Claire Saunders, Centre for Education Studies, University of Warwick

Abstract

'Risky play' is a valued play type in Early Childhood Education and Care, due to the benefits it provides to children's holistic development. However, risky play is often considered an outdoor play type; as current safety concerns and lifestyle changes are resulting in children spending more time indoors, provision of risky play is therefore under threat. This article aims to explore the extent to which this provision can be facilitated indoors. Sandseter's (2007a: 243) categorisation of risky play as play with 'great heights, high speed, harmful tools, near dangerous elements, rough-and-tumble play and play where children can disappear or get lost' is used to identify the physical risky play behaviours of focus in the study. A case study, in a Norwegian kindergarten, was conducted using observations of children's play and interviews with practitioners to analyse the facilitation of risky play indoors and the associated attitudes of practitioners. The findings identified that some risky play types are possible indoors, although prominent barriers were evident. Sandseter's categorisation is also recognised as restrictive when applied to the indoor environment, resulting in the call for further research to expand the study's findings, and to reconsider how risky play indoors is categorised.

Keywords

Risky play, indoor play, indoor learning environments, Norwegian play environments, early childhood, childhood risk-taking.

Introduction and context

Children's play is a major component in early years provision as it benefits children's holistic development (Sumaroka and Bornstein, 2009: 294-97; Play England, 2015). This article focuses on 'risky play', a topical, physical play type that benefits children's physical and socio-emotional development from exposure to risky experiences (HSE, 2012: 1). Previously, this play has been recognised as predominantly conducted outdoors (Sandseter, 2009c: 6); therefore, a societal shift towards children spending more time indoors (Keung, 2016: 230) creates uncertainty regarding children's access to risky play. Limited literature has considered this phenomenon indoors, and therefore research is required to prevent this play type from being stifled.

Multiple definitions of risky play are offered within previous, relevant academic studies, taking both adult and child perspectives. Exploring boundaries, feeling out of control, challenge, potential for injury and feelings of excitement and scariness are commonly cited in these definitions, suggesting the common characteristics of risky play (Little, Wyver and Gibson, 2011: 115; Sandseter, 2007b: 248; Stephenson, 2003: 36; Ball, 2002: 51; for further definitions of risky play, also see Sandseter, 2009c: 6-7; Brussoni et al., 2012: 3136; and Kaarby, 2004: 123-25). Sandseter (2014: 434) later highlighted that the benefits of these experiences must be balanced with the potential for injury to outweigh potential negative outcomes of this play.

Sandseter (2007a: 243) expanded these definitions by providing a categorisation of risky play activities, elaborating on how risky play can be evident within play activities. The categorisation is as follows: play with 'great heights, high speed, harmful tools, near dangerous elements, rough-and-tumble play and play where children can disappear or get lost'. Sandseter's categorisation will be the focus of this article based upon its frequent citation in associated literature (see for example Little, Wyver and Gibson, 2011: 115; Brussoni et al., 2012: 3136; Hill and Bundy, 2014).

Children appear to spending an increased amount of time indoors, which Duderstadt (2014: 56) estimated as 80-90 per cent of their time. Keung (2016: 230) recognises numerous causes including safety fears, increased traffic and the renovation of outdoor spaces making them less inviting for children's play.

Emphasis is often placed on physical, active play outside (Gleave and Cole-Hamilton, 2012: 1, 7-8) (where physically risky play can be considered a form of active play, as discussed by Brussoni et al. (2012: 3136)); therefore, threats to current provision of outdoor risky play are evident. Children's increased engagement in activities indoors is not solely due to safety fears, as Keung (2016: 230) also identifies an increase in media and the widespread availability of computer games, televisions and the internet as influences. Although these factors provide a threat to children's engagement in risky play irrespective of the environment, Sandseter (2007b: 248; 2007c) identifies that children have an innate desire to experiment with risk, and an ability to seek risky play in any environment (Sandseter, 2009a: 446) providing a rationale for the provision of risky play indoors to overcome some of these factors.

This article explores the facilitation of risky play indoors. As this is a relatively new area of research, focus is placed upon Sandseter's (2007a: 243; 2007b: 249-53, 2009b: 98-103) categorisation of risky play types to determine the possibilities for their facilitation indoors as it is popular within existing literature (e.g. Little, Wyver and Gibson, 2011: 115; Brussoni et al., 2012: 3136; Hill and Bundy, 2014). Using the research question 'To what extent can pre-identified forms of risky play be facilitated indoors?', the study aimed to observe risky play indoors, with comparisons to provision outdoors, to determine the prevalence of these play types. Childcare practitioners' views were also considered to identify potential barriers and facilitators to these play types indoors.

Literature review

The use of indoor environments for risky play has received limited recognition (Weinberger, Butler and Schumacher, 2014: 196), making available literature sparse. The following review contains literature surrounding Sandseter's (2007a: 243; 2007b: 243-53; 2009b: 98-103) categories of risky play and the necessary affordances for risky play.

Definition and social construction

Sandseter (2007a: 238) originally categorised risky play to overcome weaknesses within literature and provide a comprehensive account of risky play and associated behaviours. This categorisation was furthered with specific examples in subsequent studies. These examples include (Sandseter 2007b: 249-53; 2009c: 7) 'play at great heights', including climbing, balancing at great height and jumping from still and flexible surfaces; 'play at high speed', including bicycling, running uncontrollably and swinging; 'play with dangerous tools', including cutting tools and strangling tools such as ropes; 'play with dangerous elements', including play near cliffs and deep water; 'rough-and-tumble play', including wrestling and play fighting; and 'play where children disappear or get lost', including play were children go exploring or think they are out of sight of adults.

Lupton (2013: 45-46) recognises that the concept of risk is socially constructed, with multiple factors influencing a person's perception of risk, including knowledge and cultural factors. Differences in the facilitation of risky play within early years provision are therefore recognisable. In particular, Little, Sandseter and Wyver (2012: 305-13) recognised a more facilitative perspective among practitioners in Norway compared to Australia, resulting from a more relaxed, less litigious environment. Facilitation of risky play generally may differ between early years contexts, especially internationally.

Environments

Despite focusing on outdoor environments, Sandseter (2009a: 446) recognises that children can seek risky play in any environment, suggesting possible affordances indoors. As current literature focuses predominantly on affordances outdoors, recognition of the barriers to risky play indoors are prevalent with little literature attention given to facilitators. This limited focus may have caused a current unawareness of the environment's potential (Weinberger, Butler and Schumacher, 2014: 196), providing a narrow reflection of the affordances available inside.

Gibson's theory of affordances (1986: 127-38) explains some elements of the outdoor environment that foster risky play. As elaborated by Heft (1988: 33) the outdoor environment contains 'climb-on-able', 'jump-on/down-off-able' and 'swing-on-able' features as a result of natural features such as grass and hills, providing examples of facilitators of risky play (Sandseter, 2009b). Hawkins (2013: 77) recognises that other natural elements such as weather and fire also facilitate risky play, identifying the strength of affordances in outdoor environments (Hughes, 1996, cited in Hawkins, 2013: 77). The lack of these natural elements indoors may create an initial barrier to the facilitation of risky play.

Furthermore, practitioners have been identified as holding attitudes that favour sedentary play activities indoors (Murphy, 2015: 191-92), which Johnson, Christie and Wardle (2005, cited in Hirose, Koda and Minami, 2012: 1612) previously identified as a result of perceptions surrounding the indoor environment as a calmer, formal learning space. For example, restrictive attitudes towards rough-and-tumble play indoors have been recognised as a result of the noise and 'chaos' it brings (Storli and Sandseter, 2015: 2005). These attitudes provide a possible barrier to risky play indoors as Kyttä (2004) recognises that adults often control children's play.

Walsh (2016: 54) identifies that children are able to concentrate and explore new skills better in quieter areas. Although Walsh (2016: 54) discusses the facilitation of quieter areas outdoors, it would be logical to suggest that where indoor areas are quieter and calmer as a result of the increased conduction of sedentary activities, children's ability to concentrate on their risk-taking would increase, for example the use of dangerous tools. Although more research is required to confirm this hypothesis, the sedentary atmosphere of indoor environments may afford risky play by maintaining children's concentration, providing good-quality risky play experiences.

Tovey (2007: 18) also identifies that children's ability to be active is restricted indoors because of general limitations to open space. Tovey (2014: 217) later specifically recognised that rough-and-tumble play is a form of physically active play; therefore, indoor environments can be perceived as particularly restrictive in facilitating this category. However, Engel (2003, cited in NICE Public Health Collaborating Centre - Physical Activity, 2008: 22) discusses the possibility of facilitating active play indoors if furniture is organised to provide a large open space. Within the NICE Public Health document, suggestions are given to divide indoor spaces into separate zones to facilitate various and specific play types, including physically active play. This suggests that the organisation of indoor spaces may afford risky play, especially rough-and-tumble play, if arranged accordingly.

Furthermore, Tovey (2010: 80-81) recognises that risk-taking is a general characteristic of play and can be present in many types of play, not just physical play. Instead, Tovey suggests that both real and imagined dangers provide a sense of risk-taking for children, as well as other play types involving pushing boundaries. Wood (2013: 131) identified children's participation in 'pretend play' as an example as they explore emotions and uncertainty through imagined risks. Hirose, Hinobayashi and Minami (2007, cited in Hirose, Koda and Minami, 2012: 1612) identify that dramatic play involving 'pretending' is more commonly conducted indoors. Indoor environments may therefore facilitate alternative forms of risky play and risk-taking experiences for children, although limited literature exists surrounding this concept. Although only physically risky play is considered within this article, consideration should be given to this perspective to consider possible wider affordances of indoor environments.

Methodology

A case-study approach was used to provide an in-depth exploration of risky play indoors, based upon the lack of previous research. Although case studies do not provide generalisability (Mukherji and Albon, 2010: 82), the depth of the findings were the priority.

The case study was formed of a class within a Norwegian 'barnehage' (kindergarten), observing the play opportunities available and investigating practitioner attitudes surrounding risky play indoors. Bilton (2010: 28, 79) recognises that Norwegian practices surrounding risky play are held as an exemplar as they culturally support children's risk-taking. The opportunity to conduct the research in the barnehage was provided by the researcher working as an assistant there.

Triangulation helped to reduce potential bias resulting from the researcher's position. Appendix 3 details the themes in which triangulation was prevalent.
The case study incorporated nineteen children and four practitioners. Despite being a 'convenience sample' as a result of the researcher's role (Denscombe, 2014: 43), initial observations prior to data collection confirmed the suitability of this 'case', as liberalism was observed surrounding risky play.

Research methods

Participant observations and interviews with practitioners were conducted in this study. Observations were used to identify and analyse children's play behaviours to determine which, and to what extent, the categories of risky play were facilitated indoors. Practitioner interviews were conducted to explore their thoughts and facilitation of risky play indoors, to gain a greater understanding of possible affordances and barriers in light of Kyttä's (2004) recognition of adults' control of children's play.

Overall, twenty five-minute observations, ten indoors and ten outdoors, were conducted to enable comparisons. Children's participation in Sandseter's (2007a: 243; 2007b: 249-53; 2009b: 98-103) categories of risky play and supporting field notes were documented on to written observation schedules. Appendix 1 details the exact definitions used. Observations were conducted on nine separate days between September and November 2014 chosen by the setting, averaging two observations each day. Indoor observations occurred in the morning, and outdoor observations in the afternoon, reflecting the setting's daily timetable. Observations consisted of observing the play behaviours in one, pre-chosen, space indoors and outdoors that was frequently utilised by all the children and that held affordances for each risky play category as cited by Heft (1988: 33) and Sandseter (2009b: 98-103). The children involved resulted from random sampling (Denscombe, 2014: 36) as any of the nineteen children had the possibility of being observed. On average seven children were observed in a single observation with group sizes containing between five and ten children at any one time. The observations were mixed gender with children aged between 2 and 6 years old.

One-to-one, semi-structured interviews were used to investigate practitioners' attitudes towards the facilitation of risky play in indoor settings. Four female practitioners participated, based upon their competencies in English to ensure their viewpoints were accurately represented. The sample varied in age and education level, but consisted of two barnehage assistants, a room leader and a Special Educational Needs specialist practitioner. Topics discussed included differences between indoor and outdoor environments; planning the provision of risky play indoors and outdoors; and their attitudes towards and facilitation of risky play in both venues. An interview schedule is provided in Appendix 2.

Triangulation between these research methods enhanced the validity of the findings, particularly to reduce misrepresentation of practitioners' views as the interviews were conducted in English instead of their native language.

Ethical considerations

The 'risk' element of risky play and the involvement of children created multiple ethical considerations for this study. A situational ethics approach was taken (Oliver, 2010) accommodating the nature of risky play. Consequently, each observation was monitored and intervention occurred where necessary to promote safety.

Informed voluntary consent was gained from all participants, the barnehage and the children's guardians, with written explanation provided to guardians in English and Norwegian to ensure understanding. The children had the project explained to them by their lead practitioner and they provided verbal consent. Robson and McCartan (2016: 224) identifies that it is difficult to ensure children have given informed consent; therefore their consent was also treated as an ongoing process with consideration to other expressions of consent, such as willingness to engage, as suggested by Langston et al. (2014: 152-53).
Documentation detailing the research focus, procedure, sample and arrangements surrounding consent, confidentiality, protection of participants and account for participants' cultural values and privacy was submitted to and approved by a supervisor at the Centre for Education Studies, University of Warwick. This approval was documented with the course secretary.

Data analysis

Data from the observations was analysed to identify the numerical occurrence of the types of risky play in an indoor and outdoor environment for comparison. The data from the interviews was coded to identify relevant affordances and barriers towards the facilitation of the risky play categories inside. Further considerations resulting from considering these two data sets together form the discussion.

Results and analysis

The specific findings from each research method are detailed first, followed by an expanded analysis.

Observational data

Overall, 23 of the 57 (40.35%) documented incidences of risky play occurred indoors, suggesting a reduced extent to which risky play was facilitated indoors. Some observed play types fell into two categories, such as swinging on a swing both at height and speed which were documented as fulfilling both categories. 55 single play behaviours were observed. Three indoor and one outdoor observations contained no risky play behaviours.
Prominent differences were recognised within the types of risky play (Sandseter, 2007a: 243) observed indoors and outdoors, presented in Figure 1:

Figure 1: Incidences of each category of risky play. Incidences of risky play indoors are depicted by the left-hand bar for each category with those occurring outdoors on the right.

As represented in Figure 1, all observed play with 'dangerous tools' was conducted indoors suggesting that affordances for this play were available indoors that may have been absent outdoors. 'Rough-and-tumble' play was equally observed in each venue, suggesting a common facilitation in both venues. To provide context, five incidences of play with dangerous tools consisted of play involving scissors, the other involved using clay tools (knives) to cut clay. However, the figure indicates vast reductions in play at 'great heights' and 'high speed' indoors in comparison, suggesting prevalent barriers to the facilitation of these two play types.

Field notes

Disparities were also recorded regarding how play at great height and high speed were conducted indoors and outdoors. The intensity of both types was observed as reduced indoors. For example, observations of play at great height indoors consisted of children climbing on to furniture that was table-height, compared to observations of children climbing trees, climbing-frames and walls, and using swings outdoors which were a greater height from the ground. Similarly, play at high speed indoors consisted of children running or spinning on the spot quickly compared with the use of swings as well as running outdoors. Environmental differences are therefore apparent within the specific activities children participated in, thus creating different considerations to what is perceived as 'risky' in each environment.

Furthermore, play also occurred with equipment at great heights, as opposed to children being at a great height themselves which was focused on by Sandseter (2007b: 249-53; 2009c: 7). In play such as tower-building with blocks, a potential for injury and experiences of challenge and uncertainty were present despite this deviation.

Stricter rules were also prevalent within the indoor environment. For example, it was recorded that children were prohibited from sliding on chairs with wheels, reducing the possibility of a potential activity at high speed. No observations of similar restrictions were made outdoors. Increased open space was also noted outdoors, as the indoor environment contained more furniture, such as tables and equipment. One injury was observed indoors from running at high speed as a child ran into a piece of furniture.

Finally, it was documented that the children went on a one-off trip to an indoor soft play area, with various soft shapes, ladders and climbing equipment. Direct observations were not conducted in this venue as it was not the selected observation venue; however, it should be considered that different indoor venues hold different affordances for risky play.

Practitioner interviews

A common theme throughout the practitioner interviews was the belief that outdoors provided a 'better', more effective environment for encouraging risky play. The results were then coded to identify prominent facilitators and barriers towards the facilitation of risky play indoors.

Environmental factors

Two practitioners explicitly felt that opportunities for risky play were less intense indoors compared to outdoors. Three practitioners elaborated, identifying that elements of the natural outdoor environment, such as 'weather' and 'fire pits' provided an element of risk. This was also supported by the increased observations of play at great heights and high speed occurring outdoors, suggesting that both the physical environment and available equipment indoors may create barriers, hindering the quality of certain categories of risky play inside.

Space indoors

Three practitioners discussed the requirement of open space for the facilitation of risky play, and that this was restricted indoors. One also discussed that the organisation of the furniture indoors increased the potential for injury from these physical play types, suggesting that the indoor environment may actually be more hazardous in the facilitation of risky play. As Sandseter (2014: 434) recognises the need to balance the benefits and potential for injury when facilitating risky play, a strong barrier is presented. Recognition of this barrier is furthered as the lack of open space in the indoor environment, and occurrence of an injury indoors was documented as observational notes.

Ease of supervision indoors

One practitioner expressed the opinion that supervising play was easier indoors as a result of reduced 'space'. She explicitly cited that the indoor environment allows practitioners to sit alongside children at a table to work with them. She also referenced children's use of cutting tools and that required concentrated supervision was easier indoors as there were fewer distractions. This suggests a potential facilitator within the indoor environment, particularly for play with dangerous tools.

Practitioner attitudes towards the use of the indoor environment

Practitioners held different attitudes towards the use of and types of play conducted indoors and outdoors. Specific attitudes surrounded the encouragement of sedentary activities and quieter play indoors.

Attitudes favouring the facilitation of sedentary activities indoors were recognisable throughout the data. As well as citations to space limitations for physically risky play indoors, there was a general consensus of an overarching paradigm of the promotion of sedentary and more formal learning experiences indoors in all responses. Johnson, Christie and Wardle (2005, cited in Hirose, Koda and Minami, 2012: 1612) also recognised this perception, identifying a potentially widespread ideological barrier towards the facilitation of physically risky play indoors.

Furthering this ideology, all practitioners identified an aversion to loud noise within the indoor environment. They suggested 'the children can get much louder' when involved in risky play and therefore discussed its increased suitability outdoors. They did, however, recognise that not all risky play activities are loud, identifying that this barrier is not applicable to all risky play behaviours. Unfortunately they did not elaborate further.

One practitioner also explicitly discussed barnehage rules surrounding noise, with stricter rule enforcement also observed surrounding children's play indoors furthering this barrier. Full analysis of the barnehage's rules was not conducted; therefore the full extent of this influence is unknown. However, it does suggest a further manner in which ideologies can influence the provision of the categorised risky play behaviours indoors, individual to the setting in question.

Further discussion

Summary

The findings are summarised below by play category to consider the extent to which each category can be facilitated indoors.

Play at great height

The vast reduction and intensity in observations of this play type compared to outdoors identifies that prominent barriers restrict its facilitation. Barriers were recognised regarding the physical environment indoors, available resources (and resultant space) and ideological preferences for quiet, sedentary play indoors.

Play with high speed

Similarly, reduced incidences and intensities of the play type were observed indoors in comparison. This behaviour was displayed as running and spinning indoors. The same barriers were presented here as those affecting play at great height. Recognition can also be given to the layout of the indoor environment, making play with high speed more risky but more hazardous, which must be balanced with the benefits of experiencing this play type. As a child was observed sustaining an injury while running at high speed indoors, room layout may be particularly restrictive towards the facilitation of this category indoors.

Play with dangerous tools

All observations of this play type were within the indoor environment. Practitioner recognition of the quality of supervision indoors suggests that the indoor environment holds positive affordances for this play type. Within the observations, the use of scissors and small knives were witnessed, which can be considered as smaller tools than those such as axes and saws identified within Sandseter's categorisation (2007a: 245-46; 2007b: 253; 2009c: 7). Indoor environments can therefore be recognised as affording 'play with smaller dangerous tools'.

Play with dangerous elements

This play type was not observed in either play venue; therefore, informed conclusions cannot be drawn. As Hughes (1996, cited in Hawkins, 2013: 77) identified that natural elements such as weather and fire are outdoor phenomena, this category is potentially difficult to facilitate indoors.

Rough-and-tumble play

This play was observed equally indoors and outdoors, identifying that rough-and-tumble play can be facilitated successfully indoors. As it has been discussed that open space is a facilitator of this play type (Tovey, 2014: 217), and as it was recognised within the data that lack of open space was a limitation of the indoor environment, it can be identified that the facilitation of rough-and-tumble play is possible indoors. The extent of facilitation may depend on the individual indoor setting.

Play where children disappear or get lost

No observations of this play type were made in this study, preventing analysis. However, logical consideration could suggest that the lack of open space within the indoor environment, as previously discussed, could facilitate this play type, as there are more elements to get lost in or hide behind. This is only an assumption; further research is required for confirmation.

Indoor venues

Within the results and discussion of this article, it is evident that prominent barriers towards the facilitation of Sandseter's risky play categories are present in indoor environments. However, it is important to consider that the prevalence of these barriers depends upon the individual environment. It can be recognised that an alternative, specialist, indoor play environment accessed by the children - not part of the barnehage - contained equipment such as ladders which could facilitate play at great heights. The findings are therefore individual to the setting observed; however, they provide an analysis of possible barriers and facilitators within early years classrooms.

Suitability of Sandseter's categorisation of risky play

This study observed a reduced facilitation of the identified risky play categories indoors and detected associated barriers. Only play with dangerous tools and rough-and-tumble play could be recognised as strongly afforded by this particular indoor environment.

Sandseter's (2007a: 243, 2007b: 248) categorisation is based upon multiple studies of risky play in an outdoor environment and this study identifies that facilitation of these categories indoors is restricted, particularly resulting from physical and ideological characteristics of indoor play environments. Play at great heights and at high speed were particularly restricted indoors.

As the exploration of boundaries, feelings of being out of control, scariness and challenge are recognised as characteristics of risky play within the academic literature (Little, Wyver and Gibson, 2011: 115; Sandseter, 2007b: 248; Stephenson, 2003: 36; Ball, 2002: 51), it is possible that other, more relevant play behaviours exist within indoor environments that provide these experiences and therefore provide risky play outside of Sandseter's categorisation. This categorisation provides a limited account of the extent to which risky play can be facilitated indoors.

In response to this limitation, the findings from this study suggest two omissions from Sandseter's categorisation that could be used to provide alternative categories of risky play indoors. The first is that, within the category of 'play at great heights', Sandseter (2007b: 249) cites activities, such as climbing trees and cliffs, where the children are at great heights themselves. During the observations, supporting notes recognised that play occurred where objects were at great heights with the children playing at ground level. For example, children were seen to build tall towers with blocks, with the risk that the tower would fall over. The Froebel Blockplay Research Group (1992a, 1992b) recognise that play in this manner involves risk-taking as children experiment with boundaries (to reach a point where the tower falls over) and holds the potential for injury if the tower falls. As this play holds these characteristics of risky play (see Sandseter, 2007b: 248, Stephenson, 2003: 36; Ball, 2002: 51), 'play with resources at great heights' may be a potential risky play category with more relevance to the indoor environment.

A second omission can be recognised regarding Sandseter's (2007a: 243, 2007b: 251-52) categorisation of play with dangerous tools. Sandseter considers play using whittling knives, hammers, nails, saws and ropes as play with dangerous tools - tools which are potentially large. In this study, the children were observed using clay tools (knives) and scissors as dangerous tools, equipment which is smaller than those tools cited by Sandseter. As the use of the categorised tools was not observed in the outdoor environment either, the reasons for their absence indoors is unknown. From the observation findings and literature recognition of space limitations indoors (Tovey, 2010: 81) it could be suggested that 'play with small dangerous tools' could be a more appropriate category of indoor risky play.

Further limitations of this categorisation are also noticeable as the study focused on physically risky play only, something the adult participants also challenged. Alternative, more sedentary, forms of risk-taking were not considered. This provides a limitation as it has been identified that play types, such as dramatic play, that hold affordances for risk-taking involving imagined risks (Wood, 2013: 131) have been identified as more common in indoor environments (Hirose, Hinobayashi and Minami, 2007, cited in Hirose, Koda and Minami, 2012: 1612). Therefore, this study does not reflect the true potential of the indoor environment as a risky play venue.

Individuality

Lupton (2013: 45) recognised that risk as a concept is socially constructed; therefore individuality regarding practitioners' perceptions towards and facilitation of risky play is inevitable. Consideration must also be given to the fact that this research was conducted in Norway. As Norway is recognised as holding risky play in high liberal and cultural value (Nilsen, 2008) the attitudes, facilitators, barriers to and behaviours surrounding risky play may not be the same as in other western countries, including the UK. Although not generalisable, this study does provide a new exploration into the facilitation of risky play indoors.

Limitations and implications

A major limitation of this study is that it was conducted on a small scale, as a result of time and resource restrictions placed upon the researcher. It therefore provides restricted findings surrounding the facilitation of risky play indoors and generalisability is not possible. However, as consideration to risky play indoors has been limited in previous literature, it does provide an initial exploration into the facilitation of risky play indoors through the application of Sandseter's (2007a: 243, 2009b: 98-103) categorisation.

As a result of the small scale of this study, the implications from this study strongly surround the need for further research. Firstly, further research would improve the generalisability of and expand upon the findings presented in this article, particularly if focused solely on risky play indoors. Additionally, as this article suggests that Sandseter's categorisation of risky play may not be entirely relevant to indoor play environments, further research could observe and analyse risky play behaviour indoors more generally, potentially leading to alternative categorisations of risky play behaviour indoors.

Conclusion

This article identifies that some of Sandseter's (2007a: 243) categories of risky play, particularly 'play with dangerous tools' and 'rough-and-tumble play' can be facilitated indoors. In comparisons with provision outside, 'play at great heights' and 'high speed' were restricted inside regarding their intensity.

Barriers were prominently recognisable regarding the extent of facilitation of Sandseter's risky play categories indoors, including the available space, the lack of natural elements (such as fire and water) and practitioner attitudes towards the use of indoor learning environments. Despite this, ease of supervision and notions of indoor environments being quieter were recognised as affordances.

Further recognition was given to the restrictive nature of Sandseter's categorisation when considering the extent to which risky play can be facilitated indoors. 'Play with small tools' and 'play with resources at great height' were offered as two additional categories that provide more relevance to indoor environments. Further research is required to provide a more thorough analysis of opportunities and behaviours indoors to provide a more extensive analysis of the facilitation of risky play indoors. Consideration should also be given to wider forms of risk-taking, not just children's physical risk-taking.

---

Acknowledgements

My heartfelt thanks go to Dr Jaqueline Dynes and Anne Sine Van Marion, who supported me during my time in Norway, and to Jaqueline for being a devoted supervisor. Thanks also to the barnehage in Norway that supported this research, and to Ellen Beate Hansen Sandseter for providing valuable advice during my research project.

List of figures

Figure 1: Number of incidences of each category of risky play.

List of tables

Table 1 (Appendix 3): Distribution of triangulation

Appendices

Appendix 1

Observation definitions

The following definitions were used within the observations. They are based upon findings in Sandseter (2007a, 2007b, 2009b).

Categories of risky play:
• Great heights
Definition outdoors: Play that is conducted a few feet from the ground, including climbing apparatus and trees.
Definition indoors: Play will be considered at a standard 'table-height' (approx. 43cm) from the ground as indoor resources differ. This includes cabinets and tables.
Examples include: climbing, jumping from still/flexible surfaces, balancing on high objects, hanging/swinging at great heights.

• High speed
Definition outdoors: Movements outdoors which are observed as slightly faster than a gentle running/biking speed will be classified as risky play. Consideration will be given to whether children look in control of their movements.
Definition indoors: The same definition will be held, but children often move slower indoors; therefore, consideration will be given to increased speeds, including running, which is not common indoors.
Examples include: swinging at high speed, running uncontrollably at high speed, bicycling at high speed.

• Dangerous tools
Definition outdoors and indoors: Tools which have the potential to provide serious injury. This predominantly includes cutting tools.
Examples include: cutting tools such as knives, saws and axes; and strangling tools, such as ropes.

• Dangerous elements
Definition outdoors: Play involving natural elements which have the potential of causing harm to children will be identified as risky play. This includes weather conditions such as rain and ice. Examples include: play near cliffs, deep/icy water, fire pits.
Definition indoors: Any natural elements that could have the potential of causing harm to children that are used inside. Examples include: water or flames.

• Rough-and-tumble play
Definition outdoors and indoors: Rough, physical play with potential for physical injury. This can involve actions such as wrestling, or the use of tools, including sticks.
Examples include: wrestling, fencing with sticks, play fighting.

• Disappear/get lost
Definition outdoors and indoors: play out of the sight of adults, which children can find fun and thrilling (Sandseter, 2009b). This also involves play where children are absorbed in thinking they are out of sight of adults, but may not be.
Examples include: going exploring.

Appendix 2

Practitioner interview schedule

Introduction

For this study, risky play is considered regarding the notion of physical risk-taking with a potential result of physical injury, and the experiences children gain from this. It can be considered as play where children may feel out of control, maybe due to height or speed and experiencing and overcoming fear, as identified by Stephenson (2003).
In particular, Sandseter's (2007a, 2007b) defined six categories of risky play which are the main focus of this study. These include:
1. Great heights: such as climbing, jumping, balancing on high objects, and hanging and swinging at great heights.
2. High speed: such as swinging, sliding, sledging, running uncontrollably and cycling at high speed.
3. Dangerous tools: such as cutting tools, knives, saws and strangling tools.
4. Dangerous Elements: such as cliffs, deep water, fire pits, ice, rain.
5. Rough-and-tumble play: such as play fighting, fencing with sticks, wrestling.
6. Disappear and get lost: such as going exploring and playing out of sight of adults.
Please consider these categories in your answers.

Interview questions

1. To what extent do you agree with this definition?
2. Regarding the environment, what do you think are the differences between the indoor and outdoor environment?
3. Do you plan for the provision of risky play such as play at great heights, at high speeds, or rough-and-tumble play, both indoors and outdoors?
4. Could attitudes regarding the types of activities conducted inside, such as less physically active activities, effect the facilitation of physically risky play inside?
5. Could risky play activities (i.e. those within the definition provided) be facilitated indoors?
6. How do you feel opportunities for risky play indoors compare to those available outdoors?
7. Do you have anything else you would like to add?

Appendix 3

Distribution of triangulation
Table 1 provides a visual representation of the presence of each theme within each data set, helping to identify the prevalence and strength of each factor towards the facilitation of risky play indoors.

Theme	Observations	One-to-one interviews
Affordances	✔	✔
The natural environment	✔	✔
Space	✔	✔
Increased intensity	✔	✔
Affordances indoors	✔ (Play with resources at great heights)	✔ (Ease of supervision)
Alternative indoor venues	✔
Attitudes	✔	✔
Noise		✔
Sedentary activities indoors	✔	✔
Barnehage rules	✔	✔
Challenging Sandseter’s definition for being too restrictive in its categorisations	✔	✔

Table 1: Distribution of Triangulation.

Notes

1 Claire Saunders has recently graduated with a First Class BA (Hons) Childhood, Education and Society degree from the University of Warwick. During her final year she conducted a three-month study abroad placement at Queen Maud University College of Early Childhood Education, Norway, as part of the Erasmus programme. This placement and her interest in outdoor education influenced this study which formed her dissertation submission.

References