Understanding decision-making associated with children being driven to school and moving toward active school travel. A theory-informed qualitative study

(1)

! ! ! ! !

Understanding!decision.making!associated!with!children!

being!driven!to!school!and!moving!toward!Active!School!

Travel!

A"theory)informed"qualitative"study"

"

MSc!PLANET!Europe!

Chang!Xu!

June!2017!

! ! ! Student!Number! Cardiff"University " C1570526" Radboud"University"Nijmegen " S4645154" ! Supervisors:! ! Dr."Justin"Spinney" Dr."Karel"Martens" " " " ! ! ! ! ! ! !

(2)

Abstract

Over the last decades, concerns about the rise in childhood obesity and the prevalence of car use for school journeys have prompted children’s active school travel (AST) became a hotly debated academic, public and policy issue surrounding childhood health, mobility, and environmental sustainability. The mainly underlying reason is the significant implications of AST not only for children but also for the whole community and the environment. Under this context, a great attention has been shifted to school travel behavior research around the world, associated with an ultimate goal: finding ways to guide changes in school travel behavior so that more and more children can engage into AST. It is also the aim of this study. Therefore, the study provides insight into decision-making process around driving children to/from the school, by looking at the case of Cardiff (the UK), in an effort to inform recommendations for AST promotion.

A qualitative approach is chosen in this study, referring to both the Theoretical Domains Framework and Three-Elements Model of Social Practices Theory, from both an individual's and social practices' perspectives to explore the driving decision for school journeys. Based on 20 qualitative interviews, this study reaffirms that there is a complex interplay of factors influencing about parents’ driving decision within the decision-making process. Those factors, incorporating individual and household, physical environment, and social environment levels, determine car chauffeuring as the most convenience, safety, and comfortable way for school journeys. Most importantly, this study also finds that current parents’ decision about school travel transports is highly related to a broader society, rather than a solely personal choice. The escalating “norms” and “meanings” surrounding convenience, safety and comfort have made driving, rather than AST, as an intrinsic requirement and necessity in the current society. Even, those socially shared meanings play as internal dynamics to maintain and stabilize this kind of society practices, i.e., driving, instead of active school commuting. Hence, this study provides an alternative way to consider AST promotion and highlights the needs in a whole-system solution at a society level, rather than the prevailing individualist approaches aiming at individuals’ behavior modification.

(3)

Acknowledgements

Firstly I would like to express my thanks to my supervisors Dr. Justin Spinney and Dr. Karel Martens who helped me in every step of the way and encouraged me to deal with the challenges encountered in the research process. Without their valuable input to my study, this thesis would be impossible for me as it is the first time for me to conduct a social research, particularly in a foreign country, the UK. This piece of work finally becomes a quite beautiful ending for my master study.

I would also express my gratitude to all the participants of this study who not only gave their time to provide rich information about this topic but also offered me with strong motivation and encouragement to accomplish this work.

Thanks to the Erasmus Mundus Programme that provides me a precious scholarship for an overseas study. It can be an unforgettable and valuable study experience in my life, which also influences my future life.

I would also like to express my appreciation to all the members and teaching staff of the Master of Science degree in European Spatial Planning & Environment Policy at Cardiff University, the Master of Science degree in Urban and Regional Planning at Radboud University, and PLANET Europe Programme for their diligent contributions, works and knowledge they shared with me, as well as their friendly help in the whole learning process.

Finally, I will never forget my classmates in the joint master programme and my friends who supported and encouraged me during all the time of my learning overseas and express my love to my family, especially my mother for their long time supports and understanding.

(4)

4 Results and analysis"..."40 4.1 School Travel Decision"..."40 4.1.1 Parental role and responsibility in the school trips"..."40 4.1.2 Habitual nature of current school trips"..."42 4.1.3 Re-consideration during primary-secondary transition"..."43 4.1.4 Changes in school travel over time"..."45 4.2 Parents' perspectives on driving decision and factor “negotiation” process"..."46 4.2.1 Convenience"..."47 4.2.2 Safety"..."49 4.2.3 Comfort"..."51 4.3 Parents’ perspectives on AST"..."54 4.3.1 Inconvenient"..."54 4.3.2 Unsafe"..."57 4.3.3 Uncomfortable"..."62

5 Finding and discussion"..."65

6 Conclusion and recommendation"..."73 6.1 Summary of the findings and policy implications"..."73 6.2 Contribution and consideration for further research"..."77

(6)

Appendices

Appendix 1 AST in the policy context

Appendix 2 Designed Interview questions according to the TDF

Appendix 3 Main interview questions and possible prompt according to topics Appendix 4 Sample Demographics Summary

Appendix 5 Results of thematic coding with quotes

List of Figures

Figure 1 Definitions of the theoretical domains Figure 2 Three-Elements Model

Figure 3 Social-constructed school travel decision framework

List of Tables

Table 1 Percentage of children obese and overweight (including obese) in the UK Table 2 Characteristics of participants interviewed (n = 20)

Table 3 Results Summary Influential Factors for Driving Table 4 Results summary Perceived barriers and facilitators

(7)

(8)

CHAPTER 1: INTRODUCTION

The aim of this chapter is to provide a foundation for the rationale of this study. First, this chapter provides an overview of research background within the current UK and Wales context. The research aims, objectives and approach are summarized later. The structure of the thesis is outlined at the end of this chapter.

1.1 Background

1.1.1 Prevalence of childhood overweight1 and obesity2

Since 1980, the number of overweight and obesity children has increased at an alarming rate worldwide (Public Health England, 2016). The rising prevalence of overweight and obesity among childhood period has also been described by the World Health Organisation (WHO) as one of the most serious global public health challenges of the 21st century. According to the latest data published in The Lancet (studying in 34 Organisation for Economic Cooperation and Development (OECD) countries, Wang et al., 2016), in 2013, 23.8% (95% uncertainty intervals 22.9 to 24.7) of boys and 22.6% (21.7 to 23.6) of girls aged 2 to 19 years in those OECD countries were estimated to be overweight or obese.

Table 1 Percentage of children obese and overweight (including obese) in the UK (source from: Public Health England, 2016)

In the UK, this prevalence is also common and ranked 9th out of the 34 OECD countries for prevalence of overweight (including obesity) for boys and girls combined. Compared with other countries of the UK, Wales has been found that has the highest prevalence of childhood obesity in the UK and is one of the worst in Europe. About 34% of children in Wales were classified as being overweight or obese, """"""""""""""""""""""""""""""""""""""""""""""""""""""""

1_"_{Overweight = one standard deviation body mass index for age and sex}

(9)

including 19% as obese (Jones et al., 2013; Public Health England, 2016). In addition, it has been forecasted by the UK Government’s Foresight Programme that the current tendencies would continue and increase to 25% of obese by 2050, unless action is taken now (Welsh Government, 2015).

1.1.2 Health and economic impact of childhood overweight and obesity

Therefore, as the prevalence and severity of childhood obesity increase, on the back of this issue, there’s a deep concern related to both short-term and long-term health risks since childhood obese often leads to obesity in adulthood. Therefore, in the short term, childhood obesity can affect the ability of those children to participate in everyday activities and then impair their wellbeing, while longer-term health problems include the increased risk of coronary heart disease and strokes in their later life. Besides, a growing body of scientific evidence indicated that many chronic diseases, such as coronary heart disease, type II diabetes and osteoporosis, for which obesity and physical inactivity are risk factors, can also be suffered at a younger age (Buttitta et al., 2014; Cole, 2000; Daniels, 2006; O’Sullivan, 2013).

In addition, overweight and obesity can lead not only to health-related complications, but also easily result a serious psychological impact. For example, overweight children and adolescents often suffer from symptoms of depression and anxiety, poor self-esteem, and social stigma, which may increase their vulnerability to mental health problem (Buttitta et al., 2014; Daniels, 2006; Puhl & Heuer, 2009).

The economic consequences of childhood overweight and obesity are also significant. Reports in Wales estimated between £1.4 million and £1.65 million every week of cost for treating diseases resulting from obesity in Wales alone (Welsh Government, 2015). For example, figures from the Child Measurement Programme for Wales Report 2011/12 categorized 12.5% of Welsh children aged 4-5 years as obese. Based on the GLA costings of £31 per obese child per year, this would equate to a cost of £114,390 for that single age group in a single year (Welsh Government, 2015). In the UK wide, along with the increasing of obesity rates, National Health Service (NHS) costs attributable to overweight and obesity are forecasted to double to £9.7 billion per year by 2050. Despite the direct healthcare costs, problems associated with

(10)

obesity can also impact on social services, housing, the public sector workforce and emergency services, which will increase extra financial burden.

1.1.3 Current trends of journey to school and the loss of children’s independent mobility

Over the past four decades, the rise in childhood obesity has occurred at the same time children has undergone a significant change how they get to school in all countries (Jeekel, 2016). There has been a noticeable shift to children being driven to and from school, and away from active modes (walking and cycling) (Easton & Ferrari, 2015; Jeekel, 2016; McDonald, 2008; Mitra, 2013; Su et al., 2013). In the UK, the big shift from walk to the car for the school journey was between 1985/1986 and 1995/1997 when the levels doubled (from 22% to 40% for primary children and from 10% to 20% for secondary children) (Mackett, 2013). To a large degree, these trends can be explained by the significant increase in car ownership since 1985/86 (ibid.). As showed by National Travel Survey (Department for Transport, 2015a), the proportion of households without a car had declined from 38% in 1985/86 to 30% by 1995/1997 while the proportion of households with more than one car increased over the same period, from 17% to 26%. Similarly, a growing of study has also confirmed the negative and significant correlations between car availability and active travel to school (Van Goeverden & Boer, 2013).

Since 1995/1997, the proportion of children aged 5 to 10 being driven to school has risen steadily from 40% in 1995/1997 to 48% in 2015 whereas during the same period, the number of trips by foot for these children has reduced from 53% to 44% (Department for Transport, 2015b). Among children aged 11 to 16, in 1995/1997, 41% of trips to school were made on foot and 21% were by car; by 2015, the number of trips by foot and by car had changed to 37% and 26% respectively (ibid.). In terms of cycling, compared to some countries, such as Denmark, Finland or the Netherlands where about 50% pupils travelled to school by bike, journeys to school by bicycle for both primary and secondary British schoolchildren have been a steady rate, swinging just between 1% to 2% during the past two decades (NTS, 2015; Goeverden & Boer, 2013; Nielsen et al., 2013).

(11)

This trend has also been found in Wales. The use of car for travelling to school by primary school children and by secondary school children has reached to 51% and 27% respectively in 2014/15 (Statistical Bulletin, 2015). Even for journeys of less than half a mile in length, 30% of primary children are regularly taken by car and 56% are for 0.5 to 1 mile trip. For longer school trips (1-3 miles), the prevalent mode of transport for secondary school children is also by car, accounting for 41% in 2014/15 (ibid.). Additionally, along with the increasing in car use for children’s school journey, it has also reflected a corresponding decline in independent mobility for them in the form of walking and cycling during the past decades. There is a wide range of studies suggesting that child instigated “free” play has fallen compared to this in the past, including unsupervised travelling to and from school (Fyhri & Hjorthol, 2009; Mitra, 2013; Page et al., 2010; Shaw et al., 2012).

Data from the NTS (Department for Transport, 2015c) in the UK provided some insight into levels of usually accompanied school journeys of children aged 7-13 years between 2003 and 2015. According to the data, during this period there was an increasing from 61% to 67% in the proportion of accompanied school journeys. For 7-10 year-old children, the accompanied journeys even reached to 90% in 2015.

1.1.4 Children’s health and development impact

In this regard, as for children, it has been often argued that the major consequence brought by the increasing of car use and the decline in children’s independent mobility (CIM) is the greater risk for getting physical and mental development problems, ranging from obesity to reduced development of cognitive skills.

On the one hand, there is considerable evidence that lack of physical activity can lead to a number of adverse health conditions and obesity is one of those health risks, as discussed in the former part. Similarly, travelling to and from school, as the normal and common everyday events for children, provides then a good opportunity to fulfilling their daily requirement for physical activity. Therefore, shifting from active commuting by walking to passive commuting by car, to a large degree, cuts down the level of daily physical activity for children, which replace moderate intensity physical activity with sedentary behavior, thus leading an unhealthy lifestyle. In particular,

(12)

given childhood or adolescence, as the key period of individual’s development, the unhealthy lifestyle developed in this period is likely to be maintained in their adulthood.

On the other hand, related to CIM in school journeys, it is the claim that children who are always driven to school or activities have less opportunity to experience their neighborhoods or the broader local environment, which poses a huge risk for not only their physical health, but also mental and cognitive development (Fyhri et al., 2011). As argued by Brown et al. (2008), independent mobility is always associated with children’s level of physical health, the development of motor skills, as well as their cognitive development, by helping them to learn mapping and way-finding abilities, which is also noted by Page et al. (2010). Additionally, Brown et al. (2008) also noted freedom of children movement is of importance in helping children build relations with others beyond their immediate family, which play a significant role in their emotional development. Therefore, it is not hard to find that the childhood development consequences, caused by the car-dependent and the loss of CIM, are multifaceted and unexpected.

1.1.5 Traffic, social, and environmental impact

In addition to health and development impacts, it has often been claimed that the mass use of the car for school journeys is bound to have a serious of undesirable consequences, not only for individuals, but also for community and environment. First of all, the increase of motorized school journey affects public transport patterns, reduce the street activity, cause localized congestion around schools and contributes to the sharp road traffic peak around eight o’clock each morning, which no doubt brings inconvenience for the whole community. In particular, it increases the traffic danger for pedestrians and cyclists, especially for children who are more likely to be involved in traffic accidents. As explained by Swedish researcher Stina Sandels (cited by Johansson, 2015), given the mental and cognitive development stage within childhood, children under 12 is not ready to deal with real traffic conditions. According to the data from NTS (Department for Transport, 2015d), although the number of British children killed or seriously injured in traffic accidents has been a

(13)

steady decrease in the last five years, the proportion of children still accounted for 4.9% and 1.5% in any road accident and injury accident respectively in 2015.

Similarly, the “traffic danger” has often been reported by British parents as the main reason for their children’ accompanied school journeys and car usage (Department for Transport, 2015c). Ironically, the prevalence of car use for school journeys brings a vicious cycle (Carver et al., 2013; Fyhri & Hjorthol, 2009; Goeverden & Boer, 2013). As children are increasingly driven by their parents, traffic volume and traffic danger increase, making the walking environment less safe for others and promotion on walking and cycling more difficult. Consequently, fewer parents allow their children to travel to school by active or unsupervised modes due to the concern of “traffic danger” and the vicious cycle is thus created.

Despite the traffic impact associated with the mass use of the car for school journeys, it is also worth noting that this trend brings a severe environment impact, e.g. climate change and air pollution. It is widely known that transport is a major source of air pollution in the urban areas of the UK, as well as many European countries (Defra, 2013). In the UK, it is estimated that road transport contributes 20–30% of national emissions of air pollutants (RAC Foundation, 2014). Particularly, air pollution can cause damage to biodiversity and has a range of effects on health. Hence, air pollution is also regarded as world’s largest single environmental health risk. In a recent estimate, WHO reported that in 2012 around 7 million people died, about one in eight of total global deaths, are caused by air pollution exposure. As for school-age children who are particularly vulnerable to traffic related air pollution, the health impact has also be demonstrated in the study of Pooley et al. (2010). Pooley et al. (2010) stated that those children who traveled to and from school by foot, by bike, and by bus were more likely to be influenced given the increased exposure to traffic fumes (NO2, PM10).

1.2 The problem and rationales for the study

Over the last decades, concerns about the above currents—the rise in childhood obesity and the prevalence of car use for school journeys has prompted children’s AST became a hotly debated academic, public and policy issue surrounding

(14)

childhood health, mobility, and environmental sustainability. The mainly underlying reason is the significant implications of AST not only for children but also for the community and the environment. On the one hand, AST is widely regarded a way to enhance children’s physical and mental health by increasing their habitual and routine physical activity and by promoting their psychosocial skills. Related to health benefits, “independent mobility”, as a further component of children’s active transport, is also considered to be beneficial for children to develop their psychosocial skills, including mapping and way-finding abilities, as well as the relations with others beyond their immediate family (Carver et al., 2013). On the other hand, the promotion of active transportation, as an ideal low-cost strategy, is always proposed to realize the overall sustainability goals for the whole society through minimizing environment, social and economic impact caused by ever-increasing motor vehicles. Under this context, promoting AST has thus received more attention from policy makers and planners around the world (Carver et al., 2013; Meester et al., 2014; Lang et al., 2011; McMillan, 2007).

Despite the well-documented benefits of AST and the stress in many policy documents, unfortunately, many parents habitually drive their children to/from school, rather than use other active modes. The problem is thus: how more effective AST promotion can be achieved in the practices. This socio-political problem, identified above, becomes the first rationales for this study.

In addition, this study also has a scientific relevance. Although a considerable amount of research in recent years has conducted in this field, there are gaps in current school travel research, identified more in detail in Chapter 2, which limit the understanding of effective ways to promote AST practices. These can be summarized into three aspects:

(1) The gap lies in context-specific research. Although a range of empirical studies has already reported a wide variety of findings, it has often been argued that there are distinctive inconsistencies across studies due to study design limitation, measurement bias and cross-country variations (Potoglou & Arslangulova, 2017). In this regard, the difficulties in the generalization of the findings have been concerned especially when there are existing differences, e.g. culture, natural environment, policy environment,

(15)

etc. across countries. Therefore, the need for further context-specific research is highlighted by many researchers (Cavill & Watkins, 2007). Particularly, a majority of AST research is drawn from the USA and Australia (Sallis et al., 2004; Pont et al., 2009), while very little research concentrated on Wales (the UK) case where with a high prevalence of childhood obesity and car use for school journeys.

(2) The gap exists in method limitations. The majority of existing research has employed a cross-sectional, survey methodology. Although such a quantitative approach is an important way in identifying the correlates associated with a given transport mode, while it doesn’t lend itself to explore the dynamic nature of decision-making processes and the interrelationship between those factors. From this sense, it is hard to explain how these factors are experienced or identify what barriers are perceived for AST, thus limiting the role of those results in guiding behavior changes and informing an intervention for AST. (Abrahamse & Steg, 2009; Easton & Ferrari, 2015; Mandic et al., 2015; Mitra & Buliung, 2012, 2015; Potoglou & Arslangulova, 2017). On the contrary, qualitative approaches may be particularly helpful for investigating school travel behavior and unraveling the complexities of this kind of behavior, while very little qualitative research have been published. (3) The third gap is the lack of theory-informed research. It is widely acknowledged that behavior change needs a strong theoretical basis and evidence from a serious of systematic-review and meta-analysis in active transportation interventions has also suggested that interventions informed by theoretical basis have greater efficacy than non-theory based studies (Arnott et al., 2014; P Chillón & Evenson, 2011; Moghtaderi, Burke, & Troelsen, 2015). However, for the moment, there is a lack in exploring school travel behavior based on a behavioral theory and aim at informing a theory-based AST recommendation.

In order to fill the gaps discussed above and deal with the practical problems, this thesis seek to conduct a theory-informed qualitative investigation into the decision-making around children being driven to/from school in the Wales context, in an effort to determine the barriers and facilitators for AST engagement, and to propose recommendations for AST promotion.

(16)

1.3 Aims, research questions and approach

Against the above background, the aim of this thesis is to explore the parental

decision-making process about driving their children to/from the school, by looking at the transition period from primary to secondary school for the case of Cardiff (the UK), in an effort to inform recommendations for AST promotion.

Based on the research aim, the research questions of the research are as follows: 1. What are the factors associated with parental decision-making in driving their

children (12-15 years) to/from the school after primary-secondary transition? 2. How do these factors experienced by those parents within the school travel

decision-making process?

3. What are perceived barriers and facilitators for them to engage into the AST? 4. What recommendation can be given to increase AST in Cardiff?

1.4 Target population and the selection of primary-secondary transition

Although some study shows that child characteristics and attitudes can be a variables affecting the decision-making process on mode choice (Easton & Ferrari, 2015; Guell et al., 2012; Panter & Jones, 2008), a strong parental influence on the final transport mode choice up to a certain age has always been reported previously (Carver et al., 2013; Hopkins & Mandic, 2017; Johansson, 2006; McDonald & Aalborg, 2009). Therefore, in order to fully investigate the decision-making around children being driven to/from the school, this study particularly focuses on the perceptions and experiences of parents. In view of the target population, parents with children between the ages of 12-15 (after the school transition of primary to secondary) is chosen in this study for several reasons.

Firstly, it is because of the habit discontinuity hypothesis, described by Verplanken et al., (2008), positing that behavior change is more likely to be happened when habits are broken by a change of context. More specifically, when the context in which a behavior is habitually performed is subject to change, the environmental cues that automate behavior are no longer present and there is an increased likelihood that

(17)

individuals will re-consider their behavior and adopt alternatives (e.g., walking or cycling to school). As for the case of school travel behavior, it can also be considered under habitual control, at least, to some extent, because journeys to school are characterized by repetition. Evidence from a recent study in evaluation the habits role in school travel decision has also confirmed this point (Murtagh et al., 2012). Hence, it is valuable to study at primary-secondary transition and research the 12-15 age group student’ school travel behavior as the just change of school provides an opportunity to break the previously habitual behavior. As stated by Murtagh et al., (2012), interventions delivered shortly before, during or after a context change have been shown to be effective at changing habitual behavior.

Besides, this age of transition from childhood into adolescence is characterized by a downward shift in physical activity and an increase in sedentary behavior. Evidence showed that fewer children achieve the public health recommendations of 60 minutes moderate- to vigorous-intensity physical activity per day when they enter adolescence. Given the fact that active travel to school has been associated with higher daily levels of physical activity and higher cardiorespiratory fitness in children and adolescents, the promotion of active travel among those age group is of significance for them to maintain physical activity and establish lifelong health habits.

Finally, transition from childhood (by the age of 12) into early adolescence (12-15 years) is also a key time when both independent mobility and capacity (children's ability to move around in their neighborhood and handle the traffic condition alone) changes (Page et al., 2010). Hence, 12 can be a threshold for AST promotion. As explained by Swedish researcher Stina Sandels (cited by Johansson, 2015), given the mental and cognitive development stage within childhood, children under 12 is not ready to deal with real traffic conditions. The enhancing of capacity and independent during this period has reinforced the significance of promoting AST among this age group.

1.5 Study area

In this thesis, the city of Cardiff is selected as a study area for both methodological and pragmatic reasons. In terms of methodological consideration, the phenomenon of

(18)

Cardiff fits the research issue being investigated in the study: yet, the prevalence of car driving for school journey is evident, accounting 41% of primary school students and 31% of secondary school students. Pragmatically, the city of Cardiff is selected as a study area firstly due to its increasingly significant role in the Wales development. Cardiff, as the capital city of Wales, is now dynamic, fast growing and forecast to have a higher population growth rate over the next 20 years than any other city in the UK. In particular, along with the establishment of Cardiff Capital Region (CCR) and the gradual investment of CCR city deal, Cardiff, as the core city, has been endowed many duties and expectation to improve productivity, drive innovation and support job growth throughout south-east Wales over the following decades. It can be said that Cardiff is on the edge of becoming both a regional power within the UK and a global capital city. However, for the moment, Cardiff has its own transport challenges that may militate against its potential growth and development. One of the most major challenges of transportation is traffic congestion due to the heavy car dependence, ineffectiveness of the existing public transport and less attractiveness of active travel among the city. As expressed in Cardiff Local Development Plan 2006-2026 (Cardiff Council, 2016, p28), the current Cardiff Council attaches great emphasis to increase the use of sustainable travel modes and networks (particularly walking and cycling) in order to establish Cardiff as a sustainable travel city. But for now, at least, very little research related to active travel is drawn from Cardiff cases, let alone concerning the factors affecting active travel to school in Cardiff context. The research in understanding the reasons why heavy car dependence has come about among school-aged children is of importance to inform the further effective policy-making and targeted interventions. Finally, due to the consideration of time constraints and personal experiences, focusing on Cardiff is a particularly feasible strategy for the researcher to gain an access to parents who are driving their children to/from school and obtain sufficient data to fulfill my research aims.

1.6 Thesis structure

In addition to this introduction, this thesis is divided into six chapters reflecting the research process as follows.

(19)

CHAPTER 2: Following this introductory part, chapter two presents a summary of

the literatures, which have been reviewed in the preparation of this thesis. This chapter considers three main areas, namely current evidence in AST, behavior theory, and theoretical framework underpinning the study.

CHAPTER 3: This chapter explains in detail the research strategy and methodology

used to carry out this thesis. It includes the study design, the method for data collection and analysis, the research limitations, and ethical consideration.

CHAPTER 4: This is the empirical chapter and proceeds to provide a discussion of

collected data, informed by the theoretical framework, thus corresponding to the research aim and objectives. Therefore, the parents’ general perspectives on the school travel decision are summarized firstly, followed by a demonstration of factors “negotiation” process underlying driving decision and parents’ perspectives about AST.

CHAPTER 5: Based on the emerging findings, this chapter aims at using the

Three-Elements Model to raise a discussion about how and why parents’ driving decisions, instead of active modes, are socially constructed, in terms of meanings, materials, and competence, materials domains, with an effort to providing implications for AST promotion.

CHAPTER 6: This last chapter of this master arrives at some conclusions set against

the research questions and aim. In addition, it makes a reflection on study’s limitation for developing future work and its contribution to knowledge and practices of this thesis.

(20)

CHAPTER 2: LITERATURE REVIEW 2.1 School travel study current evidence

2.1.1 Influential factors in school travel decision

Previous studies of school travel behavior, as it has evolved over the last decades, have yielded a wide variety of insights into the decision that parents/guardians or children make about their school transportation. In reviewing the broader literature (Meester et al., 2014; Easton & Ferrari, 2015; Henne et al., 2014; Johansson, 2006; McDonald, 2008b; McMillan, 2007; Mitra & Buliung, 2012; Ridgewell et al., 2009; Trapp et al., 2011) and recent reviews review (Chillón & Evenson, 2011; McMillan, 2005; Mitra, 2013; Panter & Jones, 2008; Pont et al., 2009; Potoglou & Arslangulova, 2017; Sirard & Slater, 2008; Stevens, 2010), the majority of existing study focuses on understanding the determinants of children’s travel behaviors, which has often been demonstrated as a product of multiple levels of influence, involving a number of individual, social and environmental factors. Drawn on social-ecological approach3, which has frequently been used to conceptualize school travel behavior (Carver et al., 2013; Curtis et al., 2015; Meester et al., 2014; Li & Zhao, 2015; McMillan, 2005; Mitra & Buliung, 2012, 2015), those correlates of AST can be generally divided into four levels, namely child, household, urban environment, and context level.

Child level. Although many studies theorize school travel behavior largely as a

product of parental decisions, child characteristics, e.g. age, gender, race/ethnicity, and attitudes, have often been tested its relations with the rates of active commuting in many empirical studies. For example, many research shows that boys are more likely than girls to actively travel to school (Evenson et al., 2003; Harten & Olds, 2004; Rosenberg et al., 2006; Schofield & Mummery, 2005;). Also, the apparent variation in school travel behavior across ethnic groups is found by several studies (McDonald, 2008a; Yarlagadda & Srinivasan, 2008). However, regarding individual factors, it is worth noting that there are inconsistencies across studies. An example is that in the same country, e.g. the US and Australia, some studies show that active commuting to """"""""""""""""""""""""""""""""""""""""""""""""""""""""

3_{" Social-ecological approach conceptualizes physical activity as a product of multilevel effects, incorporating}

individual, household, neighborhood, and social-political environment, and emphasize reciprocal interaction between those levels.

(21)

school is more common among older pupils than younger pupils (Evenson et al., 2006; Merom, Tudor- Locke et al., 2006), whereas other studies show the opposite pattern (Blair et al., 2001; Evenson et al., 2003; Schofield & Mummery, 2005). As stated by (Potoglou & Arslangulova, 2017), this inconsistent pattern cross studies maybe due to the availability of directly relevant data or measurement bias.

Additionally, in terms of child’s individual factors in school travel behavior, it is explained by Mitra (2013) that some factors, largely reflecting individual self-efficacy, can have both a direct and particularly indirect affect on the outcomes of travel behavior, e.g. influencing caregiver attitudes or evaluations. In other words, their physical ability and psychological cognition, e.g. ability in handling potential dangers in the surrounding environment, can impact their parents’ attitudes or evaluations, which in turn influences the outcomes (Meester et al., 2014; Yelavich et al., 2008). In particular, interplaying with the environment, a pupil develops and matures. Therefore, the age of transition from childhood into early adolescence (12-15 years) is characterized as a key time when both independent mobility and capacity (children's ability to move around in their neighborhood and handle the traffic condition alone) changes (Page et al., 2010). Page et al., (2009) has also shown in the PEACH project that children within the transition period with greater independent mobility engage in a greater volume of physical activity. The enhancing of capacity and independence during this period has reinforced the significance of promoting AST among this age group. Besides, given the fact that child's self-efficacy are often connected with parents' attitudes or evaluations, in order to promote AST effectively, it emphasizes the significance in research parents' decision-making process during their children's age transition.

Household level. The influences of household factors are always complex and

important, and precisely because parents have a crucial role as the chauffeur or the permission-giver in the school travel. According to the existing studies, the major household-level influences on school travel outcomes include social-economic status, the composition of the household, activities of each household and household attitudes (Mitra, 2013; Abrahamse & Steg, 2009; Johansson, 2006; McDonald, 2008).) Firstly, regarding social-economic status, most previous studies have looked at this kind of variable, i.e. family income, parental education and employment (Abrahamse

(22)

& Steg, 2009; Shokoohi et al., 2012). The family income is usually explored as it is correlated with the number of household cars, which directly determine the travel options available to the household.

Besides, in the case of school travel, social-economic status also works indirectly to shape travel needs for the school run, e.g. choosing a better neighborhood with long distance to school and vice versa. Furthermore, previous research also suggests that composition of the household and the schedule of each household member may influence the decision of school transportation. Evidence shows that parents are likely to combine their working commuting with their children's school run. As a result, when the school travel interferes with parents' work schedules, children are less likely to use active modes (McMillan, 2007). Also, according to McDonald (2007) and McMillan (2007), having siblings have a positive association with active travel and negative association with inactive travel, while Yarlagadda & Srinivasan, (2008) reports when multiple children in the family attend school, their mothers are more likely to drive the children.

Obviously, according to previous studies, inconsistent results regarding family factors are also found. Except data or measurement bias, a possible explanation for this inconsistent pattern is differences across countries. A good example is presented by Potoglou & Arslangulova (2017) that in the US, pupils from high-income family are more likely to travel to school actively than those from low-income, while there is an opposite pattern in Brazil. These contradicting results limit the possibility to generalization. This indicates the need for more context-specific research in this issue. In addition, in view the factors at the household level, it has also stated that psychological factors do matter in the transportation decision-making, e.g. family perceptions, attitudes, and habits. The family perception and attitudes are often associated with children's self-efficacy and physical environment, which is discussed above (child level) and in the following part (urban environment level). Despite these two, it is worth mentioning the habits' role in the household decision on school transportation. Evidence suggests that children's school travel behavior is likely to be under a strong sense of habitual control (Murtagh et al., 2012; Shannon et al., 2006). Referring to Shannon et.al, (2006, p2) explanation, it is likely because "both repetition

(23)

(i.e., they are typically made each day of the school week) and situational stability (i.e., they take place at approximately the same time of day, have the same start and end points and typically constitute the same route) are the main characteristics of school-based travel". As the results, the repeated past behavior can easily predict household future travel behavior, which is notoriously difficult to change.

As stressed by many studies (Bamberg, 2006; Jones & Ogilvie, 2012; Murtagh et al., 2012.; Stanbridge & Lyons, n.d.; Verplanken, Walker, Davis, & Jurasek, 2008), it is valuable to place attention on the period of habit discontinuity, e.g. home relocation, school transition, in an effort to understand how active commuting might be promoted more effectively. It is mainly because of the habit discontinuity hypothesis, described by Verplanken et al. (2008), positing that behavior change is more likely to happen when habits are broken by a change of context. More specifically, when the context in which a behavior is habitually performed is subject to change, the environmental cues that automate behavior are no longer present and there is an increased likelihood that individuals will re-consider their behavior and adopt alternatives (e.g., walking or cycling to school). Murtagh et al., (2012) states that interventions delivered shortly before, during or after a context change have been shown to be effective at changing habitual behavior. Yet, there has been little study focusing on dealing with habits in school travel behavior, and focusing on the period of school transition, e.g. primary-secondary transition.

Urban environment level. Not only the urban form, as a result of urban spatial

planning, but also built environment in the neighborhood, is always discussed in the school travel study and other travel behavioral studies (McMillan, 2005; Mitra, 2013). It is because that both individual level and household level are within and exposed to the physical environment. The structure and quality of surrounding environment easily influence their day-to-day activity indirectly and directly. On the one hand, in the case of school travel, the most common factors discussed in the previous study are home-to-school proximity (distance), traffic safety, levels of urbanization, street connectivity, and infrastructure and facilities (Mitra, 2013). The home-to-school distance is regarded as the influential factor associated with the decision on active travel mode. Many studies report that the closer pupils live to school, the more likely they choice cycling and walking for school commuting (Curtis, Babb, & Olaru, 2015;

(24)

Mitra & Buliung, 2012). Also, it has been consistently shown across studies that children who live in more urbanization and high-density area are more likely to use active transportation compared to those who attend schools in rural or suburban areas (Bringolf-Isler et al., 2008; McDonald, 2007; Merom et al., 2006). As explained by McDonald, (2007), it may be due to the structure of the transportation system (e.g., road layout), and geographical organization of buildings, connectivity, distance, and pedestrian facilities.

On the other hand, further research on the surrounding environment has also been shifted to both real environment and the perceptive environment. The later is even considered by some researcher to be more important than the real environment in the case of school travel decision-making (KERR et al., 2006; Lang et al., 2011; McMillan, 2007). It is because parents' perceptions and attitudes are often found that directly determine the pupils' actual transportation modes. For example, parental concerns of risk and safety around the school and negative perception about neighborhood built environment have always been identified in the existing literature as a negative factor in pupils active school-based travel. Thus, research in how the real environment is translated into parents' perceptions is needed in the understanding of school travel behavior and making a progress in AST (D’Haese et al., 2015; De Meester et al., 2014; Easton & Ferrari, 2015a; Raktim Mitra & Buliung, 2012).

Context level. The context, including social-political context from national to local

level, as well as the local natural environment, has frequently been discussed in the school travel behavior (Mitra & Buliung, 2015). In the recent years, some studies try to focus on how the urban planning and transportation policies affect transportation choice for school journeys. For example, Local Transport Plans, as well as School Travel Plans, is often considered as an important approach to influence the practices of school travel through funding the infrastructure projects, adjusting road construction and speed limitation (Baslington, 2008; Hinckson, 2016; Rice, 2008). Although there is not a consensus on the type of policy, most researchers still believe that policies are necessary to improve the AST. Despite policies and planning, some soft transportation measure and initiatives in community-based organizations, e.g. Safe Routes to School, the Walking School Bus, and the Walk to School, have often been discussed. Few evaluations study of SRTS programs have been published,

(25)

especially in US, to report a success story of increasing the number of children walking to school (Boarnet et al., 2005), while small or non-significant effects of this initiatives has also been stated by some studies (Chillón et al., 2011; Murtagh et al., 2012). Two possible explanations, given by Murtagh et al., (2012, p1), are that “initiatives or interventions have been developed without a theoretical basis and active travel is strongly controlled by habits, which are notoriously difficult to change”. This indicates that taking into consideration the effects of habituation and theory-informed study is further required in this areas.

2.1.2 Previous qualitative studies in school travel

Summing up the above, a considerable number of studies related to school travel behavior have been published, contributing to a better understanding of this kind of travel behavior and a guiding for AST promoting. Yet, most of the published literature related this topic has been generated from quantitative studies, by employing a cross-sectional, survey methodology, as also stated by many researchers (Carver et al., 2013; Lang et al., 2011; Mitra, 2013; Panter & Jones, 2008; Pont et al., 2009; Richichi, 2009; Sirard & Slater, 2008). Although such a quantitative approach is an important way in identifying the correlates associated with a given transport mode, as presented above, it doesn't lend itself to explore the dynamic nature of decision-making processes and to answer the question how to encourage AST effectively (Faulkner et al., 2010). As argued by many studies (Abrahamse & Steg, 2009; Easton & Ferrari, 2015a; Mandic et al., 2015; Mitra & Buliung, 2012, 2015; Potoglou & Arslangulova, 2017), evidence from those quantitative-led literature mainly suggests that school travel behavior is affected by multiple levels of influences and those correlates are hard to be generalized or even inconsistent. To capture the interplay of those underlying factors within the decision-making process and unraveling the complexities of this kind of behavior, qualitative approaches are required. As explained by Sallis et al. (2006), qualitative research facilitates an understanding not only about the ‘what’ but also the ‘how’ and ‘why’. Under this consideration, few qualitative studies on school travel behavior have been published. Because this study is qualitative in nature, the available qualitative research on this topic has been reviewed in order to identify future qualitative research directions and prepare for this study.

(26)

The review shows that most of these qualitative study focuses on exploring parents' or children' perception on active travel to school, in order to identify the potential barriers and facilitators for AST promotion (Hopkins & Mandic, 2017; Kirby & Inchley, 2009; Ahlport et al., 2008; Guell et al, 2012). The frequently mentioned barriers include safety concerns, distance, walking routes/sidewalks, household travel schedule, and some nature environment constraints, e.g. weather, season and hills. Some studies also seek to investigate parents' or children' perception particularly on AST initiatives, e.g. Safe Routes to School, Walking School Bus and School Travel-Plan programme, for valuable suggestions which formed the basis of future plan implementation (Atrooshi, 2017; Hinckson, 2016; Carver et al., 2013). Relatively small part of these qualitative studies turns to explore the dynamic process of household's decision-making, in an effort to understand the relationships among influencing factors of school travel transportation decision and how these factors influence the final travel mode choice (Faulkner et al., 2010; Lang et al., 2011). In this regarding, two studies are found in the AST literature. One was conducted in Auckland, New Zealand (Lang et al., 2011), the other is in Canada (Faulkner et al., 2010). To the best of my knowledge, there is no study drawn from the UK case, let alone the Wales cases.

In these studies, sample strategies vary from study to study. For example, in exploring various barriers and facilitators of using AST, Ahlport et al., (2008) employ a very different sample, focusing on children in 4th and 5th grades, and their parents who all to live close enough to walk or bicycle to school, namely within 1.5 miles and from four primary schools in North Carolina. The sample includes both active travelers and non-active travelers. Although this sampling strategy has some advantage, such as comparing differences between children and parents perception, and between active travelers and non-active travelers, it still has some shortcomings. The main drawback is due to the focus on the author-identified "reasonable distance" in this case and it limits the exploration within those who live far from this distance and about acceptable distance. Also, in the study of Lang et al., (2011), they seek to examine the factors that lead parents to select travel modes for their children's trip to school and address the question: why do some parents drive their children to primary school, while other parents encourage walking? One school from a suburban in Auckland,

(27)

New Zealand is selected as the study area and three type of parents (N=30), those who generally drive their children to school, whose children generally walk, and whose children had changed from being driven to walking are all selected. Although this study has a large sample size of 30, with an equal split of three types, all participants are recruited from one school. To some extent, this sampling strategy limits the generalization of the findings, especially for other schools with different types of built environment. As stressed by Richichi (2009), when informing policy or intervention for AST, it is important and necessary to consider schools within diverse types of neighborhoods, for a better representing in diverse built environment and SES.

In terms of methods, focus group and interview have often been used in qualitative studies of school travel for collecting data. By enabling parents, children, and other stakeholders to voice their own opinions and experiences of commuting to school, a range of rich contextual information that has not been considered in the many survey-based studies emerges in this data collection process. For example, in the study of Lang et al., (2011), focus group is adopted by researchers to produce in-depth, interactive discussions within parents around the topic of school travel behavior, e.g. the factors influencing participants' transport decisions, their perceptions of walking and driving, and the social meanings attached to these. Also, some studies also conduct focus group sessions with students and parents together (Hopkins & Mandic, 2017, Hinckson, 2016; Ahlport et al., 2008). The focus group method in those study offers a rich window into the diverse views and perceptions underlying school travel behavior.

Despite focus group, due to flexibility, interview method has often been used in many qualitative studies (Atrooshi, 2017; Guell et al., 2012; Faulkner et al., 2010). For example, Guell et al., (2012) recruit 65 interviewees (parents of primary school students), in the City of Denver (the US), to re-contextualize certain factors that are often interpreted as ‘barriers' in quantitative research. Also, Atrooshi, (2017) conduct semi-structure interviews with 16 parents of the primary school student in Northern Ontario (Canada), guided by a social marketing approach, to explore their perception s of a Walking School Bus in order to inform an effective messaging and promotional strategy. Compared to focus group, interview method provides research with more opportunities to explore the individual's dynamic and complex process in school

(28)

transportation decision, thus contextualizing those influential factors, barriers, facilitators for AST.

2.1.3 Summary research gaps and implication for further study

Although the above studies provide crucial evidence about children’s school travel behavior, additional research is required for the following three reasons:

Firstly, although previous empirical studies have proposed a wide variety of findings, it has often been argued that there are distinctive inconsistencies across studies due to study design limitation, measurement bias and cross-country variations (Potoglou & Arslangulova, 2017). Besides, the difficulties in the generalization of the findings have also been concerned especially when there are existing differences, e.g. culture, natural environment, policy environment, etc. across countries. Therefore, the need for further context-specific research for informing local AST policies and interventions is highlighted by many researchers (Cavill & Watkins, 2007; Richichi, 2009). Particularly, a majority of AST research is drawn from the USA, Australia, and New Zealand, while very little research concentrated on Wales (the UK) case where with a high prevalence of childhood obesity and car use for school journeys. Case studies such as this are thus necessary to gain a more comprehensive understanding of school travel behavior in the Wales. Furthermore, it is important to remember that schools even in one city can vary in terms of the types of neighborhoods, such as the built environment and SES. When design a sample framework, providing schools within diverse types of neighborhoods, e.g. different built environment and social factors, need to be taken into consideration, especially for a better implication of AST policies and interventions.

Secondly, most of the published literature related this topic has been generated from quantitative studies, by employing a cross-sectional, survey methodology. Although such a quantitative approach is an important way in identifying the correlates associated with a given transport mode, while it doesn't lend itself to explore the dynamic nature of decision-making processes and to answer the question how to encourage AST effectively (Faulkner et al., 2010). In the filed of school travel behavior study, capturing the interplay of those correlates revealed by previous

(29)

quantitative study within the decision-making process and unraveling the complexities of this kind of behavior is warranted (Carver et al., 2013; Lang et al., 2011; Mitra, 2013; Panter & Jones, 2008; Pont et al., 2009; Richichi, 2009; Sirard & Slater, 2008). Furthermore, Richichi (2009) argue that self-selection in the quantitative study can weaken some of the research findings on school travel behavior, thus highlighting the requirement of further research for more new insights.

Thirdly, there is a lack of attention for the role of school-transition in AST in the existing literature. As stressed by many studies (Bamberg, 2006; Jones & Ogilvie, 2012; Murtagh et al., 2012.; Stanbridge & Lyons, 2006.; Verplanken et al., 2008), it is valuable to place attention on the period of habit discontinuity, e.g. school transition, in an effort to understand how active commuting might be promoted more effectively. The last but not least, the theory-informed research design and analysis are further required in the school travel study. Indeed, to date, few qualitative studies on school travel behavior have been published to revealed the perceived benefits and barriers associated with AST by conducting interview or focus groups discussion with children or parents, while very little of these study explore this travel behavior from a theoretical perspective. However, the nature of human behavior is a very complex issue. For a better understanding of human behavior, a theory is necessary and important. As stated by Singleton, (2013, p4), "a theory is a valuable tool in the scientific study of phenomena, especially when related to human behavior – which includes travel behavior". It is mainly because that a theory presents a systematic and accessible way of understanding and explaining events, behaviors, and/or situations. Regarding understanding travel behavior and prompting behavior change, theories can also be useful as it guides the search to understand why people do or do not practice a given behavior, suggests how to develop more effective ways to influence and change behavior, and particularly helps to identify appropriate targets for a more effectively intervention. Evidence from a serious of systematic-review and meta-analysis has shown that adopting a theoretical approach to guide the development of interventions aiming at behavior change is of importance for the greater efficacy (Arnott et al., 2014; Chillón & Evenson, 2011; Moghtaderi et al., 2015). Given that intervention development work in this area requires improvement

(30)

and the lack of theory-based empirical school travel study, more works employing a theory-based qualitative method are further required.

2.2 Current AST promotion measures A shift from individualist approaches to social practices modification

Over the past decades, the growing concerns about the prevalence of car use for school journeys and associated impacts have prompted children's AST to become a hotly debated public and policy issue surrounding childhood health, mobility and environmental sustainability (See Appendix 1 AST in a policy context). Pressure is growing on transport sectors to tackle this problem behavior with a view to providing sustainable school commuting so that it can make a positive contribution to children themselves and the whole society. So this has led to the development and implementation of a range of transport policy measures aiming at reducing or changing car use, leading a mode shift from driving to active modes, i.e. walking and cycling (Bamberg et al., 2011; Gärling et al., 2009). However, changing the behavior of a target population has been described as challenging and hard work (Spotswood et al., 2015).

Historically, the UK government has largely relied on "hard" measures, include improvements to infrastructure, management of public transport services, enhancement in proximity and accessibility, prohibition or rationing of car use, increasing the cost of car usage (ibid.). From the theoretical perspectives, hard planning approaches mainly through modifying the objective environment to guide travelers to judge the environment and travel options' consequences and then provide them with sufficient reasons to change current car travel behavior (Bamberg et al., 2011). However, given the difficulties in implement of those hard measures due to public opposition or political infeasibility, they are not effective in realizing car-use reduction (Cairns et al., 2008). Also, due to the budget limitations, some of them are even hard to be real implemented.

Therefore, over the past decade, “soft” policy measures, which use techniques of information dissemination and persuasion to influence car users to voluntarily switch to sustainable travel modes, have attracted increased attention. From the theoretical

(31)

perspective, the mechanism of “soft” measures is to target individual decision making as the point of change, by altering their perceptions, and by motivating and empowering them to switch to alternative travel options, typically relying on individualist approach and psychological approaches. As summarized by Bamberg et al., (2011), the most common measures cover voluntary change measures, psychological and behavioral strategies, or mobility management tools. In relation to AST promotion, frequently implemented measures include School Travel Plans (STPs), Safe Routes to School Strategy, School Travel Initiatives, and Travel awareness campaigns. However, as stated by Stevens, (2010), the change of behavior has been negligible after many years of encouraging in the travel model shift through those mainstream measures. As suspected by many researchers, there is a big question over whether those measures “which assume that people lack information or motivation, or that they need help, and that once one or more of these is supplied they will be more inclined to change” really work on prompt a voluntary behavior modification (Bonsall, 2009; Spotswood et al., 2015, p23). However, in line with the neo-liberal direction of UK policy, individualist approaches, allowing individuals to manage their behavior change rather than forcing a change through the regulation instruments, are still politically popular in the UK in terms of behavior change intervention in recent years. Meanwhile, given the fact those approaches are able to render policy responses relatively straightforward, it can also partly explain why it is politically popular towards behavior change (Hargreaves, 2011).

Specifically, individualist approaches look to individuals’ beliefs and attitudes as predictors of behavior, and largely refer to social psychological research, which uses theories such as the Theory of Planned Behavior (TPB), developed as an extension of the earlier “theory of reasoned action”. Central to these approaches and underlying theory is the assumption that if the necessary cognitive components can be identified and modified, behavioral changes will automatically emerge (Hargreaves, 2011). Therefore, apart from the first suspicion of its effectiveness, in particular, critics have also argued that such approaches are excessively individualistic and fail to realize that social values, social relations, and built environment materials are also intrinsic to the performance of behavior, rather than only variables existing within individuals’ decision-making processes (Guell et al., 2012; Hargreaves, 2011; Shove, 2003).

(32)

Consequently, a reinvigorated approach based on theories of social practice (SPT) has sought to address this shortcoming brought about by those individualist approaches. Unlike the TPB, in general, the central proposition o SPT is the recognition that individuals' body and mind become the ‘carriers' of social practices, performing various behaviors that the practice requires. In other words, for the purposes of SPT, the individual's beliefs and attitudes associated with behavior choices are no longer the unit of inquiry, but instead it centers on "how practices form, how they are reproduced, maintained, stabilized, challenged and ultimately killed-off; and on how practices recruit practitioners to maintain and strengthen them through continued performance, and on how such practitioners may be encouraged to defect to more sustainable practices" (Hargreaves, 2011, p84). In particular, the key implication of this theory is that "the sources of change behavior lie in the development of practices themselves", as emphasized by Warde (2005, cited by Hargreaves, 2011, p83). In this sense, the utilizing of this theory complements to the disadvantage of the TPB as it has often been argued that the TPB is not useful or effective when it come to guide a design of intervention aimed a real behavior shift as the utilizing of this theory is more likely to help for finding cognitive targets for change than offering suggestions on how these cognitions might be changed. Owning to those valuable contributions in understanding travel behavior and providing implication for population-level behavior change, indeed, SPT has attracted attention recently and has raised a radical discussion about the most proper way for sustainable behavior changes, while relative studies are sparse.

2.3 Theoretical Framework

In order to address research questions and fill the research gaps, the theoretical framework of this study referred to both the Theoretical Domains Framework (TDF, originally proposed by Michie et al., 2005 and refined by Cane et al., 2012) and Three-Elements Model (Shove et al., 2012, based on Social Practice Theory), from both an individual's and social practices' perspectives to exploring the driving decision for school journeys. To be specific, the constructs outlined by the TDF have been used to design a theory-informed interview guide that was used to explore parents' perspectives of school travel and then uncover the influential factors associated with car driving choices, while the Three-Elements Model have been used

(33)

to further discuss the parents' driving decision as part of social practices, with a focus on providing implications for AST in Cardiff, including identifying barriers and facilitators for AST engagement within the context.

2.3.1 Theoretical Domains Framework (TDF)

This framework was developed by Susan Michie and team through an expert systematic consensus process that synthesizes a multitude of coherent behavior theories 128 key theoretical constructs from 33 different psychological and organizational behavior theories—into a single framework that includes 12 theoretical construct domains (Michie et al., 2005): (1) knowledge, (2) skills, (3) social/ professional role and identity, (4) beliefs about capabilities, (5) beliefs about consequences, (6) motivation and goals, (7) memory, attention and decision processes, (8) environmental context and resources, (9) social influences, (10) emotion, (11) behavioral regulation, and (12) nature of the behavior. Figure 1 provides an overview of the 12 domains from the TDF and their descriptors.

Understanding decision-making associated with children being driven to school and moving toward active school travel. A theory-informed qualitative study