Identifying and characterizing risky play in the age one-to-three years

Abstract

While research has investigated risk-taking in play for children from the age of four years upwards, less is known of risky play with children under four years. A small-scale observational study with children from five childcare settings with differing characteristics was undertaken to explore the occurrence and characteristics of risky play for children under four years of age, in relation to the current understanding of risky play. The study found similarities across the different contexts, which seem to reflect the characteristics of risky play for children aged one to three years. The findings suggest that the existing definition and characteristics of risky play are appropriate for two- and three-year-old children, but for one-year-olds, the study found discrepancies indicating deviations from existing definitions, indicating that the concept may not be so useful for this age group. To develop understanding of risky play, this article suggests new categories and an adapted definition.

Full text

Identifying and characterizing risky play in the age one-to-three years
Rasmus Kleppe, Department of Early Childhood Education, Oslo and Akershus University College of
Applied Sciences, Oslo, Norway
Edward Melhuish, Department of Psychological Sciences Birkbeck University of London, London,
United Kingdom
Ellen Beate Hansen Sandseter, Department of Physical Education and Health, Queen Maud University
College of Early Childhood Education, Trondheim, Norway

1
SUMMARY: While research has investigated risk-taking in play for children from the age of
four years upwards, less is known of risky play with children under four years. A small-scale
observational study with children from five childcare settings with differing characteristics
was undertaken to explore the occurrence and characteristics of risky play for children under
4 years of age, in relation to the current understanding of risky play. The study found
similarities across the different contexts, which seem to reflect the characteristics of risky play
for children one-to-three years of age. The findings suggest that the existing definition and
characteristics of risky play are appropriate for two- and three-year-old children, but for one-
year-olds, the study found discrepancies indicating deviations from existing definitions,
indicating that the concept may not be so useful for this age group. To develop understanding
of risky play, this paper suggests new categories and an adapted definition.
Keywords: Risky play, exploratory play, children under three, one-year-olds, exploratory
study

1
Introduction
There has been growing interest in why children take risks and what the effects might be
(Aldis, 1975; Boyer, 2006; Byrnes, Miller, & Schafer, 1999; P. Christensen & Mikkelsen,
2008; S. Christensen & Morrongiello, 1997; Pellegrini & Smith, 1998; Readdick & Park,
1998; Sandseter, 2010b; Sandseter & Kennair, 2011; S. J. Smith, 1998; Stephenson, 2003).
The literature defines risk-taking as actions with a probability for undesirable results or
negative consequences, and suggests that the abilities to understand situations; to assess own
capabilities; and to avoid excessive risks, are important for development. Children express
and practice these abilities typically as play, hence the term risky play (Sandseter, 2007).
Consequently, also the intrinsic value of play must be considered (Lillemyr, 2009; Sutton-
Smith, 1997).
Although it is well established that children from four years and up take risks in play, little is
known of younger children’s risky play (Bjørnestad et al., 2012, p. 21). With a global increase
in enrollment of children under three in Early Childhood Education and Care (ECEC) (Engel,
Barnett, Anders, & Taguma, 2015), we need knowledge of all aspects of children’s play in
such contexts. Hence, this paper investigates the occurrence and characteristics of risky play
by children in the age range one - three years in ECEC-contexts and considers whether
existing definitions of risky play apply or if alternative definitions or adaptations are
necessary.
Previous research
Children’s risk-taking in play has been studied since the 1970’s (Aldis, 1975; Bruner, 1976),
but it is under-developed with limited consensus on definitions. However, the literature
indicates that children’s risk-taking in play has common characteristics. Generally, research
suggests that risk-taking in play imitates real-life risks through play (Aldis, 1975; Sandseter &
Kennair, 2011), and includes curiosity, exploration, deep concentration, fear and excitement.
Children explore their surroundings and their capabilities through trial and error, and their
behavior involves a balancing act between exhilaration and fear, as the child either masters
the challenge or withdraws because of fear (P. Christensen & Mikkelsen, 2008; Cook,
Peterson, & DiLillo, 1999; Sandseter, 2010a). This “edgework” (Lyng, 1990) involves the
child increasing the risk, e.g., climbs a little higher each time or ventures further from the
adult (Aldis, 1975; S. J. Smith, 1998). Although fear has its natural place in risky play, the
literature tends to focus on fun and thrill, with overt sounds and body language such as
screams, laughs and big movements (Mårtensson, 2004; Readdick & Park, 1998; Sandseter,

2
2007). Sutton-Smith (1997) sees fun and exhilaration as strong motivational factors conducive
to repetition of some risky play. In this vein, risky play is linked with vigorous physical
activity, specifically sliding, swinging, climbing, bike riding, balancing over drops, jumping
down, chasing and play-fighting, shooting with bows and arrows, rolling on the ground and
whittling sticks (Hughes, 2012; Kaarby, 2005; Sandseter, 2010b; S. J. Smith, 1998;
Stephenson, 2003). Notably, research suggests that such vigorous physical activities happens
more outdoors (Aarts, Wendel-Vos, van Oers, van de Goor, & Schuit, 2010; Cosco, Moore, &
Islam, 2010; Storli & Hagen, 2010), and risky play is seen typically as outdoor play
(Staempfli, 2009; Stephenson, 2003; Tovey, 2007). Rough-and-tumble play is regarded as
risky play by researchers as it has the potential for (unintentional) harm (Blurton-Jones, 1976;
Humphreys & Smith, 1984; Pellegrini & Smith, 1998; P. K. Smith, 2005). Bringing these
perspectives together, Sandseter (2010b) offers this definition: “[risky play] involves thrilling
and exciting forms of physical play that involve uncertainty and a risk of physical injury”
(2010b, p. 22). Additionally, she identifies six categories of risky play: 1) Play with great
heights (danger of injury from falling), 2) play with high speed (uncontrolled speed that can
lead to collision), 3) play with dangerous tools (that can lead to injuries), 4) play near
dangerous elements (such as fire, water or heights), 5) rough-and-tumble play (where children
can harm each other), and 6) play where the children can get lost.
Thus research links risky play to exploratory behavior and an observable balancing act
between fear and exhilaration. There is an emphasis on vigorous physical activity, overt
bodily expressions, fun and thrill, mostly outdoor activities and risk of physical injury.
Conceptualizing risk and play
Historically, risk-taking behavior is regarded as something to be avoided (Boyer, 2006; Lyng,
1990; Malaby, 2002). Certainly, risk-management per se is not concerned with balancing cost
and benefit but with reducing risk (Adams, 2001, p. 16). Similarly, risk-taking from a
psychological perspective has focused largely on maladaptive social functions named “the
prevailing developmental psychopathology model” (Ellis et al., 2012, p. 598). In contrast,
literature on risky play focuses largely on the intrinsic value and learning potential of the
behavior. This notion, that risk-taking is part of life and has both positive and negative effects,
and should therefore be investigated more comprehensively, seems now to have wider
influence (Boyer, 2006; P. Christensen & Mikkelsen, 2008; Ellis et al., 2012). Adams (2001)
suggests expanding the understanding of risk by distinguishing between objective and
subjective risk. Objective risk involves pre-defined, observable or measurable risk, while

3
subjective risk involves how individuals perceive risk in different situations. Sandseter
(2009a) suggest that objective risk can be observed as the environmental characteristics of the
situation, e.g. height, speed, unstable surfaces, etc. Subjective risk can be observed as
individual characteristics, i.e. how the child expresses its experience through body language,
facial expressions, sounds or words. While exploring objective risks, the child will adjust its
subjective experience and expressions (Aldis, 1975; Apter, 1992; Sandseter, 2009b), a process
that can be interpreted as self-regulation (Byrnes, 2013). Applications of Vygotsky’s concept
of zone of proximal development (1978, p. 84), resonate with this, as children explore their
surroundings, and, by giving themselves increased challenges, create their own zone of
proximal development (Johnson, Sevimli-Celik, & Al-Mansour, 2012).
While there are conflicting perspectives (Lillemyr, Dockett, & Perry, 2013), there are
common characteristics in play theories, namely that play is intrinsically motivated, voluntary
and “purposeless”, meaning that the activity in itself is more important than its ends (Johnson
et al., 2012; Lillemyr et al., 2013). This links play and learning, with the implication that an
activity can be simultaneously purposeless and functional. This paradox (Martin & Caro,
1985) can be resolved in recognizing that play has both immediate and mediate effects,
simultaneously bearing intrinsic value and learning potential (Lillemyr, 2009; Pellegrini,
Dupuis, & Smith, 2007). Additionally, for lack of research under four years, this study needs
terminology for play in relation to age. As with play and learning, this dimension yields
conflicting theoretical positions, such as the assumption that play follows a universal,
sequential, age-related developmental trajectory (Pellegrini & Smith, 1998; Piaget, 1954; P.
K. Smith, 2005), as opposed to the concept of play as a culturally situated and complex
phenomenon (Engdahl, 2007; Løkken, 2000; Merleau-Ponty, 2012). This complex debate is
omitted here, and, rather, both perspectives are applied pragmatically, in aiming at descriptive
identification and characterization of play at certain ages.
Generally, we need to be circumspect regarding any preconceptions from previous research
on older children. Thus interpretations of observations here are based on a basic
understanding of risky play: “[…] risk taking involves the implementation of options that
could lead to negative consequences.” (Byrnes et al., 1999, p. 367). Risk is unavoidable, and
has both negative and positive effects, so, risk-taking in play can be potentially valuable, for
the child to both experience the potential excitement and joy, and providing practice in
dealing with risky situations, which are sometimes inevitable.

4
Method
In exploring a new phenomenon, qualitative approaches with few participants are often
recommended (Creswell & Plano Clark, 2011; Johannessen, Tufte, & Christoffersen, 2010, p.
114) and the ethnographic role as participant observer is emphasized as particularly suitable
to gain insight into children’s lives (Corsaro, 2003; Gulløv & Højlund, 2003; James & Prout,
1997; Lange & Mierendorff, 2009). Hence, in this study, so-called short-term ethnography
(Pink & Morgan, 2013) was chosen as the main data-collection technique. Pink and Morgan
(2013) suggest that this method is especially appropriate within theoretically informed,
applied research. In contrast to traditional ethnography, where the ideal is long-term
participation, this exploratory study has a narrow focus, and applies less intrusive and time-
consuming data-collection (Knoblauch, 2005; Millen, 2000). Still, the purpose of the
ethnography is to obtain rich or thick descriptions (Geertz, 1994). To achieve this,
observations should include many situations, also situations normally occurring outside of the
staff’s view, and Corsaro (2003) suggests that this can be obtained by behaving differently
from regular staff. From the start, the staff generally appeared playful and involved with the
children and the role of “detached observer” was chosen (Gulløv & Højlund, 2003, p. 40).
Carrying a notebook and a video camera strengthened the position as both different and
detached. On day 3 of data collection, Daniel (3) indicated the achievement of this role: “[…]
points at me and shouts: Look! He’s not an adult! Because … he does not have children! He
is a child! He does not have children, hah!” (Video 0016, ECEC Center 3 (nature center),
Day 3).
In line with recommendations for short-term ethnography, multiple data collection techniques
are utilized. Together with ethnography, mapping is applied to provide an overview of
complex situations and comparable data on extent and context (Cosco et al., 2010). Video is
applied to increase the level of detail in descriptions (Knoblauch & Schnettler, 2012).
Participants
There are several ways to select participants for a qualitative study. According to Seawright
and Gerring (2008), selecting qualitative samples has the same “twin objectives as random
sampling; that is, one desires (1) a representative sample and (2) useful variation on the
dimensions of theoretical interest.“ (2008, p.296). In this study it was realistic to prioritize the
latter and five ECEC center groups were included. Two ECEC center groups were selected

5
from the BePro-sample (BePro, 2013)
1
(Norwegian longitudinal study, including 206 ECEC
center groups) based on their varied scores (high and low) on the Infant Toddler
Environmental Rating Scale (ITERS-R) (Harms, Cryer, & Clifford, 2006). The ITERS-R does
not address risky play directly, but several items touch on the topic, such as physical activity,
supervision and safety. The ECEC center groups varied scores strengthen the potential for
generalization, meaning that similarities in different contexts are more likely to represent
general patterns (Gobo, 2008). For specific purposive sampling reasons (Teddlie & Yu,
2007), two nature ECEC centers were included, where children spend most of their time
outdoors, in a natural environment, and offer increased probability of relevant observations.
Research suggests that vigorous physical activity, and therefore risky play, will occur more
often outdoors (Aarts et al., 2010; Cosco et al., 2010; Sando & Lysklett, 2012; Storli &
Hagen, 2010). For similar reasons, one infant-toddler group was included. After the first
seven days of data-collection, the observations indicated deviations amongst one-year-olds
from the predominant understanding of risky play. Therefore, to strengthen detailed
descriptions, it was decided to observe only one-year-olds for parts of the remaining data-
collection.
Groups consisting of one-to-three-year-olds were observed for 10 days. One-year-olds, aged
1,1 to 1,11 years, were observed for 4 additional days. The groups were observed between
August and February the following year and the participants consists of 28 boys and 25 girls,
with 26 one-year-olds, 20 two-year-olds and 7 three-year-olds. The low number of three-year-
olds reflects Norwegian practice, where children move to the older age group within the
semester they turn three.
Ethical considerations
Research is necessary to obtain knowledge for the welfare of children. Observing children is
therefore sometimes necessary, but measures should be made to secure the rights and integrity
of study participants. The study adheres to all ethical standards and privacy policies of the
Norwegian Social Science Data Service and Norwegian Data Protection Authority, which
ensures participants’ confidentiality and anonymity. The approval presupposes informed
consent from all parents of children, which was obtained. Still, children themselves should
have a say, and the possible experience of intrusiveness was of high priority. The staff would
inform the children of a visit by a stranger and the purpose of this visit, to the best of the
1
The ITERS-R data used in this article are acquired through two projects funded by the Research Council of
Norway, ‘Better Provision for Norway’s Children in ECEC’ (BePro) and ‘Searching for Qualities’.

6
children’s comprehension. Most importantly, the children could give “ongoing consent”
(Flewitt, 2005, p.556), meaning that if a child showed signs of discomfort related to the
presence of the observer, the observer would withdraw. In addition, the study’s focus is risky
play and there would be occasions where children might be physically injured. In such cases,
continuous judgement was necessary to decide whether to intervene, and avoiding injury was
given priority over the role as detached observer. No such situations occurred during the
observations.
Observations
The full groups were normally observed throughout the day, for about 7 hours. With few
children in each group and/or physical limitations, e.g., fences or closed doors, it was possible
to observe all children most of the time. In all five ECEC center groups, children’s everyday
life had several routines, such as diaper change, meals and naps. The children were observed
in all activities and transitions between activities. However, the major issue was to determine
whether a behavior could be characterized as risky play, so describing all activities would be
infeasible and unnecessary. Therefore, any situation that was perceived potentially dangerous,
either by the child, staff or the observer was mapped and described to answer two basic
questions: Are there environmental or individual characteristics in the situation that indicates
risk? And: Can the child’s behavior be characterized as play? For the sake of interpretations,
the descriptions elaborate on actions, facial expressions, body language, voice/sounds and
verbal expressions of both staff and children. In addition, the following information was
collected for each instance of risky play.
• Who – with codes for individuals, gender and age. In this paper, children have a
fictitious name and age in brackets, e.g. Lene (1,5).
• What – with descriptions of activities leading to coded categories.
• Staff reactions/involvement – with description of interaction to be coded for later
analysis. Here staff have a fictitious name and staff level in brackets. Teacher = (T),
Assistant = (A), e.g. Espen (T).
• Location – with codes for Inside/outside
• Sociability – with codes for Alone/Together
• Duration – with codes for Long/Short

7
Video-recordings were done for two days, one day in a nature ECEC center and one day in
the infant-toddler group. The videos were coded similarly to the field notes (Knoblauch &
Schnettler, 2012).
Mapping
The purpose of the mapping is to provide supplementary information to the qualitative
descriptions; to establish to what extent risky play occurred and to collect comparable,
contextual data (Cosco et al., 2010). Several mapping tools were reviewed, but all instruments
were missing terminology related to risky play. Alternatively, risky play could be defined as
vigorous physical play, a category in several instruments. However, if the children in focus
would display other types of risky play, relevant observations would be missed. Therefore, a
mapping tool was developed for the present study. The categories of the mapping reflect the
codes described earlier, and could be represented quantitatively.
The mapping was piloted in two ECEC centers to investigate the relevance of the codes and
the usefulness for observations and analysis. Small adjustments were made to the mapping
format throughout without altering the basic content so that early and later mappings are
comparable. Since the mapping has not been subjected to inter-rater reliability, the mapping
will not be emphasized as evidence as such, but as support for the general patterns and
descriptions.
Analysis
From the first days of observations, children were observed playing in ways that could be
identified within existing definitions of risky play, but there were also observations of
children in similar play without experiencing any risk, and sometimes vice versa; children
experienced risk without showing any thrill or there was no risk of injury. To determine
whether the play could be characterized as risky, the two criteria of environmental and
individual characteristics (objective and subjective risk) were applied in the analysis. This was
combined with previous observations of children’s behavior in similar situations. Staff
reactions were considered in assessing environmental characteristics, as an outside observer
might overlook some risky aspects of a situation.
Analytic sample
The present paper’s basic variable for analysis is Instances of risky play. One “play” or game
counts as one instance and can include many children and/or repeated risk-taking. For
example, if a group of children were chasing each other, and at the same time climbing and

8
play fighting with sticks, this would count as one instance. Repetitive play such as sliding or
swinging would also count as one instance. All instances were coded as described earlier with
Who, What, Staff reactions/involvement, Location, Sociability, Duration, for statistical
analysis.
This gives a sample where n represents the total number of instances of risky play observed in
12 days. Comparison of these instances across different contexts is made feasible through
precise observational criteria and coding (Gobo, 2008). As mentioned, on three of the days,
only one-year-olds’ risky play was observed and mapped, even if there were two- and three-
year-olds present. If summed together, this would over-estimate one-year-olds’ number of
instances compared with older peers. Therefore, these three days cannot be compared
statistically with the observations of the full groups, and are considered as a separate sample.
This gives two samples of Instances of risky play: Sample 1, including children one-to-three
years (N=198), Sample 2, with one-year-olds only (N=46).
Individual children were described and mapped in detail, including age in months. However,
in the statistical analysis, age categories are one-, two- and three years. Individual differences
in age-related development are more nuanced than this, but splitting age into months in the
statistical analysis would give very small numbers in each age group. Moreover, one main
finding is related to the ability to walk. Since this ability normally is developed in the second
year and stabilizes in the third year (Goodway, Ozun, & Gallahue, 2012), this categorization
was considered sufficiently detailed.
Findings and discussion
Regarding appearance and content of play, there are variations when it comes to how each
child expresses itself and engages in risky play. However, based on the described criteria,
risky play was observed in the age group one-to-three years in all five ECEC center groups on
all days of observation. The mapping and descriptive similarities found across different
contexts, suggest consistent patterns. Number of instances varied between 12 and 28 in one
day, with no apparent differences between nature ECEC centers and ordinary centers. These
patterns can be regarded as characteristics of risky play in the age group one-to-three years.
Common characteristics of risky play in the age group two-to-three years
According to the predominant understanding, playing with risk involves a thrill or excitement,
described by the children themselves as “it tickles in the tummy” (Sandseter, 2010a, p.76). It
can also be identified through overt expressions of excitement, fear or exhilaration (Aldis,

9
1975; Sandseter, 2009b; Stephenson, 2003). These characteristics make risky play relatively
easy to identify, also in the present study:
Example 1: Sondre (2,9) and Daniel (3,3) (check names) are climbing on the big
snowballs, bouldering (the balls are about their size and there is a whole
circle/structure of them). They climb up, try to jump from one to another or slide or
jump off. Daniel jumps off several times and slides down the “high wall”. He shouts: I
drove fast! I drove the fastest! Wasn’t that fun?! Sondre climbs to the top of the wall,
but says with a tiny voice “No” and climbs down [he is discouraged]. He slides off
from a lower boulder. He watches Daniel as he slides again from the higher boulder,
and Daniel looks back up at him and assures: I didn’t break my legs! Daniel goes on
to reassure Sondre that he dares: “It is not big!” Sondre laboriously gets in position
and mumbles to himself (I do it, I dare this) and off he goes. At the bottom he shouts: I
dared, I dared! …. I dared slide down there! He walks back into the circle of boulders
while he repeats to Daniel: I dared! (Video 0016, Nature ECEC Center 3, Day 3)
This episode is interpreted as play based on its voluntary appearance, exhilaration and
repetitiveness; the play goes on for more than twenty minutes. Moreover, it is interpreted as
risky play based on the combination of environmental characteristics (objective risk),
individual characteristics (subjective risk) and increase of risk. The environmental
characteristics are the height of the boulders (double the children’s height) and the steep
incline, which gives high speed. The risk of physical injury is even adressed by the children.
The individual characteristics are firstly attributed to the fun and thrill the boys express
through their body language and excited cheers. Secondly, the subjective risk can be observed
as Sondre is balancing between exhilaration and fear. First, he slides from a lower boulder,
and while increasing the risk, by going higher, he hesitates and withdraws with fear. He
expresses both with body language and in words that he does not dare. His voice is low; his
face is towards the ground and back slightly sunk, which indicates anxiety and maybe
disappointment. Daniel continues to address the risk and simultaneously reassures Sondre that
there will be “no broken legs” and “you dare”. Sondre’s body language continues to show
hesitation, but he moves into position on top of the boulder. What he mumbles to himself is
interpreted as his mental approach; he repeats Daniel’s words of encouragement just before
going over the edge. Sondre’s reaction after sliding down confirms his exhilaration and
reward of mastering the challenge: “I dared!”

10
As both individual and environmental characteristics confirm the situation as risky play, the
aspect of increasing the risk is interesting. The rewarding thrill and exhilaration of mastering
seem to be a strong motivational factor in risky play (Sandseter, 2009b). As the child
progressively masters an increased objective risk, the subjective risk and therefore the
exhilaration decreases. Consequently, the objective risk is increased, to optimize exhilaration
(Apter, 1992; Sandseter, 2009b). This is typically observed as children each time climb a little
higher or venture further from the staff (Sandseter, 2009a). As in Sondre’s example, children
were observed doing this throughout the observations. For example, while sliding, they would
start out sliding sitting upright, and then continue to slide on their back and eventually on their
stomachs, head first. This can also be interpreted as a learning aspect of risky play, where the
child is motivated (and where the environment allows), to constantly increase the challenges
within the changing zone of proximal development.
Duration and sociability – short and long play
A prominent finding is the briefness of many instances of risky play. These instances put the
observer to the test because the situations are literally over in a few seconds. One typical
example of such play would be:
Example 2: Fredrik (2,1) is walking around by himself next to the fireplace outside the
main building. He walks carefully up a rock on the ground; the rock is pointy and
about 30 cm high. He gets to the top and says “Ooooi”, stands up and stretches his
arms out to the side. He has a big smile. He loses his balance slightly, catches himself
by crouching quickly, and then jumps off. (Field notes, Nature ECEC Center 3, Day 1)
This way of playing and using the environment seems typical for the age group. Parts of the
day, the children wander about, and they engage with anything they might come across. In
most of the established categories, be it playing with height, speed, tools, elements and even
rough and tumble, risky play comes in these brief intermezzos, as part of exploring or
engaging with their surroundings. It is identified as risky play mainly due to the individual
characteristics, for example, the thrill the child expresses. In Example 2, Daniel is careful
when walking up the rock and the reward of reaching the top is obvious in his big smile.
While a fall from this height might not lead to injury, he would probably feel pain, and the
risk is apparent by him almost falling and catching himself on the way down. This situation
lasted just under 30 seconds.

11
The rule for coding an instance short was that it lasted approximately 1 minute or less. Any
play lasting longer than 1 minute was coded as long. The reason for this was the distinct
briefness of many situations. Situations lasting for two minutes and longer, even up to 30
minutes, had more similarities between them than with the very short ones. The similarities
include that they often involve two or more children and sometimes staff; the play often has
components of role-play, and rough and tumble. There are also longer sessions with more
repetitive play such as swinging or sliding. This type of play does not have the social features
of role-play or rough and tumble, but is often sociable, meaning the children play two or more
together, for example swinging. The mapping shows that long sessions of play are dominantly
social (71%), while few of the short sessions are social (19%).
In summation, two-to-three-year olds exhibit risky play in much the same way as described in
previous research. Both the objective risk is apparent in the environmental features and the
subjective risk is apparent in the individual bodily expressions; overt and easy to identify.
When given the opportunity they engage in play with height, speed, dangerous tools and
elements, rough and tumble play and a few instances of running away or hiding from the staff
(disappear/get lost). Three categories stand out: Playing with speed (25%), Rough and tumble
play (12%) and Playing near dangerous elements (39%) (Table 1).
Table 1: Categories of play 1-3-years (Sample 1)
Frequency
Percent
Valid
Height
15
7.6
Speed
49
24.7
Impact
15
7.6
Rough'n'Tumble
24
12.1
Tools
8
4.0
Elements
77
38.9
Run_away
4
2.0
Vicarous risk
3
1.5
Other
3
1.5
Total
198
100.0
Common characteristics of one-year-olds risky play
Where two-to-three-year-olds’ risky play largely resembles previous research, similar
situations involving one-year-olds could appear different, either based on the environmental
or the individual characteristics, or both:

12
Example 3: The group has just finished eating and Nicolai (1,7) goes over to the
“balance bowl”, it is a flat bowl, slightly concave, approx. 10cm deep and approx.
50cm across. It is now turned over on the floor, forming a low convex structure.
Nicolai climbs up, hands and feet on the bowl. Safely on top, he tries to raise to a
standing position, but gives up and slides off. He makes no sounds and keeps a stern
face throughout. Sandra (T) puts Celine (1,3) on the floor (she has been sitting by the
table). She crawls quickly and determined to the bowl, crawls up on it. When on top,
she just sits there. Face blank, watches a bit around. She then crawls off after 1 min
and then crawls back up. At 00:10, she almost slides off and catches herself. She then
continues to climb and move around the top for a while. (Video 0031-34, Infant-
toddler group, Day 1)
Throughout the observations, children under two years were in situations where the objective
risk could be identified, but where the individual characteristics were not observed as
expected. Children would climb, slide or swing with very little overt body language of
hesitation, thrill and excitement. Sometimes, also the objective risk, observed as the
environmental characteristics in Example 3, proved difficult to identify. Here, the low,
slightly convex structure can easily be overlooked as representing any risk, certainly no risk
of injury. Rather, the risk is attributed to two individual characteristics in the example. Firstly,
Nicolai is a steady walker, but as he increases the risk by trying to stand up on the bowl, he
withdraws. Several other one-year-olds attempted this and some succeeded. Secondly, when
Celine is put on the floor she crawls (she cannot walk) directly and eagerly to the bowl and on
to the top, i.e., she seems highly motivated. Whether she is experiencing fear is impossible to
interpret from her body language until she almost falls off. She catches herself quickly and
moves to safety on top. She then continues to move around on the bowl. The movement of
catching herself is interpreted as a subjective risk, an experience of fear, even if the
experience is very brief and the fear probably not strong.
As in Example 1 and 2, Nicolai’s and Celine’s behavior is interpreted as play based on its
voluntary appearance and intrinsic motivation. However, the excitement described in
Example 1, is not observed. This resonates with age-related theories of physical development
and play. Pellegrini & Smith (1988) describes this sequential; starting in infancy with
rhythmic stereotypes, i.e. gross motor movements with no apparent purpose. In the preschool
age, exercise play, i.e. gross locomotor movements in the context of play, is predominant,
while rough and tumble play occurs increasingly in the late pre-school years and is seen as the

13
predominant physical play in the primary school age. While exercise play can be both solitary
and social, rough and tumble has a distinct social character. This describes a general trend in
age-related play development; from partly solitary and repetitive play to predominantly social
and more complex play (Goodway et al., 2012; P. K. Smith, 2005; Sutton-Smith, 1997). Seen
this way, risky play follows a general pattern, and Celine’s and Nicolai’s play can be observed
as subtle, i.e. less sociable, less extrovert, more repetitive and more focused because of their
age, specifically related to their agility. The more agile a child would be, the more his/her
risky play would resemble the existing definition. The mapping supports the assumption of
age-related play, suggesting that children, with age, increasingly engage more in long and
social play. However, one- and two-year olds were observed involved in long sessions of
social risky play, such as rough and tumble, and the play appears both social and complex.
Therefore, the concept of exercise play seems too narrow, and, rather, the social toddling
style, as described by Løkken (2000) and Engdahl (2007), seems a more appropriate
description of under-three’s risky play, including one-year-olds.
Extent of risky play
The mapping also suggests differences between one-year-olds and two-to-three-year olds
regarding the extent of involvement in risky play. Of the 198 instances of risky play in
Sample 1, one-year-olds were involved in approximately 25% of the instances and two- and
three-year olds were involved in the remaining 75% (Table 2). Similarly, involvement is
highest among two- and three-year olds, with involvement up to 17 instances in a day. No
one-year-old was observed being involved in more than 10 instances in a day. Additionally,
among the one-year-olds there were several that did not involve in risky play, while among
the two- and three-year-olds there was no individual with less than 2 instances in a day. These
figures are skewed in favor of the older children due to higher presence on days of
observation, not only because of fewer one-year-old participants in Sample 1, but also
because one-year-olds sleep and participate in more routine care. The one-year-olds simply
have less time to play. Still, the differences in frequency of risky play in Sample 1 remain
large. Comparing the average occurrence of risky play per day in Sample 1 and Sample 2
suggests the same, 25 instances per day in Sample 1 and 15 instances per day in Sample 2.
Table 2: Involvement/Age (Sample 1)
Frequency
Percent
1 year
40
20.2
2-3 year
140
70.7
Mixed group
18
9.1

14
Total
198
100.0
Adaptation of categories
Initially it was presumed that existing categories might be inappropriate, therefore,
categorizing was avoided until the in-depth analysis. Indeed, in contrast to the two- and three-
year olds, one-year-olds play proved difficult to categorize. As described above, much
observed play could not be identified within existing categories based on environmental or
individual characteristics. However, the observations indicated subjective risk, albeit subtle.
Therefore, in line with the short duration of many instances (Example 2), the exploration of
objects/surroundings was added as an individual characteristic, and the name of the category
was changed from Playing near dangerous elements to Playing with dangerous elements.
Elements that could be perceived as dangerous were also extended, including elements such
as darkness, loud sounds/voices and unknown objects or environments. Playing with
dangerous elements fits with Lyng’s terminology edgework (1990), which includes
interpreting behavior as testing boundaries, literally or emotionally/mentally, and approaching
the edge of ones abilities. This can be applied to all types of risky play, but with regards to
one-year-olds, dangerous elements, as defined here, are probably more within their zone of
proximal development, rather than, for example, high speed or dangerous tools. This
decreases also the aspect of risk of physical injury, as in Example 2 and 3. When separating
one-year olds in Sample 1, two categories stand out: Speed (17%) and Elements (63%). In
Sample 2, the proportion of Playing with elements is even higher (69%).
Finally, some instances eluded existing categories. One type of play had the common
individual characteristics of crashing, either themselves or an object, into something.
“Impact” is defined as “the action of one object coming forcibly into contact with another”
(Oxford Dictionaries, 2015), which seems a good description of the children’s play; either if
they repeatedly threw themselves onto a mattress or crashed their tricycle into a fence. A new
category was therefore named playing with impact. The staff’s reactions were used as an
environmental characteristic as they sometimes reacted with frightened surprise.
Some observations had elements of fear, tension or excitement that were categorized as Other.
The analysis showed that some of these observations had common characteristics. These were
situations where the risk was only observed by the children. When a ski-jump was to take
place, a group of two-year-olds would sit down next to the jump and watch. Similarly, one-
year-olds were observed watching through the window older children slide or play rough and

15
tumble outside. This is suggested here as an emergent category named “Vicarious risk”.
According to Apter (1992), this experience can have the same arousing effect as a “real”
experience, and in this context is additionally interpreted as a pre-phase of risk-taking, with a
potential learning aspect.
Conclusion
The present paper suggests that the existing definition and characteristics of risky play are
appropriate for two- and three-year-old children. Regarding one-year-olds, the study suggests
several deviations from the existing understanding of risky play. In this context, the term one-
year-old must be seen in relation to motor development and in particular, the ability to walk.
One-year-olds show less risky play than older peers, and when playing, they express less
emotion, especially while alone. They do not show the same overt, easy-to-identify body
language and facial expressions as their older peers. Typically, one-year-olds’ risky play is
more frequently brief and solitary compared to two- and three-year-olds. One-year-olds’ main
risky activity is playing with dangerous elements, where the term dangerous must be
emphasized as subjective. Their risky play involves exploring and testing their surroundings
and their bodies in relation to these. To expand the understanding of risky play, this paper
suggests adding “Playing with impact” and “Vicarious risk” as new categories and delineating
characteristics to - play that involves uncertainty and exploration – bodily, emotional,
perceptional or environmental – that could lead to either positive or negative consequences.
Being an exploratory study with a low number of participants, the study has limitations.
However, the number of instances observed and similarities across the different contexts
suggest the potential for generalizability of the findings. The validity of the suggested
definition and characteristics will be further tested by future studies’ ability to utilize or
reproduce these findings. Taking the rapid, global expansion of childcare into consideration,
equally important would be how childcare centers deal with this type of play, e.g., creating
zones of proximal development for all children. The described behavior among one-year-olds,
presumes high levels of attention and sensitivity among ECEC-practitioners and research
should elucidate how practitioners observe, structure and/or engage in this type of play.

16
References
Aarts, M.-J., Wendel-Vos, W., van Oers, H. A. M., van de Goor, I. A. M., & Schuit, A. J. (2010).
Environmental Determinants of Outdoor Play in Children: A Large-Scale Cross-Sectional
Study. American Journal of Preventive Medicine, 39(3), 212-219.
doi:http://dx.doi.org/10.1016/j.amepre.2010.05.008
Adams, J. (2001). Risk. London: Routledge.
Aldis, O. (1975). Play Fighting. New York: Academic Press, Inc.
Apter, M. J. (1992). The Dangerous Edge: The Psychology of Excitement. New York: Free Press.
BePro. (2013). Better Provision for Norway's children in ECEC. Retrieved from
http://www.hioa.no/Forskning-og-utvikling/Hva-forsker-HiOA-paa/FoU-ved-LUI/Better-
Provision-for-Norway-s-children-in-ECEC
Bjørnestad, E., Pramling Samuelsson, I., Bae, B., Gulbrandsen, L., Johansson, J.-E., Løberg, H., & Os, E.
(2012). Hva betyr livet i barnehagen for barn under 3 år? En forskningsoversikt (978-82-
93208-13-6). Retrieved from
https://www.regjeringen.no/globalassets/upload/kd/vedlegg/barnehager/rapporter20og20p
laner/forskningsoversikt_barn_under_tre_aar.pdf
Blurton-Jones, N. (1976). Rough-and-tumble Play among Nursery School Children. In J. S. Bruner, A.
Jolly, & K. Sylva (Eds.), Play: Its Role in Development and Evolution (pp. 352-363). New York:
Basic Books.
Boyer, T. W. (2006). The development of risk-taking: A multi-perspective review. Developmental
Review, 26(3), 291-345. doi:http://dx.doi.org/10.1016/j.dr.2006.05.002
Bruner, J. S. (1976). Nature and Uses of Immaturity. In J. S. Bruner, A. Jolly, & K. Sylva (Eds.), Play: Its
Role in Development and Evolution (pp. 28-63). New York: Basic Books.
Byrnes, J. P. (2013). The Nature and Development of Decision-making : A Self-regulation Model.
Hoboken: Taylor and Francis.
Byrnes, J. P., Miller, D. C., & Schafer, W. D. (1999). Gender Differences in Risk Taking: A Meta-
Analysis. Psychological Bulletin, 125(3), 367-383. doi:http://dx.doi.org/10.1037/0033-
2909.125.3.367
Christensen, P., & Mikkelsen, M. R. (2008). Jumping off and being careful: children's strategies of risk
management in everyday life. Sociology of Health & Illness, 30(1), 112-130.
doi:http://dx.doi.org/10.1111/j.1467-9566.2007.01046.x
Christensen, S., & Morrongiello, B. A. (1997). The influence of peers on children's judgments about
engaging in behaviors that threaten their safety. Journal of Applied Developmental
Psychology, 18(4), 547-562.
Cook, S., Peterson, L., & DiLillo, D. (1999). Fear and Exhilaration in Response to Risk: An Extension of a
Model of Injury Risk in a Real-World Context. Behavior Therapy, 30(1), 5-15.
doi:http://dx.doi.org/10.1016/S0005-7894(99)80042-2
Corsaro, W. A. (2003). We're Friends, Right? Inside Kids' Culture. Washington DC: Joseph Henry Press.
Cosco, N. G., Moore, R. C., & Islam, M. Z. (2010). Behavior mapping: a method for linking preschool
physical activity and outdoor design. Med Sci Sports Exerc, 42(3), 513-519.
doi:http://dx.doi.org/10.1249/MSS.0b013e3181cea27a
Creswell, J. W., & Plano Clark, V. L. (2011). Choosing a Mixed Methods Design. In J. W. Creswell & V.
L. Plano Clark (Eds.), Designing and Conducting Mixed Methods Research (2nd ed., pp. 53-
106). Thousand Oaks: SAGE Publications.
Ellis, B. J., Del Giudice, M., Dishion, T. J., Figueredo, A. J., Gray, P., Griskevicius, V., . . . Volk, A. A.
(2012). The Evolutionary Basis of Risky Adolescent Behavior: Implications for Science, Policy,
and Practice. Developmental Psychology, 48(3), 598-623.
doi:http://dx.doi.org/10.1037/a0026220
Engdahl, I. (2007). Med barnens röst : Ettåringar ”berättar” om sin förskola (Licentiatavhandling i
barn- och ungdomsvetenskap), Stockholm Institute of Education, (40)

17
Engel, A., Barnett, W. S., Anders, Y., & Taguma, M. (2015). Early Childhood Education and Care
Review: Norway. Retrieved from http://www.oecd.org/norway/Early-Childhood-Education-
and-Care-Policy-Review-Norway.pdf
Flewitt, R. (2005). Conducting research with young children: Some ethical considerations. Early Child
Development and Care, 175(6), 553-565.
Geertz, C. (1994). Thick Description: Toward an Interpretive Theory of Culture. In M. Martin & L. C.
McIntyre (Eds.), Readings in the Philosophy of Social Science (pp. 213-231). London, England:
The MIT Press.
Gobo, G. (2008). Re-Conceptualizing Generalization: Old Issues in a New Frame. In P. Alasuutari, L.
Bickman, & J. Brannen (Eds.), The SAGE Handbook of Social Research Methods (pp. 193-213).
London: SAGE Publications Ltd.
Goodway, J. D., Ozun, J. C., & Gallahue, D. L. (2012). Motor Development in Young Children. In O. N.
Saracho & B. Spodek (Eds.), Handbook of Research on the Education of Young Children (3 ed.,
pp. 89-100). New York: Routlegde.
Gulløv, E., & Højlund, S. (2003). Feltarbejde blandt børn: metodologi og etik i etnografisk
børneforskning. København: Gyldendal Uddannelse.
Harms, T., Cryer, D., & Clifford, R. M. (2006). Infant/Toddler Environment Rating Scale - Revised
edition. In. New York: Teachers College Press.
Hughes, B. (2012). Evolutionary Playwork (3rd ed.). Oxford: Routledge.
Humphreys, A. P., & Smith, P. K. (1984). Rough-and-tumble in Preschool and Playground. In P. K.
Smith (Ed.), Play in Animals and Humans (pp. 241-266). Oxford, England: Basil Blackwell, Inc.
James, A., & Prout, A. (1997). Constructing and reconstructing childhood: contemporary issues in the
sociological study of childhood (2nd ed.). London: Falmer Press.
Johannessen, A., Tufte, P. A., & Christoffersen, L. (2010). Introduksjon til samfunnsvitenskapelig
metode (4 ed.). Oslo: Abstrakt.
Johnson, J. E., Sevimli-Celik, S., & Al-Mansour, M. (2012). Play in Early Childhood Education. In O. N.
Saracho & B. Spodek (Eds.), Handbook of Research on the Education of Young Children (Vol.
3, pp. 265-275). New York: Routledge.
Kaarby, K. M. E. (2005). Children playing in nature. Paper presented at the Questions of Quality,
September 23rd-25th 2004, Dublin Castle, Ireland.
Knoblauch, H. (2005). Focused ethnography. Forum Qualitative Sozialforschung/Forum: Qualitative
Social Research, 6(3).
Knoblauch, H., & Schnettler, B. (2012). Videography: analysing video data as a ‘focused’ethnographic
and hermeneutical exercise. Qualitative Research, 12(3), 334-356.
Lange, A., & Mierendorff, J. (2009). Method and Methodology in Childhood Research. In J. Qvortrup,
W. A. Corsaro, & M.-S. Honig (Eds.), The Palgrave Handbook of Childhood Studies (pp. 78-95).
London: Palgrave Macmillan UK.
Lillemyr, O. F. (2009). Taking Play Seriously: Children and Play in Early Childhood Education – An
Exciting Challenge. Charlotte, NC: Information Age Publishing, Inc.
Lillemyr, O. F., Dockett, S., & Perry, B. (2013). Play and Learning in Early Years Education:
International Perspectives In O. F. Lillemyr, S. Dockett, & B. Perry (Eds.), Varied Perspectives
on Play and Learning: Theory and Research on Early Years Education (pp. 1-8). Charlotte, NC:
Information Age Publishing, Inc.
Lyng, S. (1990). Edgework: A Social Psychological Analysis of Voluntary Risk Taking. American Journal
of Sociology, 95(4), 851-886.
Løkken, G. (2000). The playful quality of the toddling "style". International Journal of Qualitative
Studies in Education, 13(5), 531-542. doi:http://dx.doi.org/10.1080/09518390050156440
Malaby, T. M. (2002). Odds and Ends: Risk, Mortality, and the Politics of Contingency. Culture,
Medicine and Psychiatry, 26(3), 283-312. doi:http://dx.doi.org/10.1023/A:1021204803969
Martin, P., & Caro, T. M. (1985). On the Functions of Play and its Role in Behavioral Development.
Advances in the Study of Behavior, 15, 59-103.
Merleau-Ponty, M. (2012). Phenomenology of Perception. London: Routledge.

18
Millen, D. R. (2000). Rapid Ethnography: Time Deepening Strategies for HCI Field Research. Paper
presented at the Proceedings of the 3rd conference on Designing interactive systems:
processes, practices, methods, and techniques, New York, NY, USA.
Mårtensson, F. (2004). Landskapet i leken. (Doctoral thesis), Swedish University of Agricultural
Sciences, Alnarp, Sweden.
Oxford Dictionaries, O. D. (2015). Impact. Retrieved from
http://www.oxforddictionaries.com/definition/english/impact
Pellegrini, A. D. (1988). Elementary-School Children's Rough-and-Tumble Play and Social
Competence. Developmental Psychology, 24(6), 802-806.
doi:http://dx.doi.org/10.1037/0012-1649.24.6.802
Pellegrini, A. D., Dupuis, D., & Smith, P. K. (2007). Play in evolution and development. Developmental
Review, 27(2), 261-276. doi:http://dx.doi.org/10.1016/j.dr.2006.09.001
Pellegrini, A. D., & Smith, P. K. (1998). Physical Activity Play: The Nature and Function of a Neglected
Aspect of Play. Child Development, 69(3), 577-598. doi:http://dx.doi.org/10.1111/j.1467-
8624.1998.tb06226.x
Piaget, J. (1954). The Construction of Reality in the Child. New York: Basic books.
Pink, S., & Morgan, J. (2013). Short‐term ethnography: Intense routes to knowing. Symbolic
Interaction, 36(3), 351-361.
Readdick, C. A., & Park, J. J. (1998). Achieving Great Heights: The Climbing Child. Young Children,
53(6), 14-19.
Sando, O. J., & Lysklett, O. B. (2012). 1-3 åringers uteaktivitet om vinteren. In FoU i praksis 2011.
Trondheim: Tapir Akademisk Forlag.
Sandseter, E. B. H. (2007). Categorising risky play—how can we identify risk‐taking in children's play?
European Early Childhood Education Research Journal, 15(2), 237-252.
doi:http://dx.doi.org/10.1080/13502930701321733
Sandseter, E. B. H. (2009a). Characteristics of risky play. Journal of Adventure Education and Outdoor
Learning, 9(1), 3-21. doi:http://dx.doi.org/10.1080/14729670802702762
Sandseter, E. B. H. (2009b). Children’s Expressions of Exhilaration and Fear in Risky Play.
Contemporary Issues in Early Childhood, 10(2), 92-106.
Sandseter, E. B. H. (2010a). ‘it tickles in my tummy!’: understanding children’s risk-taking in play
through reversal theory. Journal of Early Childhood Research, 8(1), 67-88.
doi:http://dx.doi.org/10.1177/1476718x09345393
Sandseter, E. B. H. (2010b). Scaryfunny: A Qualitative Study of Risky Play Among Preschool Children.
(Philosophiae Doctor), Norwegian University of Science and Technology, Trondheim.
Sandseter, E. B. H., & Kennair, L. E. O. (2011). Children’s Risky Play from an Evolutionary Perspective:
The Anti-phobic Effects of Thrilling Experiences. Evolutionary Psychology, 9(2), 257-284.
Seawright, J., & Gerring, J. (2008). Case Selection Techniques in Case Study Research: A Menu of
Qualitative and Quantitative Options. Political Research Quarterly, 61(2), 294-308.
Smith, P. K. (2005). Play: Types and Functions in Human Development. In B. J. Ellis & D. F. Bjorklund
(Eds.), Origins of the Social Mind: Evolutionary Psychology and Child Development (pp. 271-
291). New York: The Guilford Press.
Smith, S. J. (1998). Risk and Our Pedagogical Relation to Children: On the Playground and Beyond.
Albany: State University of New York Press.
Staempfli, M. B. (2009). Reintroducing Adventure Into Children's Outdoor Play Environments.
Environment and Behavior, 41(2), 268-280.
doi:http://dx.doi.org/10.1177/0013916508315000
Stephenson, A. (2003). Physical Risk-taking: Dangerous or endangered? Early Years: An International
Journal of Research and Development, 23(1), 35-43.
Storli, R., & Hagen, T. L. (2010). Affordances in outdoor environments and children's physically active
play in pre‐school. European Early Childhood Education Research Journal, 18(4), 445-456.
doi:http://dx.doi.org/10.1080/1350293X.2010.525923
Sutton-Smith, B. (1997). The Ambiguity of Play. London: Harvard University Press.

19
Teddlie, C., & Yu, F. (2007). Mixed Methods Sampling: A Typology With Examples. Journal of Mixed
Methods Research, 1(1), 77-100. doi:http://dx.doi.org/10.1177/2345678906292430
Tovey, H. (2007). Playing Outdoors: Spaces and Places, Risk and Challenge. Maidenhead: Open
University Press.
Vygotsky, L. S. (1978). Mind in Society: The Development of Higher Psychological Processes.
Cambridge, Mass.: Harvard University Press.