How does the built environment influence childhood obesity?

(1)

Jacob Evans 11124792

Master’s thesis in Sociology: Social Problems and Social Policy First Reader: Drs. G Bak-Veltkamp

Second Reader: Dhr.Dr. C Bröer 30/06/16

How does the built

environment influence

(2)

1

Acknowledgements

This thesis has been an immensely challenging but also immensely rewarding experience and the product of many hours of work. That said there are people who must be thanked for without them this thesis would not have been possible. To my fellow students in the thesis group and Mutsumi and Christian, thank you. The comments and critique I received were immensely helpful and challenged me to extend myself further, my own perceptions and knowledge were challenged and this thesis is much the better for it.

A special mention must go to my supervisor Gerlieke whose support has been invaluable, your patience, enthusiasm, extensive knowledge and excellent example has elevated this thesis and inspired me to reach a level that would not have been possible without you. You’ve been an excellent supervisor and I wish you all the best in your own research.

To my friends’ new and old thank you for your enduring support through what at times has been a very stressful time, from boring you with the details to being at times annoyingly unsociable thanks for your continued support.

Finally I must thank my parents. Without your love, humour and support this would never have been possible. You inspire me to extend myself and reach my potential, thank you for your support now and in the future.

(4)

3

Introduction

The global spread of obesity has alarmed publics, politicians, medical professionals and policy practitioners (Strauss. 2000). This spread and growth of obesity has been many years in the making and the current trends that are in place in many modern nations show little sign of reversing the current cycle, the world health organisation in recognition of the spread of obesity has deemed it an ‘epidemic’ (WHO. 1997). So too in the United Kingdom has the spread of obesity alarmed many as the growth of obesity incidences among both the adult and child population have reached worrying highs.

Despite the global recognition of obesity as a major cause of concern, few studies have been able to establish what may be behind such increases. Some posit that modern lifestyles have led to

increasingly sedentary populations as well as the increasing availability of convenient and energy dense foods. However little focus in European studies has been placed on the role of the built environment as an influence of both child and adult levels of obesity. Studies in North America and also Australia have shed some light on how the built environment can contribute to the physical activity and ultimately weight status of its inhabitants.

Though still in its infancy the use of Geographical Information Systems (GIS) by health researchers is gaining attention due to its versatility and ever increasing pool of available data. It is with GIS that this research intends to explore the link between the built environment and childhood obesity. In particular using this information system to assess the walkability of an area and also assess the availability of amenities within walking distance of a certain area. Previous studies using GIS have found mixed findings when considering the relationship between population density and obesity; whilst some studies have previously found that urban areas in the United States have been associated with higher levels of both adult and childhood obesity, more recent studies have found that urban sprawl; characterised by large areas of low density populations are actually more strongly associated with higher levels of obesity (Lopez. 2004. Lake. 2006).

There are a variety of reasons that this link between area and obesity level could be the case, some which are measurable through GIS and others that will require further research at a later time. Some studies have posited that areas characterised by lower population densities are too reliant on vehicle transport in order to reach certain amenities such as schools, supermarkets and places of work. This reliance on vehicular transport to reach necessary areas has led to an environment that is hostile to physical activity, taking this further in previous studies there has been found to be an ‘obesogenic’ environment, one constructed by town and city planners that have a negative impact on both child and adult obesity levels (Lake and Townshend 2009).

The aim of this research is to try and identify features in the built environment that could contribute to higher levels of childhood obesity in the United Kingdom. The way this will be operationalised is by comparing twenty high obesity and twenty low obesity areas together through GIS analysis to identify the potential similarities or differences between them. For example do high obesity areas have a lower level of street connectivity making walking more time consuming, or do lower obesity areas have a higher proportion of the area within walking proximity to a school making essential journeys practical through walking rather than using vehicular transport. These are questions that

(5)

4 this research will aim to provide a preliminary answer to as well as a jumping off point for further research.

Obesity as a Disease

A simple definition of obesity is offered by the National Health Service (NHS) in the UK; ‘very overweight with a lot of body fat’. Whilst simply put, this does not offer a good insight into the nature of obesity and what its effects or causes are. A definition such as the one offered above suggests that obesity can be simply put down to “eating too much and moving too little”. Weight gain and obesity are posing a growing threat to health in countries all over the world. Obesity is a chronic disease, present in both developed and developing countries, and affecting both children as well as adults.

Obesity can be considered to have a number of influences and fallouts; in particular the increased risk and incidence of non-communicable conditions such as, sleep apnoea, arrhythmia, chronic joint pain and high blood pressure (Swinburn and Egger. 2002). These conditions are not only limited to physically diagnosable ones; in fact obesity has been found to have a link between numerous health conditions and has been viewed as both a cause and effect of depression (Foresight. 2007). The spread of obesity worldwide has caused alarm to numerous international health organisations and national governments. The spread of obesity has been so extreme that it led to the world Health Organisation to declare an obesity epidemic in 1997 (WHO. 1997). In its 1997 report Obesity: Preventing and Managing the Global Epidemic, the World Health Organisation offered a slightly more detailed definition of obesity with a few extra notes of guidance;

‘Obesity can be defined simply as the disease in which excess body fat has accumulated to such an extent that health may be adversely affected. However, the amount of excess fat, its distribution within the body, and the associated health consequences vary considerably between individuals’ (WHO.1997. Pg.6)

Whilst many health systems and organisations accept the definition offered by the World Health Organisation finding and implementing an objective system of obesity measurement has not been so straightforward. There have been problems on implementing and investigating standard

measurements of obesity measurement. The most commonly used method of judgement has become that of Body Mass Index (BMI) to identity and measure individuals who are currently or potentially at risk of being overweight or obese (Strauss. 2000). In an attempt to overcome some of the shortcomings of BMI measurements there has also been a use of measurement of

intra-abdominal area fat and waist circumference so as to more accurately measure and assess obesity in both adults and children (Public Health England. 2011).

In children in particular there has been difficulty in using BMI measurement to objectively and accurately record weight status. A reason for this is the variety of sizes and growth patterns that children and adolescences will experience throughout the course of their youth. One downside of BMI is that it doesn’t take into account muscle mass when measuring weight, meaning that a young athlete could potentially be grouped together with an individual who is holding a large amount of body fat (Public Health England. 2011). As such the heights and weights of children can range widely

(6)

5 within the same age cohort, a fact that makes measurement both difficult to conduct and also to compare results.

The rise and spread of obesity has been one that has been growing over time, in 1995 there were an estimated 200 million obese adults worldwide, by the year 2000 that number had increased to over 300 million (http://www.who.int/nutrition/topics/obesity/en/). In the United States it is projected that if trends continue at their current projections in a few years 70% of people will be overweight and around 40% obese (Xu and Wang. 2015).The growth is not confined to adults alone as more and more children are overweight and obese than in previous periods of time. Many campaigns at the local, regional, national and international level have tried to raise attention to and solve the growing problem of both child and adulthood obesity. In recent years there has been increasing attention given to numerous schemes and initiatives to combat obesity, in particular North America which has the highest documented levels of both childhood and adult obesity have published many academic articles dealing with obesity and both its potential causes and remedies. Organisations such as the Campaign to End Obesity and also increased media coverage in the United States have led to more attention being placed on Obesity as a health priority. In 2012 the US television network HBO aired a four part series titled; “Weight of the Nation” highlighting the growing body of research dedicated to obesity and also the implementation of alternative solutions to try and tackle the problem.

In recent years the levels of funding allocated to obesity research have imploded as new areas of interest arise and are examined with the US Department of Health in recent years providing millions of dollars to State bodies and departments to investigate obesity further and simultaneously

promote healthy eating and exercise ( Wang, Wen and Xu. 2013).

A body of research emerging recently has looked to examine the socio-cultural aspects of obesity; in particular how human environments can influence obesity (Berke et al. 2007). As is to be expected with a phenomenon as complex as obesity it is not surprising that various forms of the human environment have been implicated as having an impact. Barnet and Casper (2001) offer an extensive definition of the range of human social environments;

‘Human social environments encompass the immediate physical surroundings, social relationships, and cultural milieus within which defined groups of people function and interact. Components of the social environment include built infrastructure; industrial and occupational structure; labor markets; social and economic processes; wealth; social, human, and health services; power relations; government; race relations; social inequality; cultural practices; the arts; religious institutions and practices; and beliefs about place and community. [ … ] Social environments can be experienced at multiple scales, often simultaneously, including households, kin networks, neighborhoods, towns and cities, and regions.’

Whilst Barnet and Casper address the extent of human social environments that are present this also poses some challenges both for academics and policymakers. Such a wide range of human social environments poses problems both for research and policy; simply put the range of potential human social environments influencing obesity is almost too wide to pinpoint exact correlations or

relationships. What this does show is that a simple approach to tackling obesity is not a feasible option, to combat the spread of obesity in both children and adults effectively it will be necessary to

(7)

6 target numerous areas (Curran et al. 2006). The majority of academic research has so far largely overlooked the role of components of the human social environment, choosing by on large to focus on the circumstances or categories that people are placed into within society. A key shortcoming with this focus of research is that many of the categories that are examined are permanent or cannot simply be adjusted; a great deal of research has focused on the link between race and ethnicity for example (Lake. 2006). Whilst there is a substantial body of research that has focused on the link between race and obesity rates little progress has been made to explain why such a

correlation may be present.

Such research does little more than pinpoint races that are statistically more likely to find

themselves overweight or obese. The same can be said of research that focuses on the link between socio-economic status or social class and obesity. Little other than observational findings have been produced, simply stating the presence of a correlation with few feasible recommendations for policy makers or researchers to try and uncover the root cause of these correlations (Lake and Townshend. 2006).

Obesity as a financial cost

Obesity is undoubtedly a large drain on government health expenditure worldwide with millions of Euros being spent every year in treatment and research (UN. 2013). When considering the financial costs of obesity they can be broken down into three areas;

1) Directly to the Individual 2) To Employers and Workplaces 3) To the State and the taxpayer

These three areas adequately categorise the costs of adult obesity, however a slightly altered approach must be taken for childhood obesity. A key component to add would be that of poor academic performance, though the previous categories are still applicable (Heshmat. 2014). Though hard to calculate the financial costs of obesity can be high. For the person dealing with obesity there can be significant medical costs and higher insurance payments for longer bouts of associated ill health, also though exact numbers are hard to pinpoint there can be a significant financial loss through missed days of work, especially if the person is in a form of insecure employment (Butland et al. 2007). In regards to absenteeism the financial cost is potentially shouldered by not only the employee but also the employer. Though exact numbers are hard to calculate it was estimated that in the US loss of productivity was higher than $66 billion annually and similarly in the United Kingdom the indirect costs of obesity (including loss of productivity) between the years 1998 and 2007 was estimated at between £2.6 billion and £15.8 billion (Hammond et al. 2010; Butland et al. 2007).

The direct costs of obesity to the state and ultimately the taxpayer are also difficult to provide an exact figure for. There is no doubt that the increasing rates of obesity have put significant strains on health services in terms of financial expenditure and resource allocation for longer periods of illness and emergency visits (Foresight. 2007). These strains will ultimately result in heavier burdens being

(8)

7 placed upon the state to try and accommodate this extra demand for health services, significant financial resources will have to be accumulated to fund health services as well as to fund research and development all of which requires significant civil service involvement (Public Health England. 2011). In the US medical expenditure on obesity was calculated as nearing $147 billion annually (Finkelstein et al. 2009). Similarly the costs to the NHS in the United Kingdom are also very high, for the year 2006/7 the cost to the NHS in in patient stays was at £148 million in England alone, since then the level of obesity and consequently spending has increased (Dr Foster Research. 2008). Worldwide obesity is the third of the three top social burdens generated by humans and costs an estimated $2 trillion a year. Internationally the only two which exceed obesity in financial cost are smoking and armed violence and terrorism (Mckinsey. 2014).

In a 2014 international study it was found that obesity was second to smoking in terms of costs to the health services in the UK, similarly the majority of UK spending on obesity is aimed largely at treatment rather than seeking to take preventative measures against the disease (Public Health England. 2011). Obesity is not a purely financial cost on society, though the numerical data give an insight into its scale and reach. Further the current approaches to reducing obesity are inadequate and require significant restructuring, obesity cannot be solved by one sector or set of actors alone, what is required is a wide and coordinated response involving various societal actors and sectors.

‘any single intervention is likely to have only a small impact at the aggregate level. Our research suggests that an ambitious, comprehensive, and sustained portfolio of initiatives by national and local governments, retailers, consumer-goods companies, restaurants, employers, media organizations, educators, health-care providers, and individuals is likely to be necessary to support broad behavioural change’ (Mckinsey. 2014).

Obesity in the United Kingdom

This section will deal with the historical rise of obesity in the United Kingdom, its future projections and current strategies in place to try and combat the problem. The level of both child and adulthood obesity in the United Kingdom is at an alarmingly high level, by far the highest in Western Europe with one in four adults being obese according to the United Nations Food and Agriculture

Organisation (UN.2013). Statistics from the same report state that obesity levels have more than trebled in the last thirty years from around 6% of men and 8% of women in 1980 to a staggering 24.9% in 2013. These increases have not gone unnoticed; earlier in 2016 the Health Secretary Jeremy Hunt labelled increasing rates of childhood obesity as a “National Emergency” and Tam Fry of the National Obesity Forum has gone as far to say that;

“For the UK, obesity will be a national tragedy if the current government's indifference to stemming it is not quickly reversed."

Researchers from Imperial College London have recently warned that as soon as 2025 British people will be the most overweight in Europe if current projections are not curbed. In Europe Malta and Turkey currently have the largest proportion of obese people; however the UK is likely to overtake them both by 2025 with the highest proportion of obese women at 37% and highest proportion of obese men alongside Ireland at 34% (Mckinsey. 2014). The rise in the level of obese women has led

(9)

8 to the Chief Medical officer for England to call on the government to take action, women are at a greater risk of becoming obese than their male counterparts both in the UK and in other Nations the world over (UN. 2013). The graph below represents the European rise of obesity between 1975 and 2014;

(UN. 2013)

Whilst all Nations represented above have experienced increases in their levels of obesity some are far starker than others. Ireland has experienced the most significant increase in obesity from 5.1% to 25.2% and in the UK it has rose from 9.3% to 28.4% a massive increase by any standard. The logical question then is; what explains this rise? There are various theories offered as to how the UK has become the nation with the highest levels of obesity in Western Europe and projected to be the largest in the whole of Europe by 2025. One theory put forward by researchers is that there has been a shift in the British lifestyle in the last 30 years which has resulted in an energy imbalance for many (Foresight. 2007). Modern life has been stylised as making exercise more of an inconvenience as many people live largely sedentary lives and forego exercise whilst using vehicular transport to forms of employment that are long hours of mainly clerical and unphysical work (Butland et al. 2007). Over the years the average BMI of a British person has increased; in 1975 the BMI of the average Briton was 23, considered a healthy weight. Currently the average BMI stands at 27, not only is this classified as overweight but also means that since 1975 on average every British person has gained 1.5 kilograms per decade (McKinsey. 2014).

Such increases will have a dramatic effect not only economically but also socially for the UK, in a recent government report it was suggested that this pattern of weight gain was leading to the UK becoming an “Obese Society”, one where being obese or overweight is considered to be normal and leads to an attitudinal shift towards what is considered a normal weight or normal lifestyle. Whilst the scale of the problem is almost universally accepted in many recent studies, the measures necessary or the timescale with which progress can be made is subject to opinion. The recent Mckinsey report stipulated that a reversal of the current trend could be achieved within twenty years, however a government paper opined that the current epidemic is three decades in the

(10)

9 making and progress will take a similar amount of time (Mckinsey.2014; Tackling Obesities: Future Choices – Executive summary. 2013).

What is often made clear but rarely acted on in a meaningful way is that a response to obesity will require various different approaches. Such a complex disease will require action in a variety of areas if there is to be a reversal of the current trend, though not limited to the NHS the body is currently bearing the brunt of the epidemic with costs continuing to rise. The vast majority of spending is concentrated on treating obesity and its related illnesses; an almost insignificant amount in comparison is spent on taking preventative or curative measures (Lake and Townshend. 2009). A variety of areas need to be targeted if the energy imbalance currently in place for many children is to be overcome. Whilst in essence this is what is necessary to effectively redress the energy imbalance many areas of everyday life and human social environments need to be targeted to facilitate easier and healthier choices than what is in the current situation. The following graph produced by the Foresight commission on obesity visually displays the numerous causes and factors of both child and adult obesity;

(Foresight. 2007)

To effectively combat obesity all of the following areas need to be in some way addressed or moderated for as in the case of biology. Policy and research should aim at addressing the above mentioned areas, especially those which can be altered or changed by effective policy intervention. When addressing childhood obesity much has been made of the role of societal influences and individual psychology, especially the psychology of parents and their associated parenting

techniques. As has been suggested by numerous reports the role of the parent is now becoming an area of intervention for policy practitioners, often characterised by education and training

programmes for parents to give them the knowledge to help provide their children with healthy lifestyles (Public Health England. 2011). Whilst such techniques are employed in the UK and other European countries such as the Netherlands there has been little assessment of the long term effectiveness or the implications of such training schemes (Foresight. 2007). In the United Kingdom there is not a standard programme in place that equally allocated resources for parent training across the country. Instead programmes are often offered in association with the NHS by another

(11)

10 organisation rather than being directly provided for (Public Health England. 2011). As such access to these schemes for participants can depend largely on geographical location, in such a situation it is difficult to envision how a proper nationwide assessment of these programmes can be feasible. Some research has cast doubt on the effectiveness of parenting interventions as a solution to childhood obesity, citing the limited scope for application and also the need to address many other aspects of society that influence day to day practices (Barnet and Casper. 2001). Whilst parenting is key to lowering childhood obesity it is only one of a set of numerous and interrelated factors, a parenting intervention is a small scale targeted effort available in only a few cases that are put before the relevant organisations. Such a scope is inadequate if policymakers are serious about tackling childhood obesity, what is needed are interventions that are both cost effective and can make large impacts on a national or regional scale (Hammond et al. 2010).

Whilst parenting is of huge importance providing the necessary knowledge will only make so much progress, what is needed is a shift of societal perception and environment to facilitate putting good practices into action. Such alterations must be made in numerous areas if progress is to be made, from the level of physical education in schools to the regulation of marketing strategies for certain foods (Mckinsey. 2014). The 2014 Mckinsey report identified the need for 44 interventions across the following 18 areas in order to bring 20% of those who are overweight or obese back to normal weight within five to ten years;

Education;

1) Public Health Campaigns 2) School Curriculum 3) Parental Education Personal Responsibility;

1) Weight Management Programmes 2) Healthy Meals 3) Surgery 4) Pharmaceuticals 5) Active Transport Environment; 1) Workplace Wellness 2) Labelling

3) High fat, salt and sugar food access 4) Health Care Payors

5) Reformulation 6) Media Restrictions 7) Price and Promotions 8) Urban Environment 9) Subsidies and Taxes 10) Portion Control

(12)

11 (Mckinsey. 2014. Pg2)

What must be noted is that whilst levels of obesity have been rising conversely levels of physical activity in the UK have been decreasing with less than half of the population performing the

recommended levels of weekly physical activity (Mackett and Brown. 2011). In recent years due to a shift to a service sector economy coupled with longer working hours has left British society more sedentary than at any other time in history. What this has translated to is less opportunity and less time to be physically active for both adults and children (Public Health England. 2011). In recent years the need to find a multi area targeted programme against obesity has led to some interesting findings and suggestions by researchers. The previously mentioned Mckinsey report recommended action across 18 areas. The Future Choices summary also advocates concerted action across numerous areas, whilst the two reports have stark differences in projected time frames both manage to capture the essence of the necessary solutions, being alterations to the current patterns of life and environments in which people currently live.

The Importance of Childhood Obesity

Obesity can be particularly problematic in children and adolescence’s. Developing obesity or being overweight can cause significant complications for a child in such a crucial period of physical and mental development. A 2007 US study found that obese children are more likely to have lower levels of self-esteem, often associated with feelings of loneliness and nervousness around peers, also obese children with decreasing levels of self-esteem were more likely to drink or smoke than their other peers in their teenage years (Strauss. 2000). As Brownell and Wadden (1984) state;

“The professional community is concerned with the medical concomitants of obesity, but the psychological and social perils are at least as important to those afflicted by the problem. The reason is clear; society does not tolerate excess weight. The effects of this overt and covert pressure to be thin can be powerful and permanent.” (Cited in Strauss. 2000. Pg.5)

As with adult levels of obesity worldwide there has been a significant increase in the level of obese children throughout the world, highlighting the importance for both academia and policymakers. A recent study published by the World Health Organisation said that between 1991 and 2014 the number of obese or overweight children under 5 had increased from 31 million to 41 million with many remaining obese into their later years and adolescence (WHO.2014).

The psychological and psychosocial costs of childhood obesity can be just as damaging to the sufferer as the physical risks also. Overweight children can potentially find themselves developing depression and can be ostracised by their peer groups due to their weight (Beck. 2016). Similarly both in the home and at school obese children can be subjected to a form of stigma viewing the failure to be at a healthy weight as being caused by poor personal qualities such as idleness or lack of will power (Washington. 2011). In the United States previous studies have found that obese children are perceived as lazy by their peers and are less likely to be selected as a friend whilst obese children are also 67% more likely to be bullied than children at a normal weight (Puhl.2013. Lumeng et al. 2010).

(13)

12 Physically, obese children are much more likely to develop non communicable diseases such as diabetes and are also more likely to become obese adults (Serdula et al. 1993). Being overweight or obese can have a significant impact on a child’s enjoyment of life and make it much harder to take part in social activities which involve physical activity such as sports teams (Lumeng et al. 2010). Academically also obese children are more likely to underperform or miss a higher number of school days than their non-obese peers (Puhl. 2013).

The physical and emotional costs of childhood obesity for the individual are high, the cost to society as a whole is potentially even higher financially and socially. As previously mentioned childhood obesity can lead to problems academically and socially, problems that can potentially be carried into adulthood with many having potentially underperformed in the classroom and having failed to develop psychosocially (Lumeng et al. 2010). Further, there is a documented link between obesity and mental health problems in both children and adults; a rise in obesity rates has the potential to in turn increase incidences of mental health problems within the population (Hammond et al. 2010). The potential for a rise in physical health conditions is also well documented with the rise of obesity; the potential risk of various non communicable diseases is much higher for obesity sufferers and can have many short and long lasting effects. As such obesity is associated with an increased risk of morbidity, long term disability and mortality (Visscher et al. 2001). Many of the conditions mentioned have also been linked to similar higher levels of vulnerability among young children (WHO.2007).

New Focus

By approaching obesity as a disease from a different conceptual framework much more progress can be made towards uncovering the reasons why some people suffer from obesity much more than others. The idea of the human social environment allows a researcher to look at the social world with a much more critical eye than rather searching for correlations. While potentially overwhelming to approach all aspects of the human social environment, by focusing on key components there is the potential to gain an understanding of how the environments human’s create both physically and conceptually influence obesity rates.

Rather than viewing the relationship between socio economic status and obesity as correlational there is the potential to uncover why this relationship truly exists or to at least shed more light on the problem. It could be that people from a lower socio-economic status find themselves

disproportionately housed in areas that are not made with physical activity or health eating choices in mind (Gorden-Larsen et al. 2006). Looking at the structural barriers to living at a healthy weight affords far more opportunities to uncover previously overlooked factors which may help explain why some groups are more likely to be obese or overweight than others (Barnett and Casper. 2001). When considering which components to examine if considering structural constraints, the built environment is a strong candidate for significantly influencing the levels of obesity. Though relatively new in European and British studies the research of the built environment has been underway and growing in North America for several years now (Crawford et al. 2011).

(14)

13 Whilst having addressed many areas this paper is concerned with one; the built environment. Many factors come into play when addressing childhood obesity, this approach is informed by the concept of the human social environment which makes it possible to approach obesity as a phenomenon informed and shaped by various factors (Barnet and Casper. 2011). Largely under-investigated in European literature the built environment is one of the areas that can be directly shaped by policy in a way that affects a person’s physical movement (Galvez et al. 2011).The built environment on its own is no cure, only in conjunction with various other areas some of which have already been mentioned can a solution be found and implemented. To look in detail at the built environment allows for a large factor to be investigated in the human environment and obesity, allowing for more understanding in such a complex area (Brisbon et al. 2005).

In recent years the impact of the built environment for obesity has become more recognised in the UK with the Commission for Architecture and the built environment being highly critical of modern urban design and calling for health to be taken into consideration in the planning of new projects (CABE.2005). As obesity and its associated conditions have risen in incidence the built environment of the UK has undergone significant alterations. From the early twentieth century until the pre WW2 era much of the neighbourhoods in the UK could be characterised as ‘mixed use’, with places of employment, commerce and homes being centred in proximity to one another within communities (Townshend and Lake. 2009). Current town and city planning has placed a focus on creating areas where car ownership is taken as a given and seen as a sole means of transportation.

Recently as the built environment has gained more attention from policymakers and researchers alike, the UK is beginning to acknowledge the impact of the built environment in not only obesity but health in general (Wakefield. 2004). The impact of the built environment on levels of childhood obesity is a particularly important area when considering the risks that children can be exposed to through their physical and social settings (Rahman, Cushing and Jackson. 2011).In the following section the different aspects of the built environment and approaches to researching them will be examined both in the British context and internationally.

The Built Environment

The built environment has been researched extensively in North America and Australia, and whilst the term may seem simplistic it is important to provide an adequate definition for such a broad term. Researchers have approached the concept in different ways with some such as Wakefield (2004) considering it in a purely physical sense;

‘The built environment includes all aspects of the environment that are modified by humans, including homes, schools, workplaces, parks, industrial areas, and highways.’ (Pg2)

Others have chosen to approach the built environment from a much broader perspective choosing to examine the contextual settings such as social networks that can interact with the built

environment to impact upon childhood obesity. Galvez et al (2011) chose to examine the built environment and childhood obesity as part of the Ecological Systems Theory, treating it as one of a series of interrelated factors in a similar way as to the Foresight Review as the graphic below represents;

(15)

14 The scope of the built environment is broad; both conceptually and in the ways it has been

researched. Whilst some have chosen to focus on a purely functional approach i.e. pavement access other researchers have chosen to look more at the effect of this environment or how residents perceptions alter behaviour (Doyle et al. 2006). A term that has gained attention in recent years is that of an ‘obesogenic’ environment, one that actively contributes to higher levels of obesity (Townshend and Lake. 2009). In seeking to give a definition to the obesogenic environment Swinburn and Egger (2002) regards it as;

‘the sum of influences, opportunities, or conditions of life have on Promoting obesity in individuals or populations’

As the field has advanced however the concept of an obesogenic environment has not been fully developed, various researchers have found conflicting findings when trying to either establish or discredit the presence of obesogenic environments (Lake and Townshend 2006). Many studies in an attempt to prove causation have failed to adequately provide evidence for such a relation rather trying to establish a weak form of causality. This study will not attempt to provide causality, instead viewing the built environment as a factor either negatively or positively influencing levels of

childhood obesity. To attempt to say that the built environment is a single cause would be

misguided, however to dismiss the significant impact that it can also play would be so also (Berke et al. 2007).

What this approach to the built environment shows is a move away from individual factors such as motivation to lose weight and knowledge that has only a minor effect, the built environment can have a far greater influence and in conjunction with minor research areas it is possible to unveil significant influences on levels of obesity (Public Health England. 2011). Formulating an objective

(16)

15 and replicable approach to the built environment has resulted in a series of alternative approaches from researchers in recent years, two strands of research have emerged; one approaching the food environment and the other looking at how the built environment either promotes or limits physical activity (Lake and Townshend. 2006).

Many researches have tried to establish the link between the built environment and obesity through examining a variety of its potential influencers such as diet, food availability, active commuting, walkability and access to physical activity facilities (Spence et al. 2009; Tester and Baker. 2009; Van Dyck et al. 2010; Crawford et al. 2008). Whilst this body of literature provides many interesting findings it also prevents certain problems for a potential researcher. To properly research the built environment it is necessary to do so with a firm theoretical and methodological framework, the current literature offers many options for such approaches. An issue arises in the variety of findings when researchers have investigated certain factors of the built environment. The food environment that researchers have tried to link with levels of obesity is adequately defined by Lake (2006) as; ‘The food environment can include availability and accessibility to food as well as food advertising and marketing’ (Pg. 264)

Including advertising and marketing introduces a series of macro issues that communities themselves may not be able to address or remedy, policy regarding advertising and marketing at least in the UK is set at the state level. Many studies of obesity have focused on smaller scale geographical area where targeted interventions could have an effect in reducing obesity, the mandate of local government in the UK to address issues such as marketing is limited and as such any impact of an intervention at the local level would remain largely hypothetical.

A key area where there has been difficulty in providing consistent results is the role of the food environment and its relation to obesity. A previous study by Public Health England found a correlation between the number of fast food outlets in an area and its socioeconomic status, suggesting that the lower an areas socio economic status was the more fast food outlets would be available. Previous literature investigating an obesogenic environment has talked of ‘food deserts’ where only energy dense food is available within a significant distance and buying healthy choices is actively discouraged by the environment (Hill and Peters. 1998). However few researches have managed to make a significant case for such ‘food deserts’ in all but a few cases, such a hypothesis is hard to provide significant causality for and would need to include consideration for the subjective nature of diet selection (Pearson et al. 2005). Nevertheless the relationship between socio economic deprivation and fast food outlet frequency is not a solid one. Various researches have produced varying and often contradictory results. To properly establish such a connection between the food environment and obesity it is necessary to understand the pathways in which the built environment affects food behaviour and then attempt to establish a link to adiposity (Giskes et al. 2007). Exposure to fast food doesn’t automatically lead to consumption, in regards to the UK Cummins and

MacIntyre (1999) failed to establish a link between neighbourhood food availability and an individual’s diet.

Contradictory findings have been produced in the UK when looking at the link between socio economic status and fast food availability, two studies based in the UK produced dissimilar results. Cummins et al (2005) found that the higher the level of socio economic deprivation of an area the more likely there was to be a McDonald’s restaurant in proximity. A Glasgow based study found no

(17)

16 association between area deprivation and fast food availability (Macintyre et al. 2005), it has been suggested that to explain fast food outlet locations more information is necessary than the socio economic status of an area (Giskes et al. 2007). There can be many factors that affect the locations of fast food outlets, such as area property prices, local business density, transport links and many others (Talin and Koschinsky. 2013). A key methodological shortcoming of some studies that have tried to prove the existence of “food deserts” or the link between deprivation and fast food availability is overlooking other forms of food outlets.

Whilst there may be a density of fast food outlets that doesn’t automatically imply that other healthy alternatives aren’t present, a more complete approach would be to map the locations of supermarkets, small grocery stores and other food shopping options other than fast food (Giskes et al. 2007). Previous research from the United States has claimed that the food environment in some ways can explain the racial and socio economic inequalities present not only in obesity but health in general (Lake and Townshend. 2006). The food environment itself is complex with various factors that can be considered associated, rather than socioeconomic status and distance to travel Pearson et al. (2005) came to the conclusion that gender, age and cultural influence had a far greater impact on dietary decisions.

The food environment whilst of importance needs refinements in methodology by researchers in order to make it a more feasible line of enquiry, current literature often fails to provide conclusive evidence of the link between the food environment and adiposity in both adults and children (Ref Needed). On a wider scale much more could be learned by focusing on under researched parts of the food environment such as food marketing and selection of portion size and content, doing so would provide stronger results than the current pattern of produced correlations (Giskes et al. 2007).

The food environment has been left aside in this research project. Due to methodological constraints finding meaningful relations between the food environment and childhood obesity would require a different line of enquiry so as to avoid producing largely observational data (Lake. 2009). This project is set to largely measure the outcomes of certain community designs in regards to physical

movement i.e. walking. The extent to which a community facilitates walking as a primary means of transport to key areas; in this case schools is of particular interest when considering childhood obesity. Direct links have been made to levels of physical activity and childhood obesity and increasing the levels of physical activity that child performs is a recommendation of both the Foresight (2007) and McKinsey (2014) reports.

Research Objectives

The main objectives of this research are to approach obesity as a disease promoted or facilitated by the environments which we create and maintain. By approaching obesity in such a way there is a potential to identify environmental features which are changeable and could have an impact on childhood obesity rates. A key progression in the field of GIS research has been the development of walkability indices rather than focusing purely on measuring vehicular transport. Whilst vehicular transport plays a role in the practicality of certain journeys, measuring the walkability of an area

(18)

17 allows the researcher to grasp how well suited an area is to physical activity, namely the most common physical activity across ages, sex, ethnicity and income group (Saelens et al. 2003). The various health benefits of walking and its relationship to weight status are well documented (Mackett and Brown. 2011).

The age group which will be the focus of analysis are children currently in year 6 in school, meaning between the ages 10 and 11. Initially the plan was to focus on both ages 10 and 11 and also ages 4 and 5, however, focusing on the lower age was dismissed. Studies have suggested that children will move in and out of weight categories and obesity as they develop, arguably a lower age cohort of ages 4 and 5 is not at the stage of development to predict future behaviours and weight status (Howe et al. 2015). The ages 10 and 11 are at a crucial period of biological, social and cognitive development, at this age also they are at a certain level of independence not available to younger ages (Eccles. 1999). This level of independence results in their relationship with the built

environment having a far greater impact in terms of movement, whether this is for transport or for leisure. The decision to focus on the year 6 age group has allowed this project to be far more

focused in its approach when considering the area characteristics of high and low obesity areas. This concentration allows for a focused approach to the areas in the sample through the use of GIS software which produces rich and meaningful data which impacts on the target age group. When examining the built environment, the study will focus on two concepts; connectivity and proximity. Whilst the characteristics of areas will differ in levels of deprivation, area and population by objectively measuring these two concepts it will be possible to establish if there is a link between the physical features of an electoral ward and its childhood obesity rate. As said, establishing a direct causal link between area features and the obesity rate will not be a feasible result of this research project. Obesity is a complex disease comprised of and influenced by a variety of factors, the goals of this project are to offer some preliminary findings that could influence further research into the relationship between the built environment and obesity and also highlight an influencer on obesity that is largely overlooked in the current body of academic literature (Lake and Townshend. 2009). Although the methodology section will deal with operationalisation in much further detail, here a brief and simple explanation of what is meant by proximity and connectivity will be offered. Proximity refers to the distance between essential (to be discussed and defined later) amenities within an area and residential dwellings. The shorter the distance between these amenities and residential dwellings the more feasible they are as walking destinations. Proximity on its own does not suggest that a destination is walkable; this is why connectivity is also taken into account to assess how well connected residences and destinations are. Connectivity is measured by the amount of street interconnections that are present in an area offering more variety in route choice and also the number of possible routes that are available for reaching a certain destination. Further specifics and the rationale for making these methodological choices are offered later in the paper. A further reason to focus on the built environment is to address the structures that impact on people, rather than approaching the analysis from the perspective of examining persons by their category and relationship to obesity (Visscher et al. 2011). By category it is meant the labels assigned to people based on their race and or social class for example, some analyses of childhood and adulthood obesity have focused heavily on the roles of categories of person, often suggesting that a person’s weight status is due to behaviours seen associated with a certain label; for example

(19)

18 suggesting that a person of lower social class hasn’t the knowledge of healthy eating and living practices or simply chooses to disregard them (Washington. 2011). What this project aims to accomplish is an approach that moves away from a form of “blaming” person’s and instead looks at structures that potentially constrain a child’s ability to live at a healthy weight (Galvez et al. 2011).

Research Design

The following section outlines the design of this research and the modelling of its concepts. In particular walkability and its influencing factors of connectivity and proximity are explained and examined in more detail and why and how this relates to the levels of childhood obesity within the research. Once Walkability has been outlined the focus will turn to practically measuring it and the data used within the project. Later discussed in the paper is the use of ethnography, after the initial results of GIS analysis were collected and analysed it was seen that there were findings which GIS alone would not be able to uncover. The use of ethnography allowed for more aspects of the human social environment to be investigated and a more complete picture to be given which was not entirely possible using the originally planned methods.

Walkability

There is a large body of research that has investigated the link between adult and child obesity and also walkability, as with studies of the food environment a series of differing results have been produced. There is a large body of literature also concerned with producing objective measurements for walkability; however with such a body of literature there is also a wide range of approaches that in turn produce different results. In search for producing objective measurements there has been a variety of datasets, software’s and theoretical approaches incorporated for measurement. This is to be expected, a large range of academic disciplines and public organisations have investigated

walkability with each adding their own interpretation to measurement strategies (Owen et al. 2004). There is good reason to try and measure walkability. Walking is the most common form of physical activity, with relatively no cost and efficient for covering short distances whilst providing a form of exercise it is an ideal area of inquiry for researchers of obesity (Saelens. 1993; Doyle et al. 2006). The use of objective measures to identify the relationship between the built environment and walking are essential, prior studies have placed a heavy emphasis on producing replicable measures of walkability (Owen et al. 2004). This study will aim to do that also by employing practical and easily replicable techniques as well as an easily quantifiable set of results.

(20)

19 As previously stated walkability has been of interest to a variety of academic disciplines and public organisations. Whilst this provides for many models to be adopted it also presents a challenge to the researcher in selecting the appropriate form of measurement for his data, software and own

capabilities. Approaching walkability as a theoretical concept is also challenging, adding the appropriate framework to define the concept will have significant impact on its measurement. Whilst there are many accepted factors of walkability there is a wide range of theoretical definitions of the concept. Another point is that the correct definition for the research methods must be selected; incorporating a working definition must be open to objective investigation through the selected methods (Feng et al. 2010). Regarding walkability as a measurement of neighbourhoods Talen and Koschinsky (2013) state;

‘The walkable neighborhood is a physical phenomenon—a bounded place in a given spatial location with selected material properties.’ (Pg.43)

Whilst this definition could be considered rather simplistic, in essence it captures the key features that make up what is examinable as walkable for the purposes of this study. The key characteristics of walkability and their setting are physical, concrete in their construction and approached as such. To consider walkability is to consider the physical constructions in a given area that either promote or impede walkability, the types of roads, how they can be travelled by pedestrians and how simple making journeys can be. Moreover when discussing walkability it can be assessed that a certain kind of quality is being sought after, whether it be how easily a journey can be made or how pleasant it is for the pedestrian to make a journey are two different approaches to walkability (Brownson et al. 2009).

Of particular interest is the link that has been made between a walkable community and an “active community”, one that is seen to promote physical activity in its users by virtue of how walkable the area is (Doyle et al. 2006). Though approaching walkability as made up of physical components the literature still interprets the measurement of physical features in various different ways. Various writers have written on how types of building, whether placing car parking facilities in certain areas, aesthetic appearance of buildings, man-made green space and road traffic effect walkability (Talen and Koschinsky. 2013). While there are various features that have been approached by walkability writers this research project aims to approach walkability by condensing the possible parameters down to those easily measured and comparable across areas. What is necessary is to tailor

walkability to the target age group, namely nine and ten year olds and constructing a measure that takes into account the way children move. For this purpose walkability is assessed from the basis of pedestrianised travel which is facilitated through the planning and design of road networks. As mentioned previously this measure will be constructed based on two concepts; connectivity and proximity. There are many facets that can make up a walkability measure; the key concept behind this study’s approach is functionality, addressing the degree to which an area is walkable. The following sections will go into more detail about the approach to and measurement of both connectivity and proximity.

(21)

20

Proximity

In the planning of any journey, regardless of the means of transport distance plays an important role. Proximity is a key measure of the distance between inhabitants and a target destination, in this case schools. For a destination to be walkable it must also be within a certain proximity, otherwise other modes of transport such as cars and public transport may be used to make these journeys (Hill and Peters. 1998). If an area is to be considered walkable it must have destinations which are within a walkable distance for its residences. Whilst proximity is easily measured by a form of distance the street network must compliment this proximity by making journeys as direct and easy as possible. Proximity has played an important role in the walkability issue being related to not only distance between destinations but also the variety of destinations that are available (Brownson et al. 2009). By researching the proximity of destinations it is possible also to assess the connectivity of an area, previous geographical methods have relied on drawing a large circumference or buffer around an area in order to identify how much of a neighbourhood is within a certain distance of a location. This measurement though logical can often overlook important factors that impact on the movement of pedestrians, namely street connectivity. For example under previous methods just because a household may fall under a certain radius of a location does not automatically translate into this being the case for a pedestrian wishing to make the journey. Proximity relies largely on the distance between destinations based on what the actual roads allow, for example an 800 metre buffer around an area may not necessarily translate into an 800 metre walk if the roads do not provide a direct access point. When constructing proximity it is essential that the researcher has available and recent road data to properly map out the true distance for a person trying to reach a destination. There is little to be gained from calculating proximity based on a radius which may be totally inaccurate.

Town planning in particular has paid much attention to the role of proximity, mapping the locations and accessibility of facilities such as parks, schools, supermarkets, train stations, bus stops and many others (Berke et al. 2007). As mentioned previously the methods for exploring proximity vary greatly when looking at previous research projects, the term proximity is defined in relation to the point of origin of which the researcher wishes to examine (Cromley and Mc Lafferty. 2011). Proximity can be applied in a variety of situations such as between a point of origin and a destination, between a point of origin and numerous destinations or the average distance between a destination and numerous destinations (Brownson et al. 2009). This research however uses the measure of proximity between numerous points of origin and the wider community in which they are situated. A previous approach was used by Curran (2006) to measure the proportion of communities within a defined proximity of certain facilities. By modifying the locations used (in this case schools) there is a possibility to estimate the proportion of the child population which are within walking distance of a school.

In this study the aim is to measure the proportion of the community that are within 800 metres of a school. While an exact number of children that are within that boundary will not be measurable by assessing the percent of households that fall within 800 metres of a school it is possible to get a good picture of how many children are within walking distance of a school. The selection of 800 metres was made based on previous literature and the age group that was to be considered, 800 metres acts as a good measure of walkability as it is of moderate distance though not too great for

(22)

21 repeated return journeys (Owen et al. 2004). Previous studies have found that a destination within one kilometre of residences has meant that they are more likely to be journeys that are walked rather than using vehicular transportation (Vargo et al. 2011). The proportion of an area within 800 metres of a school will be measured by using the ArcGIS function network service area, the specifics of this analysis will be discussed in more detail in the methodology section.

Connectivity

Connectivity is of great importance when trying to assess not only walkability but also other means of transport such as bicycle, car, bus and or metro. Much more literature has been produced focusing on connectivity than it has proximity; as such there are various ways of measuring

connectivity. Whilst street connectivity is the extent to which streets are linked or connected making travelling easier, the ways in which to measure this connectivity vary greatly. Various studies have based connectivity by using buffer analysis, whilst this is effective as mentioned previously without sufficient road data these measures are largely irrelevant (Cromley and Mclafferty. 2011). Whilst this research will use buffer analysis alongside the necessary road data this will not, on its own provide a measure that is easily comparable across areas. To achieve this more refined approaches are necessary, to provide a quantifiable measure of density which is accurate but also comparable is the end goal. To do this a variety of approaches are available, having reviewed the current literature on walkability a series of measurement techniques have emerged.

To decide on what measurements to use it was necessary to identify the availability of data, the capabilities of the software and the validity of the measurements. Connectivity is a simple concept, it is inexplicably linked with proximity and one needs the other in order to produce reliable

measurements. To decide on which methods to use key techniques from walkability literature were selected and reviewed. From reviewing the literature the following methods were selected; Internal Street Connectivity Index

Roadway Connectivity Index Gamma Index

Street Intersection Density

The following measures with the exception of the internal street connectivity index have emerged frequently in walkability literature. The Internal Street Connectivity Index is a relatively new method which allows for an easy numerical scoring of connectivity based on junctions and dead ends in an areas street network (Dill. 2004). The internal street connectivity index produces a score between 0 and 1, the maximum score of 1 indicates that an area is very well connected and has few dead ends whereas 0 indicates the opposite and an area is very poorly connected. The guidance for newly built neighbourhoods suggests that a score between 0.5 and 0.7 is the recommended range of

connectivity, less than five is seen as poorly connected and higher than seven is seen as ideal (Criterion Planners Engineers. 2001). A positive of this method is that it can be compared across areas of different size, the target areas in question range from 0.6 square miles to over 9 square

(23)

22 miles. With such a wide range of area sizes it is important to use a measurement method that can provide a comparable set of results across geographical locations.

The roadway connectivity index has been used frequently in the United States at the county and regional level to measure new development projects. The measurement is based on the concept of links and node. Essentially a node acts as part of a circuit, in this case a junction whilst the links are the lines that connect one node to another as can be seen below;

The measure is of the ratio of links to nodes over a radius of half a mile. Whilst providing a sound numerical score this measure is flawed for the purposes of this study in that it is aimed at measuring homogenous areas. The research areas that the roadway connectivity index is often associated with are similar in design and structure making comparison across areas relatively easy. The study areas in this project however differ widely in size making a comparison of only a half a mile radius unsuitable given the variations in density and population.

The Gamma index is a far more complex model developed by geographers in order to measure connectivity. Approaching connectivity from the view of a circuit the index also uses links and nodes as a form of measurement, in a similar way to roadway connectivity index by taking into account the ratio of links against the maximum possible number of links to nodes. Measured between one and zero, the score given is of a percentage of connectivity i.e. 0.7 translates to 70% connectivity. This was also dismissed as it lacked the necessary comparability but also required a larger and more varied amount of data than what is currently available for the UK in the public domain. Whilst an effective measurement the gamma index would require far more analysis and data than what is available or necessary, with only a few areas to analyse the gamma index is more suitable for a much larger number of areas.

A final consideration was measuring connectivity based on intersection density. Intersection density was by the far the most straightforward measure of connectivity that was encountered when reviewing the literature. Similar to measuring population density, intersection density involves calculating the number of intersections in an area and calculating the number of intersections in

(24)

23 relation to the size of the target area. Whilst a higher number of intersections suggest higher levels of connectivity, the established link in the literature is questionable. While some studies have tried to establish causation between intersection density and connectivity, by on large the measurements are based on correlations that don’t necessarily equate to higher levels of connectivity (Curran et al. 2006). A further problem with this method arises again in the variety of area sizes. A measure of density may show negative connectivity results for a smaller area that is still well connected and vice versa. Based on the highlighted shortcomings intersection density was eliminated as a method of measuring connectivity in this study.

After reviewing the various methods and taking into consideration the size of the study, the

availability of data and the required scoring of connectivity it was decided to use the Internal Street Connectivity Index. The Internal Street Connectivity Index offers versatility across different sized areas and allows for a scoring system that is transferrable across areas. This method though relatively new has been used successfully in previous studies of walkability and connectivity and provides credible measurements of connectivity. Whilst the other methods all had various proven strengths, they would not have provided the fit to the project that was necessary, having been used largely for similar sized areas it would have been poor practice to try and imitate these measures and would have provided scoring irrelevant for the sample size of this study.

When assessing what methods to use it was important to consider the type of data that was used in previous studies. If the methods required data that were not available in the public domain then they were dismissed as feasible. In the interest of replicability and validity all the data is available in the UK from the public domain and can be accessed by anyone free of charge. It was found that the data available was more than sufficient for the level and scope of analysis, being standardised, available and consistent throughout geographical locations in the UK.

Methods

This section deals with the practical applications of investigating the walkability of the target areas. Though not discussed in this section the researchers own observations were employed in a small ethnography to try and uncover the reasons for disparities which emerged between two areas as a result of the GIS analysis (Loucaides. 2004. Hammersley and Atkinson. 2007). Investigating

walkability was done through the use of Geographic Information System (GIS) software ARCGIS 10 available from the company ESRI. The use of GIS has been previously mentioned and is well documented in the literature, the software ARCGIS 10 is one of the latest versions available from ESRI which has been producing GIS software used by governments, public organisations, universities and public organisations for many years. The latest version was made available through the

Geography department at the UVA on a student trial license for personal use and also on every computer at the UVA.

The GIS software has the capability to perform numerous complex analyses, given the appropriate dataset GIS is a powerful tool for researching the built environment. It is not without its faults however, simple operations can be very time consuming and is not an easily scripted software meaning that the most mundane of tasks can take a series of repetitive steps. One of the most time

(25)

24 consuming processes is what is called clipping, in essence taking a large piece of data and limiting it to a certain scale for closer analysis. The vector data for each electoral ward is allocated to a certain geographical level or a square for each region of the UK and assigned a code; these folders contain a mass of geographical information which can be seen in a number of areas representing roads, houses, important buildings, rivers, railway stations, elevation features and various other features. Extracting the relevant vector data for each ward from a series of regional coverage files is time consuming but also necessary to detail the analysis to the low level that is required for making detailed geographical analyses. The ArcGIS software works best when dealing with small scale sets of data making for easier processing and analysis, the analysis required numerous data types and actions to perform analysis. The following data was used to perform the analysis;

Polyline Road Shapefiles Building Footprints

Important Building Footprints Conversion of footprint to point data Electoral Ward Boundary line data Ordnance Survey Street Raster Base map

There were a variety of technical actions that had to be performed before the data would be ready to be analysed across wards. What was first necessary was to place the electoral ward boundary lines on the Raster base map, doing so allows for the identification of area by code rather than trying finding the boundary on the map. Once the electoral ward has been located it is then easy to save it as its own file on the system and limit analysis purely to one area, doing so greatly decreases the processing time for the software and makes functions such as clipping and feature identification much simpler. Once a ward was located the next step was to add the data from the region that it belonged to, from this large regional dataset the data could be limited to only the electoral ward in question. What is required from each regional dataset is building footprints for each structure within an area, important building footprints that identify features such as education facilities and hospitals and finally line data for roads to model connectivity. An important piece of preparation for analysis was converting building footprints for schools into point data on the ArcGIS map, doing so requires the creation of a new feature dataset and can be achieved either interactively through adding points to the map or by entering the relevant coordinates. The most efficient way of doing this was by using ArcGIS interactive features to add a point to the map, the use of coordinate systems for such a small handful of points would have been time consuming and quite complex for the action that was required.

The conversion of schools to points makes analysis much easier, ArcGIS works better with point data for analysis of locations and accessibility, as will be discussed the network analysis function in ArcGIS requires locations to be in point form so as to properly calculate travel along the roads. To use set locations as objects of proximity and connectivity analysis ArcGIS works best when roads are represented as lines to model movement, when ArcGIS works with road lines it can model travel