicates that the play spaces overall were not designed in a way that promoted symbolic play. Research shows that children’s symbolic play is higher in the ECEC indoor environment compared to the outdoor environment (Storli and Sandseter 2019), and this would indicate that there are more appropriate materials for this kind of play indoors.

In accordance with the notion that developing symbolic play requires somewhat enhanced cognitive skills (Garner and Bergen 2006), symbolic play was the only play type where the results show an age effect, with older children engaging in more symbolic play than the younger ones.

Limitations

Drawing on cross-sectional video observations from children's everyday environment in ECEC, there are limitations to this study. No causal relationships, only associations between children’s play and the available spaces and materials can be established. The findings that there is a limited amount of symbolic play and that materials are not associated with children’s engagement in symbolic play may be a result of the limitation due to problems of identifying and discovering symbolic play, which sometimes could be more subtle and not expressed as clearly as the other play types. Using play types as mutually exclusive codes led the assessors, in many instances, to code the play that was most clearly expressed in the situation, possibly favouring functional play and constructive play. The possibility of using dual coding (more than one type of play possible at the same time) might have resolved this problem, but would also have made the data material more complex and somewhat messy. Another limitation for discovering symbolic play may have been the choice of using 2-minute videos as a unit. This may have made it more difficult for the researchers to discover symbolic play that often need longer time to develop and become visible for the onlooker. The results may be highly influenced by the cultural context of the study. Norwegian ECEC has a strong emphasis on free play, outdoor play and providing diverse and challenging play environments (NMER 2017). Still, associations similar to those found in the present study are also found in other cultural contexts (Dyment and O'Connell 2013; Zamani 2016).

Implications

The children participating in this study are executing a variety of play activities and utilising a wide range of spaces and materials in free outdoor play. More than a third of the total observation time is identified as functional play, like running, riding bikes, climbing and tumbling. Together with constructive play, these two categories constitute most of the time children play in this study. Symbolic play represents only a small amount of the total time the children spend in outdoor play. The outdoor environments in ECEC institutions in the present study were primarily designed for physical active play, with space to move and equipment facilitating active play. In this sense, the children in this study perceive much the same potential affordances (Kyttä 2004) and utilise them in a somewhat predictable manner. On the other hand, the present study also demonstrates some surprising and unpredictable findings, particularly looking at how children utilise, or rather do not utilise, materials in their outdoor play. Materials being negatively associated with functional play and not associated with symbolic play is not as expected. These results could indicate that, even though Norwegian ECEC institutions’ outdoor environments in general are considered to be child-friendly, the availability and type of material to support functional and symbolic play might not be very good. Or, children choose not to use them the way adults expect them to do. The dynamic ways in which children perceive and make use of features in their environment is complex and characterised by both predictability and unpredictability. For practitioners and facilitators (e.g. landscape architects) of children’s outdoor play, an understanding of predictability of how children perceive affordances and actualise them should be basic knowledge. At the same time, practitioners and facilitators must be aware of how the individual child should be able to use his/her own experiences, ideas, perceptions and creativity to develop the play in dynamic and unpredictable ways (Fromberg, 2006). Unpredictability is one of the most important factors in children’s play (Lester and Russell 2010), and according to Fromberg (2006) unpredictability represents unique moments in play where children create meaning and expand their experiences. When developing children’s play environments, one can direct some kind of predictability in how children may actualise affordances in the environment. Nevertheless, there will always occur unexpected ways of how children perceive and interact with the surrounding environments, and the design of child-friendly environments should provide and support both predictability and unpredictability. As such, the findings in this study should be a wake-up call to landscape architects and pedagogues to celebrate the dynamic and unpredictable in planning and facilitating outdoor play environments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Research Council of Norway under Grant 270727; Norges Forskningsråd.