Introduction

Early childhood experiences have lifelong impacts. Positive experiences during the early years of childhood can lead to the establishment of long-term habits promoting good health and well-being (Housman, 2017; Shonkoff & Phillips, 2000). Indeed, decades of research, across the developed world, have consistently demonstrated that high quality early childhood education and care (ECEC) positively impacts children’s cognitive, social, emotional, and physical development, and has lasting positive effects on their educational attainment and labor wages as adults (Havnes & Mogstad, 2011; van Belle, 2016; van Huizen & Plantenga, 2018). Universal ECEC would not only transform the lives of children worldwide, it could also—in promoting life long learning opportunities for all—contribute to more equitable and sustainable development (Engdahl & Furu, 2022; UNESCO, 2022; Weldemariam et al., 2022).

The most effective ECEC programs are those that are consistently high in quality, staffed by engaged and qualified adults, and offer children developmentally appropriate pedagogy and stimulating physical and social environments, while also being attentive to children’s basic and special needs (Jalongo et al., 2004; Whitebread et al., 2019). Today, a wide range of ECEC options exist, with varying pedagogical emphases, including play-based, academics-based, and dual language immersion preschools (Essa & Burnham, 2019). Despite differences in their pedagogical approaches, most preschools fit the category of ‘traditional preschools’ in which children spend most of their time indoors, engaged in a mix of playing and learning among peers surrounded by the products of modern civilization, including artificial lighting, climate-controlled rooms, and ready-made toys (Essa & Burnham, 2019).

A small but growing class of alternative preschools—known collectively as nature-based ECEC—focus instead on offering children enriching experiences in outdoor natural settings (Nilsen, 2008; Sobel, 2022). At many nature preschools, particularly those classified as ‘forest kindergartens’, children spend most or nearly all of each day outdoors in nature (Larimore, 2016; Nilsen, 2008). Appropriately attired, nature preschool children spend hours playing with and among the living and non-living materials and structures found in nature, even in inclement weather (Änggård, 2010; Sobel, 2014). Nature preschools began and continue to thrive in the Nordic region but in the last few decades have also become increasingly popular elsewhere, including in Germany, UK, Australia, and USA (Campbell & Speldewinde, 2019; Kahn et al., 2020; Lysklett, 2017; MacQuarrie et al., 2015; Schäffer & Kistemann, 2012).

The growing popularity of nature preschools can be explained in part by the increasing ‘indoorification’ of childhood in recent decades (Sobel, 2022) and its attendant adverse physical effects on children (Hanscom, 2016; McCurdy et al., 2010). There is also growing concern that in spending more time indoors compared to their parents and grandparents, children today are increasingly alienated from nature, to the detriment of their mental health and well-being (Louv, 2005). For example, scholars have pointed to time spent indoors on digital devices as a potential contributor to a wide range of cognitive, physical, and socioemotional deficits increasingly affecting children, teens, and even young adults (Hanscom, 2016; Louv, 2005; Sobel, 2022).

How might differences in time spent indoors versus outdoors in nature affect young children’s growth and development? There is, as yet, limited empirical evidence to evaluate young children’s experiences, behavior, or outcomes across different ECEC environments (Johnstone et al., 2022). Central to discussions about how different types of environments might differently impact young children’s development is the concept of affordance (Gibson, 1979)—the idea that different environments offer different opportunities. Environments differ in the play opportunities they offer children, so that children at one preschool might engage in very different play activities than children at another (Fjørtoft, 2001; Fjørtoft & Sageie, 2000).

Play is central to children’s growth and development (Ashiabi, 2007; Fromberg & Bergen, 2006). Differences in the type and nature of children’s play activities may influence their cognitive, physical, and socioemotional growth and development (Brussoni et al., 2012). By limiting children’s play opportunities, we may be inadvertently hindering their long-term health and well-being (Brussoni et al., 2012; Eager & Little, 2011). For example, some ECEC environments may offer children more opportunities to engage in physically stimulating activities like risky play, play that is thrilling but also entails some risk of physical injury (Sandseter, 2009). Despite decades of societal aversion to risk in the context of children’s play, a mounting body of research suggests that risk taking in play is in fact good for children (Brussoni et al., 2012). Risky play enables children to learn their own physical limitations and develop perceptual-motor capacity (Jambor, 1998). It also fosters physical and psychological resilience (Gray, 2019). Studies suggest that avoiding risk in play during childhood can lead to long-term health problems including anxiety and obesity (Dodd & Lester, 2021; Eager & Little, 2011; Sandseter & Kennair, 2011). The natural environment, with its dynamic terrain and multitude of obstacles and climbing opportunities, is a stimulating playground for children (Fjørtoft, 2001, 2004; Martens et al., 2020).

Several studies suggest children are more creative or imaginative in their play in natural environments (Dankiw et al., 2023; Zamani, 2016), engaging in more pretend play, for example (Cordiano et al., 2019; Drown & Christensen, 2014; Morrissey et al., 2017). Pretend play involves role playing, make-believe, fantasy, and dramatic play (Fein, 1981). These activities help children process their lived experiences and develop the social skills they will need for adult life like turn taking, negotiation, and respecting others’ opinions and ideas (Frost et al., 2021). Given these benefits, ECEC environments that are more spacious, allow for more freedom of movement and speaking, and contain fewer built play structures may encourage more pretend play among children (Drown & Christensen, 2014).

Some ECEC environments may also foster greater equality between the sexes in children’s play activities (Änggård, 2011). Sex differences in play behavior emerge early in life and continue into late childhood, but peak during the preschool years (Berenbaum et al., 2008; Servin et al., 1999). While some degree of sex differences in play behavior is considered developmentally normal in most human cultures (Berenbaum et al., 2008), extreme sex differentiation in play during the preschool years may be detrimental to young children, depriving them of opportunities to develop skills important to their long-term growth and development (Quinlan, 2016). The natural environment offers children a topographically variable, open-ended arena to play in, with trees for climbing, space to run, plants to collect, and places to explore (Änggård, 2010; Sobel, 2014). Since children tend to engage in motor activities when playing outdoors in nature (Fjørtoft, 2001), it may be that natural environments help foster more equality in play behavior between the sexes at an early and pivotal age in development (Änggård, 2011; Sandberg & Ärlemalm-Hagsér, 2011). It is also possible that children engage in fewer gendered play activities in natural environments because they are less exposed to gender norms imposed by past generations through traditional play artifacts possessing characteristics (like a toy’s shape or softness, its utility for social role play, or mechanical movement or propulsion) that elicit more stereotyped play behaviors (Davis & Hines, 2020).

ECEC can also play a role in helping children learn to navigate the complexities of social relationships. Conflict is a pervasive feature of life, even for young children (Cords & Killen, 1998; Shantz, 1987; Tobin et al., 2009), with most conflicts in this age group occurring over the use or possession of play items (Hay, 1984; Shantz, 1987). The ubiquity of leaves, rocks, and sticks in natural environments contributes to their popularity as play objects for children (Kiewra & Veselack, 2016; Lekies & Beery, 2013), and may lead to fewer conflicts among children playing in natural versus more built environments. It is also possible that the greater space available outside and the freedom of movement and expression afforded by the natural environment reduces stress and tension among children (Louv, 2005, 2011), leading to fewer interpersonal conflicts between them.

A growing body of literature suggests that one simple means of improving the overall quality of young children’s lives, and laying the foundation for good cognitive, social, and emotional health from an early age, is to offer them more access to nature. A recent systematic review found striking though preliminary support for the benefits of nature in ECEC–with many studies showing one or more gains in social, emotional, or physical development in children attending nature-based versus traditional ECEC programs (Johnstone et al., 2022). Despite these intriguing patterns, the authors concluded that most studies were limited in sample size or study design and relied on informants like teachers or parents or indirect assessments of children’s behavior (Johnstone et al., 2022). To our knowledge, comparative observations of what children are actually doing while at nature-based versus traditional preschools using quantitative behavioral sampling techniques (Altmann, 1974; McGrew, 1972; Pellegrini et al., 2007) have not yet been carried out.

Given the significant roles of both play and conflict in children’s lives, we conducted an observational study of the activity patterns, play behavior, and patterns of conflict among children attending a nature preschool and a traditional preschool, both in Norway, where access to high quality ECEC is a universal right for children (Haug & Storø, 2013). Norway is an ideal country in which to carry out this research because it has a high mean Gross Domestic Product (GDP), low income inequality, and invests heavily in ECEC, so that differences between individual families are minimized (Haug & Storø, 2013). All ECEC programs in Norway are also legally obligated to offer at least two hours of play time per day outdoors year-round, with nature-based programs offering more time outdoors, and more natural environments for children to play in (Lysklett, 2017; Sandseter & Lysklett, 2018).

We evaluated whether children at nature preschools behave differently from their peers at traditional preschools in Norway by directly observing the children themselves in their ‘natural’ milieu. As non-participant observers, we aimed to obtain insights into the choices children attending nature and traditional preschools make when deciding how to partition their time. This research helps fill in important gaps in our knowledge of young children’s behavior in different ECEC settings, and at a pivotal stage in their development. This knowledge will help inform parents, child care and public health professionals, and others interested in the long-term health and well-being of children, about the potential benefits and drawbacks of nature and traditional ECEC programs. At a broader level, this research contributes to ongoing discussions about how different ECEC approaches can help foster quality education and lifelong learning opportunities for all children (United Nations Sustainable Development Goal 4) as well as promote gender equality (Goal 5) (Engdahl & Furu, 2022; UNESCO, 2022; Weldemariam et al., 2022).

We hypothesized that children at the nature preschool would spend more time outdoors in nature, engage in more physically active forms of play, play less often with human-made items, participate in more pretend play, engage in more gender-neutral forms of play, and exhibit lower rates of conflict than children at the traditional preschool. Time spent on any one activity is time that cannot be spent on other activities (Bateson & Martin, 2021), so we also expected that children in nature preschools would engage in fewer indoor pursuits that can also enrich young children’s lives, including food preparation, sitting down to hot meals together, and gathering up and putting away toys, compared to their counterparts in traditional preschools.

Methods

Study Areas and Subjects

Two preschools were chosen for the study: (A) a nature preschool set in a forest, and (B) a traditional preschool in an urban environment. The forest surrounding the nature preschool is typical of “Nordmarka”, one of 11 forested areas around Oslo that together form the biggest peri-urban forest in Norway (Berglihn & Gómez-Baggethun, 2021). Common tree species in Nordmarka include Norwegian spruce, birch, and European mountain ash and understory species include blueberries, ferns, and terrestrial moss (Jensen et al., 2007). The terrain at the nature preschool is rocky and the topography is slightly to moderately sloped. No fences surround the preschool grounds though children were taught the boundaries beyond which they could not venture. A small school building where kids occasionally spent time indoors and a storage shed were situated on the preschool grounds. Shovels, buckets, and sleds were made available to children daily and skis were sometimes available. Day trips with teachers were made on foot or on cross-country skis (Fig. 1) into other parts of the continuous forest surrounding the preschool, including to an alpine ski slope. The nature preschool occurs at an elevation of ~ 470 m above sea level and experiences an average of > 1000 mm of precipitation annually. Snow cover is typically near-continuous from late November to late April, and was present on 87 of the 93 days of our study (NMI, 2024). Between 11 December 2019 and 12 March 2020, mean daily temperature (± S.E.) near the nature preschool averaged − 0.2 °C ± 0.2 (n = 93 days) (NMI, 2024).

Fig. 1
figure 1

Cross-country ski trip (left) and field trip (right) by members of the nature preschool and traditional preschool, respectively, in March 2020 in Oslo, Norway

Full size image

The traditional preschool was located on a grassy, slightly sloping, open lot with scattered tree cover between a subway platform, an open lot, and residential buildings. Unlike at the nature preschool, conventional playground structures (including for climbing) and sand boxes existed on the traditional preschool’s grounds. A large building with classrooms and a kitchen was located in the middle of the grounds. Children at the traditional preschool had access to a wider array of human-made items than at the nature preschool, both indoors and outdoors, including a variety of toys, costumes, tricycles, shovels, buckets, and balls. A wire link fence surrounded the traditional preschool serving as a boundary beyond which the children could not venture. Day trips with teachers on foot or by subway to other parts of the city, particularly its forests and other natural areas, were common (Fig. 1). The traditional preschool occurs at an elevation of ~ 75 m asl and experiences an average of > 800 mm of precipitation annually. Snow in winter is more intermittent than at the nature preschool and snow cover was present on only 16 days at the traditional preschool during the period of our study (NMI, 2024). Between 11 December 2019 and 12 March 2020, mean daily temperature (± S.E.) near the traditional preschool averaged 2.0 °C ± 0.3 (n = 93 days) (NMI, 2024).

Both preschools were in west Oslo and located ~ 6 km apart. Norway has low income inequality (World Bank, 2019) and a highly educated populace (OECD, 2024) and the parents of the children at both preschools in our study fit this profile. Though we had more children in our study population at the nature preschool (n = 29) than at the traditional preschool (n = 17), the demographic profiles of the children in our two preschool populations were similar with comparable ratios of boys to girls in each population (Table 1).

Table 1 Demographic details for the children studied at the two preschools in our study in Oslo, Norway
Full size table

This study was part of a larger research project on preschool children’s behavior and gut microbiome. Participation in the study was voluntary, and participants were able to withdraw at any time without any adverse consequences. All procedures were approved by California State University Fullerton Institutional Research Board and adhered to the laws of Norway.

Data Collection

After consent was obtained at both preschools, observational research was carried out between September 2019 and March 2020 with data collection terminated on 12 March 2020 when COVID-19 lockdowns began in Oslo. N.N. spent mid-September through early-December 2019 habituating the children and teachers to her presence. Data collection began on 11 December 2019. Throughout the study, N.N. adopted a non-participant observer approach in which she recorded children’s behavior while avoiding interactions with them and attempting to blend into the background so as not to influence their behavior with one another (Altmann, 1974; Barner-Barry, 1986).

Once children and teachers were habituated to her presence, N.N. carried out observations of children at each preschool using both scan and focal sampling (Altmann, 1974). A total of 813 scan samples were conducted at the nature preschool and 607 were completed at the traditional preschool. During these instantaneous scans conducted at 30-min intervals throughout each observation period (Altmann, 1974), N.N. recorded the behavior of each child that was visible—as well as whether the child was indoors or outdoors—from left to right (or vice versa). Behaviors were classified as rest, move, eat, aggression, art, food preparation, clean up, or play. Rest includes instances in which a child was inactive, usually while sitting down. Move includes locomotor behavior, including walking, running, or cross-country skiing, that resulted in a child changing their spatial position to get from one point to another. Eat includes instances when a child consumed food. Aggression involves conflicts of interest in which a child behaved aggressively toward one or more others, including hitting, pushing, forcibly taking a toy, or physically or verbally threatening. Art includes making paper crafts as well as unstructured drawing, coloring, or painting on paper. Food preparation includes the preparation of foods that will subsequently be cooked (by the teachers). Clean up includes gathering and putting away toys and other items that were taken out and played with at earlier points in the day.

Though notoriously difficult to define (Eberle, 2014; Gordon, 2009; Smith & Vollstedt, 1985), we classified as play instances in which children (alone or with a social partner) moved about in an exaggerated manner, manipulated toys or other objects for fun rather than for a specific practical outcome, or pretended to be someone or something other than what they are. Because play was a focus of our study, we recorded additional specific details about the nature of the play children engaged in during scans. We classified play as ‘active’ when it involved physical exertion (e.g., running, climbing, sledding, downhill skiing, digging). We labeled play as ‘social’ when a child was interacting with one or more other children while playing. Further, for social play scans, we recorded the identity of each play partner of the scanned individual. We considered play to be ‘pretend’ when the scanned individual engaged in make-believe, fantasy, imaginative, or dramatic play (Fein, 1981). We recorded whether each incident of play incorporated snow (e.g., digging in snow, sledding on snow, downhill skiing). We also recorded whether play incorporated human-made objects (e.g., shovels, buckets, sleds, downhill skis, toys). Lastly, play behaviors were categorized as masculine, feminine, or neutral according to the description of gender-typed activities given in Goble et al. 2012. The several activities in our study not identical to those in Goble et al.'s (2012) regime were categorized accordingly and are presented in modified version in Table 2.

Table 2 Description of gender-typed activities for children at the traditional preschool and nature preschool in our study in Oslo, Norway (modified from Goble et al., 2012)
Full size table

Focal sampling was another technique we employed to obtain behavioral data (Altmann, 1974). Focal sampling involved following a single focal child at a time and recording all occurrences of behaviors of interest involving the child, including all incidences of aggression. Because children’s aggressive acts are typically short in duration (Shantz, 1987), the continuous nature of focal sampling makes it a better technique for obtaining rates of aggression than scan sampling (Altmann, 1974). Mean focal sample length was equivalent at each preschool (NP: 16.1 ± 0.78 min; TP: 16.0 ± 0.77 min). The number of focal samples completed at each preschool was similar as well (NP: 58 on 8 days; TP: 62 on 10 days). All aggressive acts involving the focal individual (as initiator or recipient) that occurred during these samples were recorded. A child initiating an aggressive interaction was considered the initiator and the child being aggressed against was the recipient.

Data Analysis

We calculated the activity budgets for children at each school by dividing the total number of scans for a school by the number of scans falling into each category of activity at that school. We used activity budgets to describe how children partition their time in different activities at the two preschools. Given strong similarities in time allotted to most major categories, we did not run statistical tests on them.

We used G-tests of goodness of fit (Sokal & Rohlf, 2011) to test our predictions that children at the nature preschool would (1) spend more time outdoors and in nature, (2) engage in a greater proportion of play that was active, (3) engage in a higher proportion of play that was social, (4) engage in a lower proportion of play that incorporated human-made items, (5) display a higher proportion of pretend play, and (6) engage in a lower proportion of gendered play, than children at the traditional preschool. To examine sex differences, we also used G-tests to evaluate the extent to which each of the six aforementioned predictions of differences between the preschools held for members of each sex.

We used the data from the focal samples on aggressive interactions to calculate mean rates of aggression for children at each preschool. Specifically, for each study day, we divided the total number of aggressive incidents involving the focal individuals by the total focal sampling time to determine the mean number of aggressive incidents per hour (i.e., hourly rate of aggression). We used a Kolmogorov–Smirnov test for normality (Sokal & Rohlf, 2011) on the daily rates of aggression per hour, revealing that these data were approximately normally distributed (p = 0.200). We then compared the hourly aggression rates for each day of focal sampling at the two preschools using an independent samples t-test. When calculating the relative likelihood of boys or girls initiating or receiving aggression, we carried out G-tests and controlled for the relative abundance of boys and girls in each study population. All statistical tests were carried out using spreadsheets provided by the Handbook of Biological Statistics (McDonald, 2014) with significance level set at p < 0.05.

Results

Below, we present the results of our study in three sections focusing on the children’s activity budgets, their play behavior, and their aggressive behavior.

Activity Budgets

Activity budgets derived from scan samples were broadly similar at the two preschools (Fig. 2). Most notably, children at both schools devoted equal proportions of their time to play. Children at the traditional preschool moved and rested slightly more than those at the nature preschool. More time was spent eating at the nature preschool, while baking and gathering up and putting away toys were activities engaged in only at the traditional preschool. Proportion of time devoted to aggressive interactions was low at both preschools, though higher at the traditional preschool. Although children at both preschools were mostly outdoors, children at the nature preschool spent significantly less time indoors (8%) than the traditional preschool children (32%; G-test: G = 133.8, df = 1, p < 0.001). Children at the nature preschool also spent significantly more time in natural environments (92%) than children at the traditional preschool (40%; G = 464.7, df = 1, p < 0.001).

Fig. 2
figure 2

Activity budgets of children at the traditional (n = 607 scan samples) and nature (n = 813 scan samples) preschools, depicted as % of time. Note: At the traditional preschool, the ‘other’ category consisted of art (1.0%), gathering up and putting away toys (1.0%), preparing food (0.5%), and aggressive interactions (0.3%). At the nature preschool, the ‘other’ category consisted of art (3.1%) and aggressive interactions (0.1%)

Full size image

Play Behavior

Children at the nature preschool engaged in significantly more active play (80.1% of play, n=291) than at the traditional preschool (49.3% of play, n = 217) (G = 53.4, df = 1, p < 0.001), a pattern that held for both boys (G = 19.0, df = 1, p < 0.001) and girls (G = 35.0, df = 1, p < 0.001) (Fig. 3a). The top three forms of active play were digging (mostly in snow), sledding, and climbing (mostly in trees) at the nature preschool and pedaling (tricycle/scooter), digging (mostly in sand), and climbing (mostly on play structures) at the traditional preschool (Table 3).

Fig. 3
figure 3

Percentage of play scans categorized as a active play, b social play, c pretend play, d play involving naturally occurring snow, e play involving human-made objects, and f ‘gendered’ play (i.e., female- or male-typical) by boys and girls at the traditional and nature preschools. * = significant difference

Full size image
Table 3 Comparison of percentage time devoted to different forms of active play among children at a traditional preschool and a nature preschool in Oslo, Norway
Full size table

Social play was more common at the nature preschool (56.2%) than at the traditional preschool (49.1%) (G = 1.13, df = 1, p = 0.367), a difference that was statistically significant for girls (G = 7.28, df = 1, p = 0.007) but not for boys (G = 1.42, df = 1, p = 0.233) (Fig. 3b). Percent of overall play time spent in pretend play was comparable between schools (NP: 11.7%, CP: 11.5%; G = 0.003, df = 1, p = 0.955), a pattern that held for both boys (G = 0.098, df = 1, p = 0.755) and girls (G = 0.128, df = 1, p = 0.720) (Fig. 3c).

Play incorporating snow accounted for a significantly higher proportion of total play records at the nature preschool (56.7%; e.g., sledding, downhill skiing, digging in snow, rolling snowballs) than at the traditional preschool (1.8%; e.g., sledding, digging in snow, sliding into snow) (G = 208.2, df = 1, p < 0.001), a pattern that held for both boys (G = 129.9, df = 1, p < 0.001) and girls (G = 73.8, df = 1, p < 0.001) (Fig. 3d).

Children at the traditional preschool spent significantly more of their play time playing with human-made items (76.5%) than children at the nature preschool (54.3%) (G = 27.2, df = 1, p < 0.001). This difference between schools in play with human-made items was greater for girls than boys, though significant in both cases (girls: G = 17.9, df = 1, p < 0.001; boys: G = 9.22, df = 1, p = 0.002) (Fig. 3e). Sleds (40%), hand shovels (36%), and downhill skis (5%), nearly all for snow play, were the most often used human-made items by children at the nature preschool, while toys (35%), tricycles/scooters (13%), and hand shovels (11%) were the most often used human-made items by children at the traditional preschool (Table 4).

Table 4 Percentage of play bouts with human-made items incorporating each item among children at a traditional preschool and a nature preschool in Oslo, Norway
Full size table

Lastly, play was > 4 × less likely to be ‘gendered’ at the nature preschool (9.6% of play scans) than at the traditional preschool (45.6%) (G = 87.9, df = 1, p < 0.001). The extent of this disparity between preschools was similar for boys (G = 51.9, df = 1, p < 0.001) and girls (G = 34.6, df = 1, p < 0.001) (Fig. 3f). While several types of play contributed considerably to the gendered play at the traditional preschool, almost all of the gendered play at the nature preschool took the form of pretend play (Table 5).

Table 5 Comparison of gendered play activity incidences among children at a traditional preschool and a nature preschool in Oslo, Norway
Full size table

Aggressive Behavior

The rate of aggressive behavior was significantly higher at the traditional preschool (4.02 ± 0.77S.E. incidents/h) than at the nature preschool (2.25 ± 0.49S.E. incidents/hr) (t-test, p = 0.031) (Fig. 4). Controlling for their relative representation in each study population, boys and girls did not differ significantly in their likelihood of initiating aggression at either preschool (NP: G = 0.268, df = 1, p = 0.605; TP: G = 0.266, df = 1, p = 0.606) (Table 6). Further, at both preschools, boys were not significantly more likely to direct aggression toward one sex than the other (NP: G = 0.343, df = 1, p = 0.558; CP: G = 0.712, df = 1, p = 0.399) (Table 6). However, girls directed significantly more aggression toward other girls than toward boys at both preschools (NP: G = 4.721, df=1, p = 0.030; CP: G = 8.698, df = 1, p = 0.003) (Table 6).

Fig. 4
figure 4

Hourly rate of aggressive interactions among children at the traditional and nature preschools [average number of aggressive incidents per hour of observation (mean of daily values) ± SE]. * = significant difference

Full size image
Table 6 Initiators and targets of aggressive incidents among children at a traditional preschool and a nature preschool in Oslo, Norway
Full size table

Discussion

While their activity budgets were broadly similar, including equal proportions of time allotted to play, we found several major behavioral differences between children at the nature preschool and traditional preschool in our study. Play at the nature preschool was more physically active, less gendered, and less often incorporated human-made items than at the traditional preschool, differences which existed for both boys and girls. Children at the nature preschool were involved in only slightly more than half as many aggressive acts per hour as children at the traditional preschool. On the other hand, children at the traditional preschool engaged in group activities like sitting down to a hot meal together, gathering up and putting away toys, and baking, activities not observed at the nature preschool. Both preschools offered high quality care, play-based learning, and social opportunities, but differences in the physical environments, particularly the greater abundance of trees and open spaces, and greater consistency of snow cover at the nature preschool, afforded children there different play opportunities.

Activity Budgets

Children spend a lot of time outdoors even in winter in most Norwegian preschools (Haug & Storø, 2013; Sandseter & Lysklett, 2018), including the two in our study. Still, proportion of time in natural environments was significantly higher at the nature preschool, consistent with the global mission of nature preschools (especially forest kindergartens) for children to be in nature as much as possible irrespective of climatic conditions (Johnstone et al., 2022; Sobel, 2014). Despite differences in the physical environments between the two preschools, children at each devoted remarkably equal proportions of their time to play. Moving time was also relatively similar at both preschools, though over 10% of the moving time at the nature preschool consisted of cross-country skiing, a form of locomotion not used by children at the traditional preschool. Children at the nature preschool spent more time eating because they took more meal or snack breaks (typically dry flatbreads with cheese and other toppings) while out in the forest. A single large indoor sit-down cooked lunch prepared on site and incorporating plates and cutlery was more typical of the traditional preschool, ensuring socialization to communal dining and actively encouraging healthy eating habits at a vital developmental age (Mikkelsen et al., 2014). While children at the traditional preschool were observed to collectively gather up and put away toys when teachers declared it was time to do so (by announcing ‘ryddetid’ or ‘clean up time’), this activity was not observed at the nature preschool. The reduced emphasis on children gathering up and putting away toys for themselves at the nature preschool likely stems from the children’s much greater use of natural materials in play, materials which did not need to be ‘put away’ at the end of each play session. On those occasions when human-made items were used in play (like a plastic sled for sliding on snow), once they were done with the items, nature preschool children either set them down where another child could play with them or gave them to an adult (who later put it away).

Play Behavior

Although time spent playing did not differ between the two preschools (Fig. 2), the nature and types of children’s play activities did differ markedly between them (Fig. 3). One major difference was that the play of the nature preschool children was significantly more active than that of the traditional preschool children (Fig. 3a). This finding is consistent with earlier studies reporting that preschool children playing in nature exhibit greater motor development or fitness than those without access to nature (Fjørtoft, 2001; Fjørtoft & Sageie, 2000; Grahn et al., 1997). The growing evidence that children are more active in nature is consistent with the notion that nature preschools can contribute to promoting the healthier, more active lifestyles needed to combat the rising obesity rates and problems of motor development increasingly prevalent among children in many countries (Hanscom, 2016; McCurdy et al., 2010; van Stralen et al., 2012).

Still, while the nature preschool children devoted significantly more time to active play in our study, three categorically similar forms of play accounted cumulatively for more than half of all active play at both schools (Table 3). The top active play activity at the traditional preschool (tricycling or scootering) and the second most common active play activity at the nature preschool (sledding) occupied essentially the same niche, though the former was carried out on pavement or dirt, the latter on snow. Similarly, digging was the most common form of active play at the nature preschool and the second most common active play activity at the traditional preschool, though was primarily done in snow at the nature preschool and in dirt at the traditional preschool. The third most common active play form at both schools was climbing which nature preschool children engaged in mostly in trees while traditional preschool children relied primarily on human-made climbing structures. The consistent snow cover and abundance of trees for climbing appear to have facilitated the more active play style of children at the nature preschool and afforded them more opportunities to engage in active play without utilizing human-made items.

Indeed, a second major difference in play behavior between the nature and traditional preschool children was that the children at the traditional preschool played significantly more with human-made objects (Fig. 3e). In addition to pedaling tricycles and scooters and climbing on playground equipment outdoors, traditional preschool children had access to a variety of toys indoors, including blocks, toy vehicles, costumes, and puzzles. Most play with human-made items at the nature preschool involved snow shovels, sleds, and skis. Reducing children’s reliance on human-made objects and encouraging the use of natural objects, especially for play, is a consistent goal across nature preschools (Kiewra & Veselack, 2016; Lysklett, 2017), one that was achieved by the nature preschool in our study. Exceptions were made at the nature preschool primarily for items that facilitated snow play, consistent with the emphasis Norwegian culture places on snow sports and activities (Augestad & Bergsgard, 2007; Sandseter & Lysklett, 2018). There is some evidence that being in nature with fewer human-made objects encourages more creative or imaginative play among children (Cordiano et al., 2019; Dankiw et al., 2023; Kimberly & Keith, 2014; Morrissey et al., 2017). We had therefore expected that pretend play would be more common among the nature preschool children, but our results show that pretend play was equally common among both sets of preschool children (Fig. 3c). Of course, there are other possible dimensions of creative play beyond pretend play, and one candidate is the collecting behavior we observed much more often among the nature preschool children. In total, we noted 23 instances of collecting natural items (e.g., sticks, rocks, leaves) to play with by the nature preschool children versus only four such instances by the traditional preschool children. This collecting play behavior bears some similarities to what anthropologists have observed among children in foraging societies, where it serves to develop children’s foraging skills vital to their survival as adults (Bock, 2005; Crittenden, 2016; Shostak, 1976).

A third major difference in play behavior between the children at the two types of preschools in our study was that play by the nature preschool children (10% of play) was far less often gendered than the play of the traditional preschool children (46% of play) (Fig. 3f). This result is consistent with the notion that the natural setting of nature preschools offers children a less gendered environment than traditional preschool settings and that children will engage in less gendered play activities in nature as a result (Änggård, 2011; Sandberg & Ärlemalm-Hagsér, 2011; Sandseter & Lysklett, 2018). Indeed, some parents select nature preschools for their children in the hopes that play will be less gendered in these schools than in traditional preschools (Änggård, 2011), and our study provides empirical evidence consistent with this notion.

Of course, there may be benefits to gendered play as well, particularly when children of each sex experience both masculine and feminine forms of play since these different types of play activities offer different opportunities for learning (Blakemore & Centers, 2005; Granger et al., 2017; Kersh et al., 2008; Miller, 1987; Wolfgang et al., 2001). For example, research suggests that children who regularly participate in masculine play activities like building with blocks may develop better spatial skills and subsequently math skills (Ishikawa & Montello, 2006; Kersh et al., 2008; Wolfgang et al., 2001), while those engaging in feminine toy play may develop greater nurturance, creativity, and domestic skills (Blakemore & Centers, 2005; Miller, 1987). The traditional preschool in our study with its preponderance of toys and other items made by humans afforded many more opportunities for gendered play. When gendered play occurred at the traditional preschool, masculine play was > 3 to  > 5 times more common than feminine play by both girls and boys, though feminine play still accounted for 9.5% of play in girls and 7.5% of play in boys (Table 5). The greater role of masculine play for both sexes at the traditional preschool resulted especially from their playing with building blocks and toy vehicles, and riding wheeled toys. Not only was gendered play much less common at the nature preschool, those instances which did occur almost always took the form of pretend play (Table 5). Boys at the nature preschool engaged in both masculine and feminine play, while the only gendered play girls there displayed was feminine.

Aggressive Behavior

The rate of aggression among children was significantly higher at the traditional preschool (4.02 incidents/h) than at the nature preschool (2.25 incidents/h) (Fig. 4). The higher rate of conflict at the traditional preschool may reflect the greater availability of human-made objects to fight over, greater proportion of time spent indoors, and/or the smaller, fenced-in, less naturalistic outdoor play area at this preschool. Among urban school age children, natural spaces have been found to foster more prosocial behaviors (Bates et al., 2018). Moreover, previous research suggests that objects provoke most conflicts among preschool children, and that indoor spaces and crowded conditions also tend to be associated with more conflict in this age group (Dawe, 1934; Hay, 1984; Hay & Ross, 1982; Shantz, 1987; Smith & Green, 1975). These factors, combined with the less naturalistic outdoor area at the traditional preschool in our study, may have contributed to the higher rate of conflict among the children there.

We should also note that overall rates of conflict at both preschools in our study were low relative to those reported in previous studies using comparable focal sampling methods. In a review of 11 focal sampling-based studies of preschool children in USA and Canada, Hay (1984) reported rates of aggression ranging from 4.4 to 17.4 incidents/h with a mean of 9.4 incidents/h. Though more contemporaneous studies are needed, our research raises the possibility that there is less conflict among preschool children in Norway than among their counterparts in North America. Equality, harmony, and conflict resolution are major emphases of Norwegian society (Skånland, 2010) and efforts to inculcate these values begin early in preschool (Aasen et al., 2009; NMER, 2005; Skoglund, 2019).

Most of the conflicts in our study were of low intensity involving visual or vocal threats or possession of play items, while pushing and chasing resulting in contact were less common occurrences at both preschools. Controlling for the relative representation of both sexes in our study populations, boys and girls did not differ significantly in their likelihood of initiating aggressive acts at either school. This finding contrasts with those of most previous studies which have found that boys tend to initiate more aggressive acts than girls, particularly the forms of physically and verbally threatening aggression quantified in this study (Baillargeon et al., 2007; Card et al., 2008; Huitsing & Monks, 2018; Ostrov & Keating, 2004).

Sexual selection theory predicts that most aggressive competition, particularly over resources, should occur within the sexes (Buss, 1989; Campbell, 1999; Darwin, 1871). While most studies have not reported the targets of preschool children’s aggression, those that have generally find that boys more often target boys and girls more often target girls (McGrew, 1972; Ostrov & Keating, 2004; Pellegrini et al., 2007; Smith & Green, 1975), both consistent with sexual selection theory. Girls at both preschools in our study also fit this pattern, directing a disproportionate amount of their aggression toward other girls. In contrast, boys at both preschools were no more likely to direct aggression toward same sex partners, an unexpected finding which warrants further investigation.

Conclusions and Suggestions for Future Research

We found that both the traditional and nature preschools in our study in Norway offered high quality care, play-based learning, and social opportunities for young children. Both approaches appear to be model ECEC approaches in their own right in offering “quality education” and “promot(ing) lifelong learning” by starting children along a positive educational trajectory (consistent with U.N. Sustainable Development Goal 4). However, children at the traditional preschool spent most of their time in human-built environments, indoors and outdoors, while children at the nature preschool spent nearly all day in natural environments. Our results suggest that time spent in natural environments impacts children’s play behavior. Specifically, the greater abundance of trees, open spaces, and consistency of snow cover at the nature preschool afforded children there very different play opportunities, leading them to engage in more physically active, gender-neutral play. If future research comparing nature and traditional preschools corroborates our finding that play is less gendered in nature preschool contexts, there may be greater potential for nature preschools to contribute toward advancing gender equality (U.N. Sustainable Development Goal 5). Our research also revealed that children at the nature preschool engaged in fewer aggressive acts, suggesting that the fewer premade toys over which to compete, greater space available outside, and freedom of movement and expression afforded by the natural environment reduces stress and tension among children, leading to fewer interpersonal conflicts. Nevertheless, rates of aggression at both preschools in our study were lower than those reported in preschool populations in other countries, suggesting that efforts to inculcate peaceful coexistence among preschoolers in Norway appear to be succeeding. Children at the traditional preschool engaged in group activities absent from the nature preschool including preparing food, sitting down to a freshly cooked lunch together, and gathering up and putting away toys. These activities reflect the traditional preschool children’s greater use of indoor facilities and human-made toys, and may help foster a sense of community and cooperation. Future research comparing the behavior of nature preschool and traditional preschool children in countries where (unlike in Norway) preschoolers are not required by law to spend several hours each day outdoors might reveal even more pronounced differences in the children's behavior.  

We acknowledge that empirical comparative assessments like ours are limited by the measurement tools employed–intangible qualities like the nature and quality of the relationships between the children and between them and the adults caring for them are more difficult to objectively quantify. Further, one of the most daunting challenges of comparative research on nature versus traditional preschools is obtaining sufficient sample sizes of schools and children in each category to draw generalizable conclusions (Ernst et al., 2021; Johnstone et al., 2022). A recent review of the existing literature concluded that small sample sizes result in ‘low certainty of evidence’ for nearly all studies comparing nature to traditional preschools (Johnstone et al., 2022). Though logistically challenging to organize and carry out, research across a larger number of preschools in each category, across varied environments, and across the different seasons of the year is needed. Nevertheless, our case study from Norway comparing two preschools over a single season represents a valuable contribution to the literature in that it reveals new insights into patterns of play and conflict using observational methods to quantitatively study preschool children’s behavior. It also contributes to ongoing discussions among public health and child-care professionals, policymakers, and parents regarding the relative merits of different approaches to enriching the lives of the youngest members of our societies.

Finally, longitudinal research is needed to track the social, emotional, cognitive, and physical outcomes for children in nature versus traditional preschools (Johnstone et al., 2022; Sandseter & Lysklett, 2018). Obtaining an empirically-based understanding of the long-term outcomes for children in different ECEC environments is crucial to appreciating their relative strengths and weaknesses. This knowledge would enable families to make more informed decisions as to the ECEC that best suits their particular needs and goals, as well as enable communities, especially those in urban environments where access to natural areas is often limited (Colding et al., 2020), to make the most informed investment choices for their future success and well-being.