Early Childhood Education Journal,Vol.29,No.2,Winter 2001 (2001)
The Natural Environment as a Playground for Children:
The Impact of Outdoor Play Activities in Pre-Primary
INTRODUCTION have experienced positive results from being outdoors
in natural environments, but only a few studies have
Norwegian studies have revealed a disquieting ten- been done in this field (Bang et al., 1989; Fjørtoft, 1999;
dency that children are becoming more sedentary in Grahn et al., 1997). We know far too little about how
their adolescence. They spend more time, approximately the natural environment functions as a playground for
three hours a day, on TV, video, and electronic media children, and we know even less about what effects such
(MMI, 1995). The movement pattern of children has a playground might have on learning in children. The
changed remarkably the last 10–20 years. The unorgani- physical outdoor environment, and the natural environ-
sed traditional games, which included lots of moving ment in particular, as a play habitat for children, has
around, are now changing into sitting in front of your been a topic of low priority in child research (Bjerke,
private computer playing computer games. Such scenar- 1994).
ios have resulted in several health hazards like increas-
ing obesity in early childhood (Anderson et al., 1998),
THE AFFORDANCE OF NATURE
and motor problems in children are reported in several
Scandinavian studies (Due et al., 1991; Hertzberg, 1985; Natural environments represent dynamic and rough
Gilberg and Rasmussen, 1982; Kjos, 1992; Ropeid, playscapes that challenge motor activity in children. The
1997). However, a recent study of the physical activity topography, like slopes and rocks, afford natural obsta-
among 3–7 years old Norwegian children (MMI, 1997) cles that children have to cope with. The vegetation pro-
showed that 75% of the children spend some time out- vides shelters and trees for climbing. The meadows are
doors by their own every day. The most active ones for running and tumbling. Description of physical envi-
practiced several outdoor activities such as skiing and ronments usually focuses almost exclusively on forms.
hiking in the wilderness, climbing trees, enjoying water Heft (1988) suggested an alternative approach to de-
activities, and soccer in the field. Four out of ten chil- scribe the environment, which focused on function
dren expressed a wish for more time for physical activity rather than form. The functional approach corresponds
(Hansen, 1999), but children complain about the lack of better to the children’s relations to their environment.
suitable arenas for play and free time activities, such as Intuitively children use their environment for physical
grounds for climbing, building dens, sliding, and skiing challenges and play; they perceive the functions of the
(Mjaavatn, 1999). Francis (1988) argued that children’s landscape and use them for play. The central concept
play in an unstructured environment, preferably a natu- guiding children’s examination of their environment is
ral one, gives the children a genuine understanding of that of affordance. Gibson (1979) developed the concept
reality. Rivkin (1990, 1995) highlights the values of out- “affordance” to describe an awareness of the environ-
door play, but regrets that children’s access to outdoor ments and their functional significance, or their func-
play habitats are vanishing. tional meaning. For example if a rock is big enough to
Several kindergartens in the Scandinavian countries fit the hand, it might be perceived as an object to grasp
or to throw; it affords grasping or throwing. A tree that
is appropriately branched and stemmed, will likewise be
Address correspondence to Ingunn Fjørtoft, Tel.: +47 35026333, Fax: perceived as climb-on-able; it affords climbing-on. Na-
+47 35026201; e-mail: Ingunn.Fjortoft@hit.no; Telemark University
College, Department of Teacher Training, 3679 Notodden, Norway. ture provides an environment with such possibilities and
1082-3301/01/1200-0111$19.50/0 2001 Human Sciences Press, Inc.
affordances. Frost (1992) introduced the concept “play- might be characterized as quasi-experimental approach
(Robson, 1993, Thomas & Nelson, 1985). The groupsscape” to describe different play environments. He ar-
gued that natural features are important qualities of play- were selected from three kindergartens equal in age
groups. The experimental group of 46 children from onegrounds, and that the natural features allow a wide range
of learning opportunities not available from other play- kindergarten was offered free play and versatile activi-
ties in the forest next to the kindergarten. The experi-ground options. Hart (1979) and Moore (1986) have de-
scribed the children’s preferences of wilderness and un- mental group used the forest every day for 1–2 hours
throughout the year when they attended the kindergarten.structured landscapes for play.
In Scandinavia it has become popular for kinder- Only randomly they used the outdoor playground inside
the kindergarten fence. As reference group 29 childrengartners to spend more time outdoors in the natural envi-
ronment. Some kindergartens are organized as outdoor of the same age groups from two kindergartens in the
neighbouring district were chosen. The groups wereschools, where the children, aged three to six, spend all
or most of the day outdoors in a natural environment. checked out for differences in socioeconomic living con-
ditions by multiple regression analysis, using parents’ ed-Playing in a natural environment seems to have positive
effects on children; they become more creative in their ucational and professional background as variables. The
reference group used their traditional outdoor playgroundplay, and play activities and play forms are increasing.
It is also indicated that absence due to sicknesses is for 1– 2 hours a day and visited natural sites only occa-
sionally. Both groups had the same standard playgroundlower among children in outdoor kindergartens than in
the traditional ones (Grahn et al., 1997, So
¨m, equipment, such as sandpit, a swing, a seesaw, a slide
and a climbing house in their outdoor play ground. The1998). Not the least it is evident that the children’s mo-
tor fitness is improved. They move easily around in a study started with a pretest in September. The treatment
period lasted for nine months, and was terminated with arugged terrain and cope with physical challenges, which
improve their motor ability (Fjørtoft, 1999, Grahn et al., posttest in June the following year.
Both groups were tested with the EUROFIT: Eu-1997). Although few in number, these studies indicate
that the natural environment is a stimulating arena for ropen Test of Physical Fitness,the Motor Fitness Test
(Adam et al., 1988). The test included the following testlearning in general, and for motor fitness training in par-
ticular. The present research corroborates the main find- items: Flamingo balance test (standing on one foot) for
testing of general balance, Plate tapping (rapid tappingings.
of two plates alternatively with preferred hand) measur-
ing the speed of limb movement. Sit and reach ex-
VERSATILE PLAY IN THE NATURAL pressed flexibility in knee and thigh joints. Standing
ENVIRONMENT AND THE IMPACT ON broad jump (jumping for distance from a standing start)
CHILDREN’S MOTOR DEVELOPMENT measured explosive strength. Sit-ups (maximum num-
bers of sit-ups achievable in half a minute) measured
Objectives trunk strength. Bent arm hang (maintaining a bent arm
position while hanging from a bar) for testing of func-
The notion that versatile play in a natural environ- tional strength in arm and shoulder, and Shuttle run (a
ment might have an impact on children’s development running and turning, shuttle, test at maximum speed)
constituted the background for the present study. The testing running speed and agility. Two additional tests
aim of the study was to investigate how children’s play- were introduced: Beam walking for testing dynamic bal-
ing in the natural environment might stimulate their mo- ance and Indian skip (clapping right knee with left hand
tor fitness and it was decided to focus on the affordances and vice versa), which tested cross coordination (Fjør-
of the landscape and the correlation for versatile play. toft, in press).
The main objectives were, 1) to focus on the affordances Data analyses were performed according to the sta-
of the landscape for versatile play and, 2) to examine tistical programme SPSS/PC+, the PC version of the
the impact of outdoor play activities in children’s motor Statistical Package for the Social Sciences (Norusis,
ability and mastering. 1993; Frude, 1993). Frequency analyses, means and the
T-test for independent samples and paired samples, cor-
Methods relations; multiple regression analyses and factor analy-
ses were applied for data processing (Fjørtoft, in prep.).An experimental study was carried out with five-
to seven-year-old children in kindergartens in Telemark, In this article the main findings of the study will be
presented.Norway. Because of the lack of randomization, the study
113The Natural Environment as a Playground
Results tumn time were all located behind and above and close
to the kindergarten area. This area included 5 different
types of woodland, the low-herb woodland being the
The Study Area
The site of the investigation was a small forest of dominant type of vegetation, see Figure 2. The mixture
of woodland types represented a high diversity in vege-7.7 hectars of mixed woodland vegetation, located close
to a kindergarten in Bø, Telemark County in Norway. tation elements.
The variety of woodland vegetation and the physi-The landscape pattern showed a mosaic of patches of
woodland with some open spaces of rocks and open ognomy of trees and shrubs in the area made the afford-
ances for play and play habitats an offer of multiplefields and meadows in between. The topography, ex-
pressed as slope and roughness, was varied with some choices. The shrubs constituted a mixture of scattered
species, which afforded shelter and hiding, social play,steep cliffs, slopes, and plains. Vegetation and topogra-
phy jointly afforded a diversity of play habitats for the and construction play. Very special was the flexible ju-
niper bush, which motivated for functional play (how tochildren (Fjørtoft & Sageie, 2000). The forest was lo-
cated outside the fence and behind the kindergarten get in and out) and social play (play house) as well.
Some trees were suitable for climbing depending on the(Figure 1). In the closest parts the children were allowed
to go at will, but in the farther parts the children had to branching pattern, the stem diameter, and the flexibility
of the tree. The young deciduous trees were easily acces-be accompanied by adults.
sible for climbing (Figure 3).
The spruces were more suitable for hiding than for
Play Habitats climbing due to the dense branches. The more open areas
in the pine and low-herb woodland afforded running,The children more frequently used some favorite
places in the forest. These play habitats were located chase and catch, leapfrog, play tag, and other games that
an open space can afford. The shrubs afforded hide-and-close to the kindergarten and represented specific play
habitats for summer and winter play activities (Fig. 2). seek, building dens and shelters and role-play like house-
and-home, pirates, fantasy, and function play (Figure 4).The play habitats used in the spring, summer and au-
Fig. 1. The kindergarten and the forest.
play like building snow figures and dens. The genuine
winter habitat was a meadow located next to the kinder-
garten and comprised a soccer field and the lower parts
of a ski jump arena (Figure 2). These fields were used
by the kindergarten almost solely as a skiing arena in
the winter. For children at the age of 5–7 years the more
gentle slopes of the ski-jump arena were used for differ-
ent skiing disciplines (Figure 6).
The groups matched in age with a mean age of 6.1
years and there were no significant differences in age
between the groups. It was the six-year-olds that domi-
nated both groups. The sex distribution in the groups
showed a predominance of boys in the experimental
group (27 boys, 19 girls), while in the reference group
there were more girls (18 girls and 11 boys). There were
no significant differences in test results between the
sexes. Body mass and height did not show any signifi-
Fig. 2. The forest. Vegetation map. Playscapes indicated by grey color.
The topography was undulating with terraces and
slopes and a dominant cliff traversing the area, which
afforded slopes for sliding and cliffs for climbing (Fig-
The children’s favorite places were named “The
Cone War,” The Space Ship,” and “The Cliff.” The
naming itself is illustrative for the activities taking place
there. Free play fostered creative play, and the playscape
afforded loose parts and natural objects and materials to
play with. Play activities were observed related to the
affordances of the vegetation and the topography (Fjør-
toft & Sageie, 2000).
The play habitats in the forest were also used dur-
ing the wintertime, but their functional use was differ-
ent. The cliff turned into sliding slopes of different cate-
gories. With appropriate clothing with oilskins (trousers
for wet climate), the children made high-speed competi-
tions in different sliding disciplines: on their backs, on
their stomachs, feet first, head first, and so on. The steep
slope afforded shorter but more challenging rides. The
deep snow provided affordances for tumbling, rolling,
and other acrobatics. A dense snow layer made the trees
more accessible for climbing. The play categories in the
forest during the winter season can be categorized as
functional play (climbing, crawling, making angels in
the snow, etc.), role-play like play house, construction Fig. 3. Climbing trees.
115The Natural Environment as a Playground
Fig. 4. Hiding and role-play.
cant differences between the groups, neither between the
sexes. Multiple regression analyses correlating test re-
sults with background variables, such as parents’ educa- Fig. 6. Ski-jumper.
tion and profession, showed that these variables had no
significant influence on the test results (Fjørtoft, in reach) were found within the experimental group. The
prep.). improvement within the reference group was not as
During the trial period a gradual improvement in striking (Table 1).
motor ability was observed in the experimental group. Comparing the groups at the posttest, significant
The children became strikingly better at mastering a rug- differences in favor of the experimental group were
ged ground and unstructured landscape. The impact of found in the Flamingo balance test (p <.001) and the
the environment on the children’s motor ability was doc- Indian skip coordination test (p <.01). Figure 7 shows
umented in the motor fitness tests. Table 1 and Figure 7 the interference effects from pre- to posttest in both
show the main test results of motor development in the groups, showing a significant better improvement in the
groups. experimental group in those two items.
At the pretest the reference group scored better than
the experimental group (Table 1). At the posttest the
experimental group had caught up with the reference DISCUSSION
group and significant differences between the pre- and This study has described the relationship between
posttest in all the test items except for flexibility (sit and the structure and functions of a natural landscape, its
affordances for play, and the impact on motor develop-
ment in children. A significant relation between the di-
versity of the landscape and the affordance of play was
indicated (Fjørtoft & Sageie, 2000). As described by
Gibson (1979) the affordances of a landscape are what
it offers the child. As the child perceives the functions
of a landscape and uses it for play, the landscape might
have a functional impact on children’s behavior and play
performance. As maintained by Moore and Wong
(1997) the physical diversity increases the opportunities
for learning and development. This was also verified by
the findings of the present study. The motor fitness tests
showed a general tendency that the children using the
forest as a playscape performed better in motor skills
than the children on the traditional playground. At theFig. 5. Climbing rocks.
Table I. Pre- and posttest. Mean results (SE) within the groups SPSS T-test for paired samples
EXPERIMENTAL GROUP REFERENCE GROUP
TESTS PRE-TEST POST-TEST PRE-TEST POST-TEST
FLAMINGO (no. of instabilities in 30 sec.) 4.7 (0.8) 1.5 (0.3)*** 4.0 (0.6) 3.3 (0.7)
PLATE TAPPING (Time in sec. of 50 taps) 35.0 (1.9) 28.1 (1.2)*** 29.9 (1.1) 27.4 (2.6)
SIT AND REACH (cm) 24.9 (0.8) 24.4 (0.8) 25.3 (1.0) 25.5 (0.9)
STANDING BROAD JUMP (cm) 102.8 (2.9) 113.1 (3.6)*** 103.1 (4.3) 111.3 (3.8)**
SIT-UPS (reps. in 30 sec.) 5.3 (0.6) 6.5 (0.6)** 5.9 (0.8) 7.0 (1.1)
BENT ARM HANG (sec.) 2.6 (0.4) 7.0 (1.0)*** 2.6 (0.6) 5.4 (1.1)***
BEAM WALKING (sec.) 11.4 (1.4) 7.5 (0.7)** 7.7 (0.8) 7.2 (1.1)
INDIAN SKIP (reps. in 30 sec.) 21.8 (2.2) 43.6 (1.9)*** 27.8 (2.4) 37.2 (1.8)***
SHUTTLE RUN (sec.) 31.9 (0.7) 29.7 (0.5)** 30.7 (0.8) 30.3 (0.7)
** =p<.01, *** =p<.001.
pretest the experimental group started lower than the ref- other variables such as leisure activities, were outside
the control of this study. However, the parents’ socio-erence group, but scored better in all test items at the
posttest (Table 1). This result makes it highly desirable economic background did not have any influence on the
test results and there was no reason to anticipate oneto make causal inferences, and according to Robson
(1993), when there is statistical significance, it is reason- parent group being more active outdoors than the other.
It is generally accepted that people in the countryside inable to conclude that it is the independent variable (play-
ing in the forest), which have affected the dependent Norway have similar opportunities for leisure activities
and there is a somewhat democratic distribution of atten-variable (motor fitness). This amplifies the impression
that the experimental group improved more during the dance to sports and leisure activities in the population
(Wichstrøm, 1995).intervention period than the reference group. Significant
differences were noticed between the experimental A study carried out by Grahn et al. (1997) showed
a similar correlation between the physical playscape andgroup and the reference group in balance and coordina-
tion at the posttest as illustrated in Figure 7. Growth motor abilities. The study design was more like a case
study including two kindergartens with different outdoorand maturation in the children may have affected these
results. The anthropometrical measurements, however, playgrounds. One kindergarten had access to natural en-
vironment within the playground area while the othershowed no differences between the groups, neither did
age and sex. It should therefore be reasonable to con- kindergarten had a more traditional urban playground.
The EUROFIT Motor Fitness Test was applied and thesider the gain in motor fitness in the experimental group
to be related to versatile play in a stimulating playscape. results showed a significantly better performance in the
natural play area group than the traditional group.Whether these effects might have been influenced by
Grahn’s study supports the findings of the present study
and jointly the two studies indicate the positive impact
of the natural environment on children’s motor develop-
ment. Playground studies confirm the significance of di-
versity in play equipment on children’s play behavior.
The more equipped, the more versatile and creative the
play (Frost & Sunderlin, 1985, Frost, 1992), and Moore
and Wong (1997) observed that the repertoire of chil-
dren’s behavior broadened enormously with the increase
in physical diversity of the environment. Analysis of
landscape ecology and topography for the study area de-
scribed a varied and diverse playscape, and the study
Fig. 7. Intereference effects. showed a strong relation between landscape structure
Pre- posttest results. and play functions (Fjørtoft & Sageie, 2000). The study
Flamingo balance: Exp. group: 3.8* (0.3), Ref. group: 0.9 (0.6). Indian of Titman (1994) also confirms the children’s needs for
skip: Exp. group: 20.6* (2.2), Ref. group: 10.5 (2.1).
green grounds, trees to climb and shrubs for shelter and
117The Natural Environment as a Playground
Frost, J. L. and Sunderlin, S. (Ed.). (1985). When children play.Pro-
building of constructions. A natural landscape is synon- ceedings of the international conference on play and play envi-
ymous with an enriched environment, which again stim- ronments. Wheaton, MD: Association for Childhood Education
International.ulates and promotes play and learning (Rivkin, 1995).
Frude, N. (1993). A Guide to SPSS/PC+. (2nd ed.). London: Mac-
Results from the present study add information to the millan.
significance of outdoor arenas, the possibilities for phys- Gibson, J. (1979). The ecological approach to visual perception. Bos-
ton: Houghton Mifflin Gillberg, C. and Rasmussen, P. (1982).ical play, and the impact on children’s development.
Perceptual, motor and attentional deficits in six-year-old children.
Acta Paedagogica.Scandinavia,71, 121–9.
Grahn, P., Ma
˚rtensson, F., Lindblad, B., Nilsson, P., & Ekman, A.
CONCLUSION (1997). UTE PA
˚DAGIS. Stad & Land nr. 45. Alnarp: Movium,
This study has indicated the relation between versa- Hansen, E. B. (1999). Løkke og lag.Hvordan barn trener i dag? Oslo:
tile play in the natural environment and the impact on Norges idrettsforbund og Olympiske komite.
Hart, R. (1979). Children’s experience of place. New York: Irvington.motor fitness in children. Significant effects were found
Heft, H. (1988). Affordances of children’s environments: A functional
in balance and coordination abilities. These are compe- approach to environmental description. Children’s Environments
tencies that are of great importance to the children’s Quarterly,5(3).
Hertzberg, A. (1985). Prediction of cervical and low-back pain basedgeneral mastering of their own body in relation to the
on routine school health examinations. Scandinavian Journal of
physical environment. There is a strong relation between Primary Health Care,3, 247–53.
the structures of the landscape and the functions of play. Kjos, R. (1992). Motorisk test av 5-a
˚ringer. Fysioterapeuten,8, 15–17.
Mjaavatn, P. E. (1999). Splitthopp—Epleslang og Fotball.En rapportThe forest itself represents an environment for play and
om barn og egenorganisert fysisk aktivitet. Oslo: Norges Idrettsf-
learning that stimulates motor development and foster- orbund og Olympiske komite.
ing in children. Referring to Bjerke (1994) we still know MMI (1995). Mediebruksundersøkelsen.Barn og ungdomsunder-
søkelsen 8–24 a
˚r. Oslo: Markeds-og Mediainstituttet a.s.too little about the learning effects from the natural envi-
MMI (1997). MMI’s landsrepresentative undersøkelse blant barn 3–7
ronment, and more effort should be dedicated to further a
˚r. Oslo: Oppdrag fra Norges Idrettsforbund og Norges Olym-
studies in this field. piske Komite.
Moore. R. C. (1986). Childhood’s domain.Play and space in child
development. London: Croom Helm.
Moore, R. C. and Wong, H. H. (1997). Natural learning.CreatingREFERENCES
environments for rediscovering nature’s way of teaching. Berk-
ley, California: MIG Communications.Adam, C.V., Klissouras, M., Ravazollo, R., Renson, W. and Tuxworth,
W. (1988). EUROFIT: European Test of Physical Fitness.Hand- Norusis, M. J. (1993). SPSS for Windows.Base system user’s guide.
Release 6.0. Chicago: Marketing Department, SPSS Inc.book. Rome, Italy. Council of Europe, Committee for the Devel-
opment of Sport. Committee of Experts on Sports Research. Robson, C. (1993). Real world research.A resource for social scien-
tists and practitioner-researchers. Oxford: Blackwell.Anderson, R. E., Crespo, C. J., Bartlett, S. J., Cheskin, L. J., & Pratt,
M. (1998). Relationship of Physical Activity and Television Ropeid, K. (1997). Sma
˚barnskroppene ødelagt av stillesitting. Norsk
førskolelœrerblad nr. 2- pp. 6–7.Watching With Body Weight and Level of Fatness Among Chil-
dren. Results From the Third National Health and Nutrition Ex- Rivkin, M. S. (1990). Outdoor play —What happens here? In: Wor-
tham, S. and Frost, J.L. (Eds.), Playground for young children.amination Survey. JAMA,279 (12), 938–942.
Bang, C., Braute, J. & Kohen, B. (1989). Naturleikeplassen.Ein stad National survey and perspectives. A Project of the American As-
sociation for Leisure and Recreation. An Association of thefor leik og lœring. Oslo. Universitetsforlaget.
Bjerke, T. (1994). BARN—NATUR—FRILUFTSLIV. Oslo. A/S Land- American Alliance for Health, Physical Education, Recreation
Due, P., Holstein, B., Marklund, U. (1991). Selvrapportert helbred Rivkin, M. S. (1995). The great outdoors.Restoring children’s rights
to play outside. Washington D. C.: National Association for theblant skoleelever i Norden. Nordisk Medisin,90, 1452–60.
Fjørtoft, I. (1999). Naturen som lekeplass. Rapport fra konferansen Education of Young Children.
¨m, M. (1998). Ger friluftsliv friskare barn? Nordisk Konfer-Forskning i Friluft. Friluftslivets fellesorganisasjon, FRIFO. Rap-
port nr. 1. ens Barn och Friluftsliv—friluftslivets pa
¨lsa och naturfo
˚else. Konferensrapport. Stockholm:Fjørtoft, I. (in press). Motor Fitness in Pre-Primary School Children.
The EUROFIT Motor Fitness Test explored on 5–7 years old Friluftsfra
Titman, W. (1994). Special places,special people.The hidden curricu-children. Pediatric Exercise Science.
Fjørtoft, I. and Sageie, J. (2000). The Natural Environment as a Play- lum of school grounds. London: WWF UK (World Wide Fund
for Nature)/Learning through Landscapes.ground for Children. Landscape description and analyses of a nat-
ural playscape. Landscape and Urban Planning,48, 83–97. Thomas, R.T. and Nelson, J.K. (1985). Research methods in physical
activity. Champaign: Human Kinetics.Francis, M. (1988). Negotiation between children and adult design val-
ues in open space projects. Design Studies,9(2), 67–75. Wichstrøm, L. (1995). HVEM? HVA? HVOR? Om ungdom og idrett.
En landsdekkende surveyundersøkelse. Kulturdepartementet,Frost, J. L. (1992). Play and playscapes. Albany, NY: Delmar Pub-
lishers. Oslo: UNGFORSK.