Content uploaded by Marketta Kyttä
Author content
All content in this area was uploaded by Marketta Kyttä on Jan 14, 2016
Content may be subject to copyright.
Child-friendly urban structures: Bullerby revisited
Anna Broberg
a
,
*
, Marketta Kyttä
a
,
1
, Nora Fagerholm
b
,
2
a
Department of Surveying and Planning, Aalto University, P.O. Box 12200, 00076 Aalto, Finland
b
Section of Geography, Department of Geography and Geology, University of Turku, 20014 Turku, Finland
article info
Article history:
Available online 20 June 2013
Keywords:
Child friendliness
Built environment
Affordance
Independent mobility
GIS
abstract
Definitions of environmental child friendliness offer broad criteria that are not easy to study or assess.
We suggest that due to this broadness, these definitions have produced surprisingly few attempts to
evaluate how child-friendly various types of physical environments are. The purpose of this study is to
analyse how the structure of the built environment contributes to environmental child friendliness. We
define child friendliness by two central criteria: children’s possibilities for independent mobility and
their opportunities to actualize environmental affordances.
We study how built environment qualities condition environmental child friendliness in place-based
ways by asking children and youth in Turku, Finland, to tell about their meaningful places and their
mobility to these. The data consists of over 12,000 affordances, localized by the respondents. This
experiential and behavioural place-based knowledge is combined with objectively measured data on
residential and building density, and quantity of green structures.
Moderate urban density seems to have child-friendly characteristics such as an ability to promote
independent access to meaningful places and the diversity of affordances. We find that affordances
situated on residential areas are likely to be reached alone, whereas access to affordances situated in
densely built urban cores is less independent. The proportion of green structures is not associated with
independent access. The diversity of affordances is highest in areas that are densely populated and not
very green. Green areas are important settings for doing things, and green structures around emotional
affordances increase the likelihood of liking the place significantly.
Combining children’s place-based experiences with information derived from objective measurable
qualities of the physical environment provides a valuable methodological contribution to studies on
environmental child friendliness, and the two proposed criteria of child friendliness are supported by
this study. There is no one environment that is child-friendly, but different environments have different
uses and meanings.
Ó2013 Elsevier Ltd. All rights reserved.
1. Introduction
The research literature offers an abundance of definitions con-
cerning environmental child friendliness, where the criteria for
child friendliness are often relatively broad and not easy to study
and assess (see Chatterjee, 2005; Horelli, 2007; Schulze & Moneti,
2007). These different definitions of environmental child friendli-
ness have produced surprisingly few attempts to evaluate the child
friendliness of various types of physical environments or to study
the structural variables of the urban fabric that contribute to this
matter. We argue that it may eat least partly ebe due to the
abstractness, broadness and vagueness of these definitions.
To deepen the understanding of urban characteristics promot-
ing environmental child friendliness, a more focused and oper-
ationalizable definition of environmental child friendliness is
needed. The Bullerby model by Kyttä (2008) is one candidate for
such an approach. According to this assessment model, environ-
mental child friendliness can be defined by two central criteria:
children’s possibilities for independent mobility and their oppor-
tunities to actualize diverse environmental affordances.
The Bullerby model is a theoretical tool for assessing the child
friendliness of various settings. In this article, we propose an
approach where the model is used to study how specific, built
environment qualities condition environmental child friendliness
in place-based ways. Our target in this paper is to combine both
children’s experiential and behavioural place-based knowledge
*Corresponding author. Tel.: þ358 505124554; fax: þ358 947024071.
E-mail addresses: anna.broberg@aalto.fi(A. Broberg), marketta.kytta@aalto.fi
(M. Kyttä), ncfage@utu.fi(N. Fagerholm).
1
Tel.: þ358 505124554; fax: þ358 947024071.
2
Tel.: þ358 2 333 5596; fax: þ358 2 333 5896.
Contents lists available at SciVerse ScienceDirect
Journal of Environmental Psychology
journal homepage: www.elsevier.com/locate/jep
0272-4944/$ esee front matter Ó2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jenvp.2013.06.001
Journal of Environmental Psychology 35 (2013) 110e120
with objectively measurable, place-based characteristics of specific
settings.
1.1. De finitions of child-friendly environments
The research literature offers an abundance of definitions con-
cerning environmental child friendliness. Themes like safety, avail-
able green space, variety of activity settings, independent mobility
possibilities, active socialization or “neighbourliness”, and integra-
tion of children into decision-making processes are often included
as essential criteria of environmental child friendliness (Freeman &
Tranter, 2011; Haider, 2007; McAllister, 2009). These kinds of broad,
extensive criteria are also the basis of the work of the international
network of Child Friendly Cities promoted by UNICEF. The Child
Friendly City Initiative (CFCI) encourageslocal governments to make
decisions that are in the best interests of children and promote
children’s rights to a healthy, caring, protective, educative, stimu-
lating, non-discriminatory, inclusive and culturally rich environ-
ment (Malone, 2001; Riggio, 2002; Schulze & Moneti, 2007).
An example of a more systematic definition of environmental
child friendliness embedded in both substantive and procedural
theories of a good environment is produced by Horelli (2007). The
resulting definition includes 10 normative dimensions: (1) Housing
and dwelling, (2) Basic services, (3) Participation, (4) Safety and
security, (5) Family, peers and community, (6) Urban and envi-
ronmental qualities, (7) Provision and distribution of resources and
poverty reduction, (8) Ecology, (9) Sense of belonging and conti-
nuity, and (10) Good governance. When children in different
countries were questioned about their thoughts on the dimensions
in the definition of environmental child friendliness by Horelli
(2007), only a few of these themes were brought up by the chil-
dren themselves. Safety and security, urban and environmental
qualities, and basic services were among the sets of criteria relevant
to children in Finland (Haikkola, Pacilli, Horelli, & Prezza, 2007) and
in Sweden (Nordström, 2010). When queried about these same
dimensions, Italian children mentioned urban and environmental
qualities and basic services in accordance with the Finnish and
Swedish children, but they did not mention environmental safety
(Haikkola et al., 2007). These findings resonate interestingly with
the earlier results of the Growing Up In Cities project (Chawla, 2002),
where the provision of basic services, the variety of activity set-
tings, and the freedom from physical dangers were also among the
factors that children from six continents and eight different coun-
tries indicated as primary indicators for a child-friendly environ-
ment. In addition to these three themes, green areas, freedom of
movement, and peer gathering places were also important positive
physical qualities of a child-friendly environment.
Chatterjee (2005, 2006) also finds the definitions of child
friendliness to be too broad and suggests that a child-friendly city
can only be studied as a disaggregation, made up of a number of
child-friendly places that children have a friendly relationship with.
She proposes a new theoretical concept of place-friendship that she
bases on a review of the literature on childhood friendship. Based
on the six dimensions of place-friendship, Chatterjee offers a
working definition of child-friendly places in a child’s everyday
environment, where these places:
1. provide opportunities for children to develop an attitude of
care for places that children love and respect;
2. promote a meaningful exchange between child and place
through affordance actualization in places;
3. offer opportunities for environmental learning and developing
environmental competence through direct experience in places;
4. allow children to create and control territories and protect
these territories from harm;
5. provide privacy experiences and nurture childhood secrets;
and
6. allow children to express themselves freely in place.
While we find Chatterjee’s conceptualization interesting, it still
seems to be relatively difficult to operationalize. In her dissertation,
Chatterjee (2006) questioned children in New Delhi about their
important places. Based on the data acquired from children, she
concludes that rather than having three separate dimensions con-
cerning activities of children in relatively constraint-free places,
dimensions number four and six (‘creating and controlling terri-
tories’and ‘freedom of expression in place’) could be included
under the higher level construct of ‘meaningful exchange with
places’, which introduces children to the affordances outdoors. She
thus proposes limiting the dimensions to four. Similarly, in their
recent study on Iranian children, Ramezani & Said (2012) inter-
viewed children about their important places using the place-
friendship framework and investigated whether the dimensions
can be reduced in number based on the data obtained on children’s
relations to different places. Their finding was that the six di-
mensions of place friendship could be reduced to the following
three: meaningful exchange with place, learning and gaining
competence through place experience, and having a secret place.
Meaningful exchange with place was seen as in parallel with the
actualization of affordances in place (Ramezani & Said, 2012) and
also represented the dimensions concerning the freedom of
expression, care and respect for the place, and creating territories.
What we find interesting in these two projects using Chatterjee’s
definition of child-friendly places is that the actualization of various
affordances seems to be central criteria for children’s friendly
relationship with a place when defined by children themselves.
Another critical view towards the abstractness of definitions of
child-friendliness has been aired by Whitzman, Worthington, and
Mizrachi (2010). They analysed how different Child-Friendly City
(CFC) initiatives in Australia have supported physical and social
transformations towards the institutionalization of children’s right
to the city. They see children’s independent mobility (in other
words, children’s possibility to autonomously explore the public
space) as children’s right to the city. In seven governments, they
reviewed plans on a general level and on lower level policies that
deal with young people. They revised these plans and policies in
regard to six elements: whether the plan (1) recognized children as
an interest group; (2) recognized children’s right to all public space,
not only those designed for children; (3) provided achievable tar-
gets, strategies and implementation mechanisms; (4) was inte-
grated into health and land-use planning; (5) included training for
administrators in child rights; and (6) had planners trained in
interacting with children. Interestingly, their policy scan showed
the narrow extent to which land-use planning policies were inte-
grated with CFC initiatives. The language or concepts of CFC were
not in use in the high-level plans governing land use and devel-
opment. Children were not mentioned as a specific group, but
rather in many implicit examples, they were assumed to belong in
specific places designed for children. Whitzman et al. (2010)
concluded that even if Child-Friendly Cities are a promising prac-
tice in its focus on the children’s right to independently roam the
public space, there are still difficulties in moving from the social
and health planning perspective that has informed these initiatives
towards impacts on land-use planning policies and practices.
1.2. Bringing the physical environment into the discussion
There are a few studies that evaluate environmental child
friendliness empirically, either on the neighbourhood, community
or city level. Among them are comparative studies by Kyttä (2002,
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120 111
2004) in Finland and Belarus, and work analysing the degree of
independent mobility of children in various settings (Fyhri &
Hjorthol, 2009; Fyhri, Hjorthol, Mackett, Fotel, & Kyttä, 2011;
Hillman, Adams, & Whitelegg, 1990; Tranter & Pawson, 2001). In-
terviews with families living in the city centre of Auckland, New
Zealand, revealed that the central location was seen as convenient
and less dependent on cars, while play spaces were insufficient,
apartments were not designed for family use, and fears for chil-
dren’s safety were prominent (Carroll, Witten, & Kearns, 2011).
These neighbourhood- or city-level studies do not offer a detailed
enough analysis of the physical characteristics contributing to a
child-friendly environment.
More information should be gathered on the environment’s
features that motivate everyday activity that are defined by chil-
dren themselves. This kind of information has so far been mostly
from small, qualitative studies (Veitch, Salmon, & Ball, 2007).
Among the few previous studies simultaneously studying chil-
dren’s own perceptions of their neighbourhood and the actual
potential for activity in specific physical settings is Wridt’s (2010)
research utilizing a qualitative GIS approach. She found signifi-
cant gender differences in patterns of use of the physical environ-
ment among U.S. children, and interesting differences between the
perceived places of danger and actual reported crime. This
intriguing study was, however, a small-scale project in one neigh-
bourhood with a very limited number of participants.
This is not to say that only perceived information provided by
children themselves is valid when the child friendliness of different
environments is studied. In fact, an interesting study from London,
Ontario, examined whether publicly provided recreational oppor-
tunities for children and youth are distributed unequally in the
spatial continuum (Gilliland, Holmes, Irwin, & Tucker, 2006). This
study looked at the spatial distribution of possibilities for recrea-
tion in relation to neighbourhood characteristics solely from
register-based data. In contrast, the structural qualities of the
environment are often approached solely from a subjective
perspective. For example, in a UK study, children told about their
perceptions of different elements of their physical environment in
connection with their independent mobility and participation in
play (Page, Cooper, Griew, & Jago, 2010). Studies of environmental
child friendliness that would look at more objectively measured
characteristics of the physical environment are still rare, and
studies that would combine children’s subjective experiences with
objective characteristics are almost non-existent. This is a gap the
current study hopes to fill.
In contrast to the few empirical studies concerning environ-
mental child friendliness as a whole, research about the mobility-
promoting qualities of the urban structure has yielded a large
empirical base on connections between an active, healthy lifestyle
for children and characteristics of the physical environment. (For a
recent review, see van Loon & Frank, 2011.) Residential density, the
proportion of green structure, a traffic environment that favours
pedestrian and light traffic, as well as accessibility to recreation
areas and versatile services are among the structural features of a
community that seem to support children’s active lifestyle and
independent mobility (Carver, Timperio, & Crawford, 2008; De
Vries, Bakker, Van Mechelen, & Hopman-Rock, 2007; Frank, Kerr,
Chapman, & Sallis, 2007).
On a more detailed level, features that promote traffic on foot or
by bike include sidewalks and bikeways, trafficlight-controlled
junctions, cul-de-sacs and well-functioning public transportation.
Children’s free and active movement is impeded by heavy traffic,
difficult junctions and a long distance to school (Bringolf-Isler et al.,
2008). In addition to structural features of the community, several
social, cultural and experiential features have been recognized that
are related to children’s activity possibilities in different kinds of
communities. For example, a large number of children and a strong
sense of community in the neighbourhood promote children’s
active mobility (Carver et al., 2005; Timperio et al., 2006).
Also in these studies concerning the mobility-promoting qual-
ities of the urban structure, the analysis of the environment’s fea-
tures has often been based on subjective observations made by
experts or children’s parents, whereas we suggest that the objective
features of the physical environment should rather be analysed by
using Geographic Information Systems (GIS). We agree with
McMillan (2005) and Woolcock, Gleeson, and Randolph (2010) that
questions concerning children’s active lifestyle and urban form
cannot be fully answered without a more thorough analysis of
micro-scale data on urban form and the social and ecological
variation that occur throughout cities. Nevertheless, more detailed
information is also needed on individual and household lifestyles
and place-based environmental experiences and perceptions.
There is also little debate about the relationship between more
broad definitions of child-friendly environments and more focused
aspects of the environment that motivate children to create an
inspired and individual relationship to their surroundings. The next
chapter offers one possible approach to this problem.
1.3. Environmental child friendliness in the light of the Bullerby
model
To deepen an understanding of characteristics of the urban
structure that promote environmental child friendliness, a more
focused and operational definition of environmental child friend-
liness is needed. The Bullerby model by Kyttä (2008) is one
candidate for such an approach. According to the assessment model
by Kyttä (2008), environmental child friendliness can be defined by
two central criteria: children’s possibilities for independent
mobility and their opportunities to actualize environmental affor-
dances. According to Moore (1986),“Access to and diversity [of
resources] emerge as the most important themes in child-
environment policy”(p 234).
The Bullerby model is built on the idea that the covariation of
independent mobility and the actualization of environmental
affordances (Gibson, 1986; Heft, 2001)define four qualitatively
different types of children’senvironments(seeFig. 1). A child-
Fig. 1. The Bullerby model for describing four hypothetical types and levels of child-
friendly environments (Kyttä, 2008).
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120112
friendly environment is primarily represented by the Bullerby
3
type of environment, where the abundance of mobility licenses
and actualized affordances create a positive cycle: the more
children can move around in the environment, the more and in
richer variety the affordances will be revealed. The actualization
of affordances motivates further exploration and mobility. The
opposite (i.e., the negative cycle) can also take place. In the latter
case, children are living in circumstances that are termed a Cell,
without opportunities to form a personal relationship with the
environment. In the Wasteland, possibilities for independent
mobility reveal only the dullness of the environment. Finally, the
child growing up in a Glasshouse lives in a condition where
external affordances are present, and the child can even be
aware of them but cannot actualize them autonomously. A
Glasshouse situation could occur when the media and other
sources of second-hand information give children the idea that
the environment is a field of ample affordances, but due to
mobility restrictions imposed by parents or sometimes by
communities, children do not have independent access to those
affordances.
While being aware that the Bullerby model does not include all
the essential criteria of environmental child friendliness, we argue
that the two selected dimensions are among the most crucial and
the most threatened in modern Western societies. It is also note-
worthy that the same physical environment might appear as a
Bullerby-type environment to one child and as a Cell-type envi-
ronment to another. The physical, social, and cultural environments
form an inseparable entity, the adaptation to which is partly
dependent on a child’s individual characteristics as well as the
social context (see Bronfenbrenner, 1993).
The Bullerby model is a theoretical tool for assessing the child
friendliness of various settings. In this article, we test whether the
model can be used to study how specific, built environment qual-
ities condition environmental child friendliness in place-based
ways. While the axis of independent mobility has been studied
extensively in relation to urban structure, less is known about the
axis of affordances. Our target in this paper is to combine both
children’s experiential and behavioural place-based knowledge
with objectively measurable, place-based characteristics of specific
settings. Linking the discussion of child-friendly environments to
actual places can also help communicate with land-use planners of
child friendliness, which has been shown to be problematic
(Whitzman et al., 2010).
2. Methodology
2.1. Design
This cross-sectional study focused on determining the re-
lationships between urban structure characteristics and children’s
environmental experiences and independent mobility. An Internet-
based softGIS survey (Kahila & Kyttä, 2009; Kyttä, 2011) was used to
study children’s environmental experiences and independent
mobility based on locality. In the softGIS survey, the respondents
used the Internet interface to mark on a map places that were
functionally, emotionally, or socially meaningful, and described
how accessible and likeable these places were. Respondents were
also asked to mark their home and daily routes to school and to
answer questionnaires concerning school journeys and perceived
health and well-being. Findings about active transport to school
and health and well-being are reported elsewhere (Kyttä, Broberg,
& Kahila, 2012).
The softGIS method used (see Fig. 2) is specially designed for the
use of children and youth. SoftGIS methods have been developed at
Aalto University since 2005 and have already been applied to
eleven Finnish cities, and about 9000 Finns have participated in
softGIS surveys. This methodology was honoured with the webGIS
innovation award in 2011 by Geospatial World Forum (Kyttä &
Kahila, 2011). SoftGIS methods allow residents to produce local-
ized experiential knowledge. As the experiences are gathered using
GIS-based methods, they not only comprise a separate experiential
world, but they also link to the physical environment. The localized
experiential knowledge that is gathered has coordinates and can
thus be analysed together with register-based or geographical data
included in geographic information systems (Kahila & Kyttä, 2009;
Rantanen & Kahila, 2009). The directing board of the research
project, involving public servants from multiple sectors of Turku
city administration, took part in designing the questionnaire items
to make sure the applicability of the data to planning purposes.
2.2. Subjects and communities
The study took place in the city of Turku, the oldest city in
Finland, with about 177,000 inhabitants. It is situated on the
western coast of Finland and consists of different living environ-
ments. The centre of the city is relatively dense and urban, whereas
the distant suburban areas are almost rural with their sparse land
use and open landscape.
In the coastal areas of Finland, both Finnish and Swedish are
spoken as national languages. The application was translated into
Swedish and English for the non-Finnish-speaking schools and
children who speak neither Finnish nor Swedish.
Data was acquired from 1837 5th-grade primary school pupils
(10e12 years old) and 7th grade secondary school pupils (13e15
Fig. 2. Front page of the softGIS children questionnaire and page for locating emotional affordances.
3
“Bullerby”can be literally translated as meaning a noisy village. The term is
used by the famous Swedish writer Astrid Lindgren (http://www.astridlindgren.se/)
in a number of her children’s novels where she describes the life of a group of
children living in a Swedish village, taking part in the normal everyday activities of
the village.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120 113
years old). Respondents represent 54 schools from varying
geographical locations in the city of Turku. A little over half of the
respondents were boys, and 52% belonged to the younger age
group. Of the children’s families, 37% lived in single-family houses,
33% in apartment blocks, and 30% in semi-detached or terraced
houses. Most children answered to the Finnish version of the
questionnaire (92%), while 8% answered the questionnaire in
Swedish and only 0.3% in English.
2.3. Sample and procedure
We conducted the study in two phases. In the first phase, all
elementary and secondary schools of the city of Turku were invited
to participate in the study. During 4 months, onlyabout 1000 pupils
had participated from the total of about 14,000 school children in
Turku. To gather more representative data, the data were collected
in an organized manner and were restricted to two age groups: fifth
graders (10e12 years old) and seventh graders (13e15 years old).
Between January and March 2008, two research assistants visited
54 schools willing to participate in the study again. Six schools
refused to participate. These schools were mainly schools with
special curricula (Steiner school, special schools).
Before organized data collection at schools, written consent was
obtained from children’s parents. The data collection was organized
in computer-equipped classrooms in the course of a normal school
lesson (45e60 min, depending on school policies) and was led by a
research assistant ethe teacher supervising the lesson. The chil-
dren responded independently but could ask for assistance from
either the research assistant or the teacher. The number of children
responding simultaneously varied between schools, depending on
the class sizes in different schools, but was generally around 20
pupils. The quality of Internet connections varied among schools,
and some schools faced technical problems.
A total of 3341 children participated during the second phase.
After exclusion from the database of children not in the chosen age
groups and responses received outside school hours (possible
multiple and/or unorganized answering), the sample size was
narrowed down to 1655 subjects. To gather as representative a
sample as possible, we included data from relevant age groups
gathered in the first phase from the six schools that did not
participate in the second phase (n¼182).
The final sample size was 1837 participants (1655 þ182). The
respondent rate of the second phase of data collection was 73%
(1655/2280). Of the 625 children that were not reached, 23% were
away from school during data collection, 24% did not obtain con-
sent from parents, and the answers of the rest (53%) were lost
either because of technical problems or because the child was
unable to finish the survey.
2.4. Measures
The degree of environmental child friendliness was studied on
two levels: children’s environmental experiences and independent
mobility, as suggested by Kyttä (2004). Children’s environmental
experiences were operationalized as localized affordances and their
likeability and diversity, and independent mobility as whether the
affordance is reached alone or in company of a friend or an adult.
We use both terms, independent mobility and independent access,
while discussing the independence of reaching the affordances.
2.4.1. Localized affordances
As potentially meaningful places for children, functional, social,
and emotional affordances were studied. The taxonomies used
were based on previous studies (see Table 1). The emotional
affordance category, “How does Turku feel?”, was based on Finnish
Table 1
Number of locations in the four affordance categories made bychildren in the
softGIS application.
Category, affordance No. of locations
Alone and together in Turku
I meet my friends 1108
I am in peace and quiet 651
I can be myself 412
Allowed place 317
I spend time with animals 309
I’m on my own 273
I meet new people 211
I am with grown-ups 200
Nobody is watching me 193
Forbidden place 158
Scary people 87
Place of bullying 78
I am lonely 56
Total 4053
What do I do in Turku?
I ride my bicycle 414
I play ball games 381
I run 324
Own category 321
I go swimming 292
I skate/ski 253
I go on the swings 192
I go sledding 127
I hide 99
I hang/dangle 95
I climb 91
I jump 79
I ride a skateboard 72
I play water games 70
I dig holes in the ground 54
I build things 49
Total 2913
Leisure time in Turku
I’m at my computer 469
I go shopping 301
I do sports 300
I just hang out 256
I have hobbies 234
I go to the library 179
I have fun 140
I go on adventures 133
I go to the cinema 130
I play 112
I eat out 112
I have nothing to do 97
Own category 67
I go to sports events 54
I go to a concert 18
I visit a museum 14
I go to see a show 11
Total 2627
How does Turku feel?
Safe 323
A good place to be 256
Peaceful, calm 224
Beautiful 208
Good air to breathe 207
Noisy 204
Boring 199
Dirty 188
Clean 160
Dangerous 125
Bad air to breathe 120
Ugly 106
Rowdy, rough 97
Quiet 88
Exciting 86
Own category 83
Dark 76
Total 2750
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120114
qualitative research on children’s emotionally meaningful places
(Miettinen, 2006). Some examples of the 17 different emotional
affordances questioned were peaceful or dangerous places. For the
social affordance category, an empirical study in Britain among the
same age groups (Clark & Uzzell, 2002) was applied. In total, 14
social affordances were queried under the heading “Alone and
together in Turku”, and included, for instance, places to be with
friends or forbidden places. The functional affordances were stud-
ied on a more general (activity) level, “Leisure time in Turku”, and
on a more specific level (action, operation), “What do I do in
Turku?”(Leontjev, 1978). The taxonomy for functional affordances
by Heft (1988), as interpreted by Kyttä (2002), was applied to study
the action-level functional affordances. Examples of activity-level
affordances included visiting the library and playing sports. The
action-level affordances were, for example, bicycling or climbing.
The former items represented the interests of the multisectorial
partners of our project from the city of Turku. In each category, the
order of the appearance of individual affordances in the survey
application was randomized. However, the order of the main cat-
egories on the front page of the application was stable. The survey
application allowed the respondent to map up to three localizations
for each affordance. This limitation was due to complexities the
programmers faced in storing variable amounts of data for each
respondent back in the year 2007.
2.4.2. Independent mobility
With each affordance marked on the map, the respondents were
asked how they reached the place. The options were alone, with
friends, and accompanied by an adult. Due to a mistake in the
application, this information was missing in all the localizations of
emotional affordances (22.3% of localizations). And because re-
spondents were not forced to answer the question, this data is
missing from further a 24% of other affordance localizations, in all
from 41% of the places.
2.4.3. Likeability index
Environmental likeability was addressed after the localization of
each meaningful place. The children responded on a sliding scale
from unpleasant to pleasant. The middle of the scale was marked.
The responses were stored as 0e100, with 50 representing neutral.
Unfortunately, the default value in the program was 50, and
genuine neutral responses were indistinguishable from the missing
responses. Therefore any responses of exactly 50 were discarded. In
further analysis, we used a dichotomous variable, where the value
‘1’represented positive likeability (over 50) and the value ‘0’rep-
resented negative likeability (below 50).
2.4.4. Bullerby grid
To test Kyttä’s (2004) Bullerby model of child friendliness on the
landscape scale, the original point data of the affordances set was
aggregated into a grid format using a 250-m cell size. The points fell
into a total of 1427 cells, but to obtain a valid result, cells having less
than three affordances were removed from further analysis
(N¼779). In each grid cell, the percentage of affordances reached
alone was calculated. (Emotional affordances were excluded from
this calculation because of the missing data on independent
mobility.) The overall diversity and the relative occurrence of the
different affordances were analysed with the Shannon diversity in-
dex, which is a popular measure of species diversity in ecology but
has also been used to study social data (Krebs,1989; Reed & Brown,
2003). Shannon’s index is based on information theory. It is a mea-
sure of uncertainty(disorder ina system) inpredicting what species
a random individual from a collection of species S and individuals N
belongs to (Ludwig & Reynolds, 1988). The Shannon diversity index
was calculated for every grid cell based on the relative number of
affordance points of each of the 64 themes in the grid cell.
4
Atwo-
base logarithm was used in the analysis (Krebs, 1989).
The grid cells were categorized into four environment types of
the Bullerby model according to the percentage of the affordances
reached alone (mean ¼35, SD ¼28) and to the Shannon diversity
index (mean ¼2.8, SD ¼0.9). The means were used to divide the
grid cells into the four categories. Thus, the axis of independent
mobility was
1. 35% or more of affordances within a cell reached alone, and
2. less than 35% of affordances reached alone;
and the axis of actualized affordances was
1. Shannon diversity index 2.8 or more, and
2. less than 2.8.
Using the share of affordances reached alone, rather than
without adult supervision, is a debatable choice. Children’s
mobility is predominantly social and, as has been noted among
Danish children (Mikkelsen & Christensen, 2009), independence of
mobility is not necessarily moving alone, but moving without
adults, among peers. Nevertheless, we’ve found that independent
mobility is a problematic variable in the Finnish context, where
children report very few restrictions on their mobility. To be able to
get some variation in the axis of independent mobility in the Bul-
lerby model, we use the narrow definition of reaching the afford-
ance alone as a measure of independence.
2.4.5. Mapping and GIS-based measures
The analysis of the structural characteristics of children’s
meaningful places was based on the geographical localizations that
children themselves gave while using the softGIS application. To
increase the reliability of the localizations, the softGIS application
helped children to orient themselves on the map. After the name of
a child’s school was given, the map automatically centred on it.
GIS-based measures of the urban structure were calculated
within a 50-m buffer of each affordance marked on the map and
into the 250-m grid cells of the Bullerby model. The urban structure
measures were:
Proportion of green structure: the proportion of fields, forests,
parks, and water of the total buffer or grid cell area. These were
calculated from the citywide cartographic data obtained from
the surveying department of Turku.
Residential density: housing units per hectare (hu/ha) within
the buffer or grid cell. The centroids for each building in Turku,
containing the information on housing units, floor areas and
population demographics for the building, were obtained from
the city.
Floor area ratio (FAR): calculated as the combined floor area of
the buildings within the buffer or grid cell divided by the area
not classified as green. In the buffer data, ratios over 10 were
dismissed as outliers (1.3% of buffers). These outliers were
supposedly due to incongruence in the geographical datasets
gathered on different scales.
Number of population within buffer or grid cell, calculated
from the above-mentioned building centroid data.
4
The Shannon index does not have a specific range but is dependent on the
richness and occurrence of the themes and the affordance points representing
them. Thus, a high diversity value indicates there are several of the 64 themes with
a large number of affordances marked in different themes present in the specific
grid cell.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120 115
2.5. Analysis
The research data were saved in a database from which the
children’s responses were written out in a table format, either so
that each respondent created one record or so that the record was
created by a single location (i.e., an affordance located by a
respondent). When a single respondent was the basic unit of
analysis, the respondent-based material was provided with
locality-based summaries from the data, such as the number of
different locations, information about the respondent’s home
environment, or the average distance to meaningful places. This
article covers the affordance-based material. The person-based
material has been reported upon separately (Kyttä et al., 2012).
The data were statistically analysed with PASW Statistics 18
software. The significances in the differences of means between
genders and the two age groups were studied using t-tests or the
ManneWhitney Utest, and the differences in frequencies using the
c
2
test. Analysis of variance (ANOVA) was used to study the sig-
nificances in difference between different categories of affordances
and the grid cells in the Bullerby model. The ANOVA results were
further analysed using Tukey’s test. The connections between the
urban structure characteristics and children’s environmental ex-
periences were studied with logistic regression analyses, as few of
the variables followed the normal distribution.
The GIS analyses were carried out with ArcGIS 9.3 and 10 soft-
ware and with MapInfo 8.3 software. Among the most central GIS
analyses were creating buffers around children’s homes and
calculating the residential density and the portions of green
structure and of children within these areas.
3. Results
3.1. Experientially meaning ful environmental affordances e
descriptive results
The children located altogether 12,343 affordances (see Fig. 3 for
the affordances around the city centre of Turku). The total number
of localized affordances by an individual child varied from 1 to 58,
with the mean being 8.2. The age groups differed significantly in
the number of localized affordances (Z¼5.8, df ¼1498,
p¼0.000), the younger children mapping almost 9 places per child
and the older only 7.5 on average.
The most often located functional affordances at the action level
(What do I do in Turku?) were bicycling, playing ball games, and
running. In the leisure-time activity category (Leisure time in
Turku), computer use, shopping, and playing sports were among
the most commonly located activities. Place mappings concerning
meeting friends, being yourself, and being in peace and quiet
dominated in the social affordance category (Alone and together in
Turku). And in the emotional affordance category (How does Turku
feel?), safe, feel-good, and peaceful places were marked most
frequently (see Table 1).
Boys reached their affordances alone significantly more often
(34%) than girls (28%), whereas two-thirds of the places marked by
girls were reached with friends (63%), the proportion for boys being
57%. These differences between genders were significant (
c
2
¼34.7,
df ¼2, p¼0.000). There was no gender difference in the proportion
of affordances reached in the company of an adult (9%). Interest-
ingly, there were no significant differences between the two age
groups regarding the company in which the affordances were
reached.
Affordances that were most often reached in the company of an
adult were seeing a show, going to a museum, and spending time
with adults. The smallest parental attendance was reported in
places of bullying, playing, hanging/dangling and adventuring.
Affordances most often reached alone were places where one gets
to be in peace and quiet, where one is lonely, spends time on a
computer or with animals.
In all, children liked the affordances they marked on the map.
There were small but significant differences between the age
groups and genders in average positivity towards the affordances:
girls liked their affordances a little more (mean likeability 80) than
boys (77) (t¼4.1, df ¼10,039, p¼0.000), and the older age group a
little more (80) than the younger (77) (t¼4.5, df ¼10,039,
p¼0.000).
The affordances that children liked the most on average were a
good place to be (mean ¼92, N¼212, SD ¼14), being at a computer
(mean ¼90, N¼384, SD ¼18), cleanness (mean ¼89, N¼123,
SD ¼18), and safety/security (mean ¼89, N¼268, SD ¼16).
3.2. Urban structure and affordances
Urban structure was significantly different in the four different
affordance categories. The differences applied to the floor area ratio
(F¼102.0, df ¼3, p¼0.000), the housing density of the built
environment (F¼29.8, df ¼3, p¼0.000), and the latter’sclose
companion population density (F¼33.7, df ¼3, p¼0. 0000) and for
a contrasting measure, the proportion of green structures
(F¼116.4, df ¼3, p¼0.000).
Activity-level functional affordances differed significantly from
all the other categories and were in the most densely built areas
(floor area ratio on average 0.75). Emotional affordances (mean
0.44) differed nearly significantly from both the social (mean 0.5)
and action-level functional (mean 0.38) affordances (Tukey’s
p¼0.02 and p¼0.018, respectively). Examples of affordances
where the FAR is highest are going to cinema, shopping, and
visiting library, museums and sporting events. Affordances with the
least densely built surroundings include actions such as sleigh
riding, climbing, and skiing or skating, but also beautiful places and
places where air feels good to breath.
Looking closer at the differences in housing density around
affordances, the differences distil into the action-level functional
affordances being situated in the least densely built surroundings
(mean 16 housing units per hectare). The mean housing density for
the other categories lies between 23 and 25 housing units per
Fig. 3. The affordances marked by the children in the city centre of Turku.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120116
hectare, the differences being not significant. Population-wise,
similar findings apply. The size of the population around action-
level functional affordances differs from all the other categories e
it is smaller. Also, emotional affordances and activity-level func-
tional affordances differ from each other, the latter being in
significantly more densely populated settings. The affordances
where housing and population densities are high include going to
the cinema, using a computer, playing, and spending time with
adults. These can be seen as affordances typical for residential
areas. Within the five most densely housed affordances are also
emotional places where the feeling is rowdy or rough (mean 34 hu/
ha).
The amount of greenery varies significantly between all the
categories. Action-level functional affordances (green structures on
average 44%) and emotional affordances (mean 33%) were located
in the most green environments. Leisure-time, activity-level func-
tional affordances (mean 27%) and social affordances (mean 29%)
were situated in the least green places. The most green affordance
surroundings are around places where respondents ski or skate,
swim, and play water games, but also around places experienced as
beautiful. The list of least green affordance surroundings includes
the already listed city-life affordances of shopping, visiting the li-
brary, and going to the cinema, buteven less green are places where
new friends can be made and places that are noisy.
The affordances marked by girls (mean percentage green 32%)
were in significantly less green surroundings than those marked by
boys (36%) (t¼6.1, df ¼12,309, p¼0.000). There were no differ-
ences between age groups. Correspondingly, the affordances
marked by girls were in significantly more densely built sur-
roundings (mean 23.7 hu/ha, FAR 0.58) than those marked by boys
(20.5 hu/ha, FAR 0.45) (t
hu/ha
¼4.2, df ¼12,309, p¼0.000 and
t
FAR
¼8.2, df ¼11,993, p¼0.000).
Interestingly, when the density of the built environment is
examined as housing density, the younger age group’s affordances
are in nearly significantly denser settings (mean 22.8 hu/ha) than
those of the older age group (21.3) (t¼2.0, df ¼12,309, p¼0.045).
But when we scrutinize density as floor area ratio, the opposite
holds true: the older age group has marked their affordances in
significantly denser surroundings (FAR 0.56) than the younger age
group (FAR 0.46) (t¼7.3, df ¼11,993, p¼0.000). This suggests
that the affordances of younger children concentrate on residential
areas, whereas those of older children concentrate on commercial
or central areas.
3.3. Urban structure effects on the independent access to and
positivity towards the affordances
Next we analysed the effect of the urban structure on children’s
independent access to their marked affordances. Gender was
included in all the logistic regression models but is not reported
upon separately. Housing density around affordances increased the
likelihood of a child coming to the affordance alone (OR ¼1.004, CIs
1.0 0 3e1.005, p¼0.000). The amount of population around the
affordance had a similar effect on the likelihood of accessing the
affordance independently (OR ¼1.004, CIs 1.003e1.005, p¼0.000).
An increase in the floor area ratio decreased the likelihood of
reaching the place alone (OR ¼0.786, CIs 0.734e0.843, p¼0.000).
Interpreting these results, affordances that are situated in resi-
dential areas are likely to be reached alone, whereas affordances
situated in densely built urban cores are likely to be reached with
others. The proportion of green structures did not have any sig-
nificant effect on the likelihood of independent access.
Some of the urban structural variables were associated with
children’s stated preference for affordances, namely population
around affordances (OR ¼1.002, CIs 1.001e1.0 0 4, p¼0.001) and
the proportion of green structures (OR ¼1.004, CIs 1.002e1.006 ,
p¼0.000). Gender and age group were included in these logistic
regression models, but their effect on the models is not reported
upon separately. An increase in the size of population around
affordances increased the likelihood of stating a positive value of
likeability, and interestingly, the result was different between
different affordance categories. The effect of the population
numbers on positivity was found specifically in the emotional
(OR ¼1.002, CIs 1.000e1.005, p¼0.03) and social (OR ¼1.004, CIs
1.0 0 2e1.00 7, p¼0.001) affordance categories. The more people
there were living around the affordance, the more likely the chil-
dren were to state a positive preference towards the affordance. The
green structures’positive effect on the likelihood of a positive
perception held only in the emotional affordance category
(OR ¼1.009, CIs 1.006e1.011, p¼0.000), and here the greenness
around the emotional affordance significantly increased the likeli-
hood of liking the place.
3.4. Testing the Bullerby model of child friendliness
The series of maps on Fig. 4 represent our operationalization of
the Bullerby model. The first of the maps shows the number of
Fig. 4. Child friendliness of the environments in the city of Turku as evaluated with the Bullerby model.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120 117
affordances in the 250-m grid cells and gives an idea of the
importance of the city centre to the respondents as well as the
abundance of affordance localizations around residential areas of
Turku. The diversity of different affordance categories in the cells,
shown on the second map, follows to some extent the overall
number of affordances; but diversity can also be great in cells
where the actual numbers are not the highest. Areas in Turku,
where the majority of the affordances are reached independently,
are scattered around the city, and generally the city centre gathers
lower levels of independent access, as can be seen from the third
map.
The last of the maps looks at the variety in affordances and in-
dependent access to them simultaneously and thus draws a picture
of the child friendliness of Turku, as understood according to the
Bullerby model. Large areas of the city centre fall into the Glass-
house category, where variety in affordances is large, but inde-
pendent mobility is on a low level. Bullerby types of environments,
where affordances are many and can be reached alone, border
Glasshouse environments near the city centre, but areas like this
are also scattered around the whole study area. Only larger
stretches of Cell types of environments can be found in the areas
south and southwest from the centre, on the way to the residential
areas on the islands.
According to our operationalization of the Bullerby model, 28.5%
of the cells were categorized into Cell and 26.6.% into Glasshouse
types of environments, while a little over one-fifth (21.6%) were
Wasteland environments and 23.4% were of the Bullerby type.
Variables concerning the urban structure, such as population
density or proportion of green structures, were calculated for the
grid cells, and variance in these structural variables between cells
categorized differently according to the Bullerby model was ana-
lysed. There was significant structural variation between the grid
cells in relation to size of population (F¼24.6, df ¼3, p¼0.000)
and proportion of green structures (F¼13.0, df ¼3, p¼0.000),
whereas the cells in the different categories did not vary in their
residential density or floor area ratio.
Looking at population numbers, the main result is that Bullerby
and Glasshouse types of environments differ from Wasteland and
Cell types (Tukey: Bullerby/Wasteland p¼0.000, Bullerby/Cell
p¼0.000, Glasshouse/Wasteland p¼0.008, Glasshouse/Cell
p¼0.000). Bullerby and Glasshouse represent environments where
the residential density is relatively high (on average 240 and 244
persons living in the cells, respectively), whereas Cell (mean 108)
and Wasteland (mean 126) are more sparsely populated. The pro-
portion of green structure differentiates the groups in a similar
manner, Wasteland and Cell forming the greener pair (means 42%
and 46%, respectively), Bullerby (33%) and Glasshouse (33%) being
the less green.
4. Discussion & conclusions
Combining place-based knowledge (first based on children’s
experience and second derived from objective measurable qualities
of the physical environment) provided a valuable methodological
contribution to studies on environmental child friendliness. The
study revealed that as a whole, the studied Finnish children
enjoyed widespread possibilities for active and independent
mobility and the building of a personal experiential relationship
with their outdoor environment. We chose to study environmental
child friendliness from the viewpoint of the Bullerby model,
concentrating on the affordances that an environment provides the
children and on children’s independent access to these. These two
criteria proved to offer interesting insights on different urban en-
vironments. The children had many personally meaningful places
in outdoor settings. The softGIS methodology used in the study
allowed children to mark personally meaningful affordances on the
map, and children located on average eight places per child.
Our results partly confirmed the previous findings from mobility
research about the ability of some urban environmental charac-
teristics to promote children’s independent mobility. We discov-
ered positive associations between residential and population
density and children’s independent access to affordances, whereas
building density, measured as floor area ratio, had a negative as-
sociation with children’s independence in reaching affordances.
Urban density around the affordances marked by the younger and
the older age group were different. Older children and adolescents
marked their affordances in central locations where building den-
sities were high, while affordances of the younger group were
located more in residential areas, where housing density and
population numbers were high. One might also assume that the
older age group reaches their affordances more independently than
the younger, be it alone or with peers, but age group differences in
reaching the affordance did not occur in our data. This might have
to do with the fact that the affordances of the older group are
further away from their home. Ideally, residential areas could be
developed to offer more intriguing places for youth. Indeed, not
only child friendliness but also youth friendliness of environments
should be studied (see also Woolcock et al., 2010).
Green areas are important settings for children’s experiences.
Especially action-level functional affordances were present in sur-
roundings where green structures were prominent. The greenest of
all the affordances were beautiful places. And indeed, when looking
at emotional affordances, the larger the proportion of green
structures, the likelier the positive evaluation of the place. This
reflects well-documented previous research literature on the
restorative qualities of green settings: the proximity to nature
associated with sparse building promotes mental health as a setting
for stress restoration (van den Berg, Koole, & van der Wulp, 2003).
We have also found that the amount of green structure around a
child’s home is positively associated with good perceived health
(Kyttä et al., 2012). Then again, different urban structures are
important for different experiences. As an example of this, popu-
lation numbers were bigger around social affordances than any
other category. Also, the more residents there were around the
social affordance, the more children liked the place.
The ability of the Bullerby model to reveal the essential char-
acteristics of the physical environment that contribute to envi-
ronmental child friendliness was supported in this study. The
approach was especially useful for the analysis of the resource
dimension, which is seldom studied in close connection with the
physical characteristics of the environment. Physical environments
where the diversity of affordances was vast were more densely
populated and less green than the less diverse environments. The
dimension measuring the independent access to affordances,
however, did not show differences between the urban structure of
the grid cells categorized into different environmental types of the
Bullerby model. The way the access of independence was measured
was also problematic, not only because there weremissing data but
because of the narrow definition of independence being used. In-
dependence of mobility is not necessarily moving alone, but
moving without adults, among peers (Mikkelsen & Christensen,
2009). Also, results from a previous study concentrating on chil-
dren’s activity levels in parks highlight the importance of inde-
pendence from parents or supervisors but with other children
present. The presence of parental supervision while the child was in
a park had a strong negative effect on children’s activity, while the
presence of other active children was strongly positively associated
with park-based physical activity (Floyd et al., 2011).
While we completely agree on children’s mobility being social,
in this study we used the narrow definition of reaching the
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120118
affordance alone as a measure of independence. This is due to the
low parental attendance that the children reported: 9% for both age
groups. In other, more restrictive mobility contexts, it would be
worthwhile to use the broader definition of independent mobility.
We also suggest that in future research, the accessibility dimension
should be studied, more broadly taking into account not only in-
dependent mobility, but also travel mode and frequency of visits,
possibly also the social equality of accessibility. Future research in
variable contexts could also enable the finding of some standard-
ized thresholds for the levels of independence and variability of
affordances that would have to be met for an environment to be
considered child-friendly. It is worth noting that green areas didn’t
appear as child-friendly in our Bullerby analysis. However, they
proved to be very important for the children in the affordance-
based analysis. This implies that measuring the emotional
response to the place along with the accessibility is important.
Despite the need for future research on the Bullerby model,
there was evidence that the two basic dimensions, the independent
access to and diversity of the affordances, are connected. Our pre-
vious study showed, for example, that the number of affordances
that children marked was correlated with physical activity on the
school journey (Kyttä et al., 2012). This strengthens the conception
that there is a close connection between the two central features of
a child-friendly living environment, independent mobility and
richness of affordances. The affordances were also very central to
children in the research using Chatterjee’s concept of the child-
friendly place (Chatterjee, 2006; Ramezani & Said, 2012). Maybe
the learning and competence building that Chatterjee (2005) pro-
poses as a criterion for child-friendliness could in effect be seen as
an outcome of the situation where a child lives in a child-friendly
environment, rather than characteristics of the environment. In a
similar line of reasoning, health outcomes are often attributed to
situations where a child (or an adult, for that matter) lives in an
environment that allows physical activity.
When discussing the child friendliness of environments,
concentrating on outdoor environments can be too narrow a view.
One of the affordances children themselves liked most was being at
a computer, and also the sheer number of these localizations was
the third most liked. Even if this can reflect computer use being one
of the few home-based affordances the children were offered to
locate, it should be noted that virtual and electronic environments
can act as important spaces for play, interaction and socialization
for children and youth.
The reliability and validity of research data collected via web-
based methodology demand critical evaluation. Finnish children
are probably used to communicating via the Internet because 99%
of Finnish households with children have Internet access (European
Commission, 2011). Earlier studies have shown that adolescents
tend to prefer a Web-based over a paper-based questionnaire, and
health-related surveys generally result in equal results regardless of
the method of implementation (Mangunkusumo et al., 2005).
However, web-based surveys can be vulnerable to mistakes in the
programming work, as was shown by the missing independent
mobility data on emotional affordances. Our research themes e
environmental experiences and independent mobility ewere
studied as perceived, subjective phenomena. To develop valid and
reliable measures, we used existing scales and taxonomies from
previous studies as much as possible. In most of the themes, self-
reports by the youths studied (ages 10e12 and 13e15) are prob-
ably more reliable than proxy reports from parents. And children’s
ability to report their experiences on a map can also be questioned.
Nordin and Berglund (2010) have researched the use of GIS-based
mapping methods with Swedish children aged 10e15 years and
have found them capable of reading maps and using a GIS appli-
cation for communicating their interests. Even if we did not
specifically research children’s ability to respond to the mapping
questions, it can be hypothesized that the skill level is on a similar
level as in Sweden. Orienteering is also a part of the national cur-
riculum in Finland.
Several limitations and strengths of the study should be
mentioned. First, as the softGIS method for children was used for
the first time in this study, some mishaps occurred. The order in
which the affordance categories were shown on the main page was
not randomized for different users, and thus the numbers of
different affordances, especially between categories, do not indis-
putably reflect the relative importance of these affordances for the
children. Data on independent mobility was missing from 41% of
the affordances. Half of this was due to a mistake in the application;
but the other half might be due to the vagueness of the definition.
Children might sometimes come to the places alone, while other
times with friends or adults; and thus it might be complicated to
pick just one option. Another limitation was the relatively low
quality of the register-based GIS data. This did not allow very fine
analysis of the urban environment. In future studies concerning
environmental affordances and mobility, the urban structural an-
alyses should include measures concerning the possibilities for
light traffic, public transportation options, and functions of the
urban environment, such as shops, restaurants, and playgrounds.
Finally, the Bullerby grid-based analysis of environmental child
friendliness seems to emphasize grid cells where there are plenty of
affordance localizations, and this does not necessarily reflect only
the importance of these places, but also the number of respondents
who live near that place. Then again, the Shannon diversity index
used to study the richness of the affordances is logarithmic and
takes into account the number of localizations.
The strengths of the study lie in the innovative, Internet-based
softGIS method that proved to be a promising way to study the
conditions of child-friendly living environments in a detailed
manner and in a way that also inspires children. Linking the dis-
cussion of child-friendly environments to actual physical environ-
ments in a simple way egrounded in the experiences of children
rather than in the ideals of researchers ecan also make the concept
more usable to urban planners. It is important to develop the
environmental child friendliness of whole environments and not
just focus on school journeys or places specially designed for chil-
dren. Children do not move around actively if the environment
does not offer them intriguing challenges and a rich variety of
possibilities for diverse activities. Children should also be seen as
abled and active users of their environment and as informants
possessing valuable insights into the possibilities and restrictions of
different environments.
Acknowledgements
The authors wish to thank the multisectoral steering group from
the city of Turku for extremely active and valuable co-operation.
We are very grateful to the headmasters, teachers, and naturally
the pupils of participant schools. The idea of using Shannon di-
versity index was born in discussions with colleagues at Laboratory
of Computer Cartography (UTU-LCC) at University of Turku. Mar-
ketta Kyttä worked as doctoral researcher of the Academy of
Finland during the project. The project was funded by Tekes, the
Finnish Funding Agency for Technology and Innovation, and the
Finnish Ministry of Education and Culture.
References
van den Berg, A. E., Koole, S. L., & van der Wulp, N. Y. (20 03). Environmental
preference and restoration: (How) are they related? Journal of Environmental
Psychology, 23(2), 135e146. http://dx.doi.org/10.1016/S0272-4944(02)00111-1.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120 119
Bringolf-Isler, B., Grize, L., Mäder, U., Ruch, N., Sennhauser, F. H., & Braun-
Fahrländer, C. (2008). Personal and environmental factors associated with
active commuting to school in Switzerland. Preventive Medicine, 46(1), 67e73.
http://dx.doi.org/10.1016/j.ypmed.2007.06.015.
Bronfenbrenner, U. (1993). The ecology of cognitive development: Research models
and fugitive findings. In R. H. Wozniak, & K. W. Fischer (Eds.), Development in
context: Acting and thinking in specific environments (pp. 3e44). Hillsdale, NJ,
England: Lawrence Erlbaum Associates.
Carroll, P., Witten, K., & Kearns, R. (2011). Housing intensification in Auckland, New
Zealand: Implications for children and families. Housing Studies, 26(3), 353e
367. http://dx.doi.org/10.1080/02673037.2011.542096.
Carver, A., Salmon, J., Campbell, K., Baur, L., Garnett, S., & Crawford, D. (2005). How
do perceptions of local neighborhood relate to adolescents’walking and
cycling? American Journal of Health Promotion, 20(2), 139e147.
Carver, A., Timperio, A., & Crawford, D. (2008). Playing it safe: The influence of
neighbourhood safety on children’s physical activitydA review. Health & Place,
14(2), 217e227. http://dx.doi.org/10.1016/j.healthplace.2007.06.004.
Chatterjee, S. (2005). Children’s friendship with place: A conceptual inquiry. Chil-
dren, Youth and Environments, 15(1), 1e26.
Chatterjee, S. (2006). Children’s friendship with place: An exploration of environmental
child friendliness of children’s environments in cities. North Carolina State
University.
Chawla, L. (2002). Growing up in an urbanising world. London: Earthscan.
Clark, C., & Uzzell, D. L. (2002). The affordances of the home, neighbourhood, school
and town centre for adolescents. Journal of Environmental Psychology, 22(1e2),
95e108. http://dx.doi.org/10.1006/jevp.20 01.0242.
De Vries, S. I., Bakker, I., Van Mechelen, W., & Hopman-Rock, M. (2007). De-
terminants of activity-friendly neighborhoods for children: Results from the
SPACE study. American Journal of Health Promotion, 21,312e316.
European Commission. eCommunications household survey: The results of a special
Eurobarometer survey. Available at http://ec.europa.eu/information_society/
policy/ecomm/library/ext_studies/index_en.htm Accessed 10.06.11.
Floyd, M. F., Bocarro, J. N., Smith, W. R., Baran, P. K., Moore, R. C., Cosco, N. G., et al.
(2011). Park-based physical activity among children and adolescents. American
Journal of Preventive Medicine, 41(3), 258e265. http://dx.doi.org/10.1016/
j.amepre.2011.04.013.
Frank, L., Kerr, J., Chapman, J., & Sallis, J. (2007). Urban form relationships with walk
trip frequency and distance among youth. American Journal of Health Promotion,
21(4 Suppl.), 305e311.
Freeman, C., & Tranter, P. (2011). Children and their urban environment: Changing
worlds. London: Earthscan.
Fyhri, A., & Hjorthol, R. (2009). Children’s independent mobility to school, friends
and leisure activities. Journal of Transport Geography, 17(5), 377e384. http://
dx.doi.org/10.1016/j.jtrangeo.2008.10.010.
Fyhri, A., Hjorthol, R., Mackett, R. L., Fotel, T. N., & Kyttä, M. (2011). Children’s active
travel and independent mobility in four countries: Development, social
contributing trends and measures. Transport Policy, 18(5), 703e710. http://
dx.doi.org/10.1016/j.tranpol.2011.01.005.
Gibson, J. (1986). The ecological approach to visual perception. Hillsdale, N.J.: Law-
rence Erlbaum Associates.
Gilliland, J., Holmes, M., Irwin, J. D., & Tucker, P. (2006). Environmental equity is
child’s play: Mapping public provision of recreation opportunities in urban
neighbourhoods. Vulnerable Children and Youth Studies, 1(3), 256e268.
Haider, J. (2007). Inclusive design: Planning public urban spaces for children.
Proceedings of the Institutio n of Civil Engineers. Municipal Engine er, 160(2),
83e88.
Haikkola, L., Pacilli, M. G., Horelli, L., & Prezza, M. (2007). Interpretations of urban
child-friendliness: A comparative study of two neighbourhoods in Helsinki and
Rome. Children, Youth and Environments, 17(4), 319e351.
Heft, H. (1988). Affordances of children’s environments: A functional approach to
environmental description. Children’s Environments Quarterly, 5(3), 29e37.
Heft, H. (2001). Ecological psychology in context: James Gibson, Roger Barker, and the
legacy of William James’s radical empiricism. Mahwah, N.J.: Lawrence Erlbaum
Associates Publishers.
Hillman, M., Adams, J., & Whitelegg, J. (1990). One false move.A study of children’s
independent mobility. London: Policy Studies Institute.
Horelli, L. (2007). Constructing a theoretical framework for environmental child-
friendliness. Children, Youth and Environments, 17(4), 267e292.
Kahila, M., & Kyttä, M. (2009). SoftGIS as a bridge-builder in collaborative urban
planning. In S. Geertman, & J. Stillwell (Eds.), Planning support systems best
practice and new methods (pp. 389e411). Netherlands: Springer. http://
dx.doi.org/10.1007/978-1-4020-8952-7_19.
Krebs, C. J. (1989). Ecological methodology. New York: Harper & Row Publishers.
Kyttä, M. (2002). Affordances of children’s environments in the context of cities,
small towns, suburbs and rural villages in Finland and Belarus. Journal of
Environmental Psychology, 22(1e2), 109e123. http://dx.doi.org/10.1006/
jevp.2001.0249.
Kyttä, M. (2004). The extent of children’s independent mobility and the number of
actualized affordances as criteria for child-friendly environments. Journal of
Environmental Psychology, 24(2), 179e198. http://dx.doi.org/10.1016/S0272-
4944(03)00073-2.
Kytta, M. (2008). Children in outdoor contexts affordances and independent mobility in
the assessment of environmental child friendliness. Saarbrücken: VDM Verlag Dr.
Müller.
Kyttä, M. (2011). SoftGIS methods in planning evaluation. In A. D. Hull,
E. R. Alexander, A. Khakee, & J. Woltjer (Eds.), Evaluation for sustainability and
participation in planning (pp. 334e354). London: Routledge.
Kyttä, M., Broberg, A., & Kahila, M. (2012). Urban environment and children’s active
lifestyle: SoftGIS revealing children’s behavioral patterns and meaningful pla-
ces. American Journal of Health Promotion, 26(5), e137ee148.
Kyttä, M., & Kahila, M. (2011). SoftGIS methodology eBuilding bridges in urban
planning. GIM International, 25(3), 37e41.
Leontjev, A. N. (1978). Activity, consciousness, personality. Englewood Cliffs, N.J.:
Prentice Hall.
van Loon, J., & Frank, L. (2011). Urban form relationships with youth physical ac-
tivity: Implications for research and practice. Journal of Planning Literature,
26(3), 280e308. http://dx.doi.org/10.1177/0885412211400978.
Ludwig, J. A., & Reynolds, J. F. (1988). Statistical ecology: A primer on methods and
computing. New York: John Wiley & Sons, Inc.
McAllister, C. (2009). Child friendly cities and land use planning: Implications for
children’s health.Environments: A Journal of Interdisciplinary Studies, 35(3), 45e61.
McMillan, T. E. (2005). Urban form and a child’s trip to school: The current literature
and a framework for future research. Journal of Planning Literature, 19(4), 440e
456. http://dx.doi.org/10.1177/0885412204274173.
Malone, K. (2001). Children, youth and sustainable cities. Local Environment, 6(1),
5e12. http://dx.doi.org/10.1080/13549830120024215.
Mangunkusumo, R. T., Moorman,P. W.,Van Den Berg-de Ruiter, A. E., Van Der Lei, J., De
Koning, H. J., & Raat, H. (2005). Internet-administered adolescent health question-
naires compared with a paper version in a randomized study. Journal of Adolescent
Health, 36(1), 70.e1e70.e6. http://dx.doi.org/10.1016/j.jadohealth.2004.02.020.
Miettinen, S. (2006). Lapsena lähiössä. 10e11-vuotiaiden näkökulma asuinalueensa
tiloihin [A child in a suburb. Views of 10 to 11 year old children towards the res-
idential spaces in their neighbourhood] [licentiate thesis]. Jyväskylä, Finland:
University of Jyväskylä.
Mikkelsen, M. R., & Christensen, P. (2009). Is children’s independent mobility really
independent? A study of children’s mobility combining ethnography and GPS/
mobile phone technologies. Mobilities, 4(1), 37e58. http://dx.doi.org/10.1080/
17450100802657954.
Moore, R. C. (1986). Childhood’s domain: Play and place in child development. London:
Croom Helm.
Nordin, K., & Berglund, U. (2010). Children’s maps in GIS: A tool for communicating
outdoor experiences in urban planning. International Journal of Information
Communication Technologies and Human Development, 2(2), 1e16. http://
dx.doi.org/10.4018/jicthd.2010040101.
Nordström, M. (2010). Children’s views on child-friendly environments in different
geographical, cultural and social neighbourhoods. Urban Studies, 47(3), 514e
528. http://dx.doi.org/10.1177/0042098009349771.
Page, A. S., Cooper, A. R., Griew, P., & Jago, R. (2010). Independent mobility, per-
ceptions of the built environment and children’s participation in play, active
travel and structured exercise and sport: The PEACH project. International
Journal of Behavioral Nutrition and Physical Activity, 7,17.http://dx.doi.org/
10.1186/1479-5868-7-17.
Ramezani, S., & Said, I. (2012). Children’s nomination of friendly places in an urban
neighbourhood in Shiraz, Iran. Children’s Geographies,1e21.
Rantanen, H., & Kahila, M. (2009). The SoftGIS approach to local knowledge. Journal
of Environmental Management, 90(6), 1981e1990. http://dx.doi.org/10.1016/
j.jenvman.2007.08.025.
Reed, P., & Brown, G. (20 03). Values suitability analysis: A methodology for iden-
tifying and integrating public perceptions of ecosystem values in forest plan-
ning. Journal of Environmental Planning and Management, 46(5), 643e658.
http://dx.doi.org/10.1080/0964056032000138418.
Riggio, E. (2002). Child friendly cities: Good governance in the best interests of the
child. Environment and Urbanization, 14(2), 45e58.
Schulze, S., & Moneti, F. (2007). The child friendly cities initiative. Proceedings of the
Institution of Civil Engineers. Municipal Engineer, 160(2), 77e81.
Timperio, A., Ball, K., Salmon, J., Roberts, R., Giles-Corti, B., Simmons, D., et al.
(2006). Personal, family, social, and environmental correlates of active
commuting to school. American Journal of Preventive Medicine, 30(1), 45e51.
http://dx.doi.org/10.1016/j.amepre.2005.08.047.
Tranter, P., & Pawson, E. (2001). Children’s access to local environments: A case-
study of Christchurch, New Zealand. Local Environment, 6(1), 27e48.
Veitch, J., Salmon, J., & Ball, K. (2007). Children’s perceptions of the use of public
open spaces for active free-play. Children’s Geographies, 5(4), 409e422.
Whitzman, C., Worthington, M., & Mizrachi, D. (2010). The journey and the desti-
nation matter: Child-friendly cities and children’s right to the city. Built Envi-
ronment, 36(4), 474e486.
Woolcock, G., Gleeson, B., & Randolph, B. (2010). Urban research and child-friendly
cities: A new Australian outline. Children’s Geographies, 8(2), 177e192. http://
dx.doi.org/10.1080/147332810 03691426.
Wridt, P. (2010). A qualitative GIS approach to mapping urban neighborhoods with
children to promote physical activity and child-friendly community planning.
Environment and Planning B: Planning and Design, 37(1), 129e147.
A. Broberg et al. / Journal of Environmental Psychology 35 (2013) 110e120120
