BARRIERS AND STRATEGIES TOWARDS A MODAL SHIFT FROM
CAR DRIVING TO MORE SUSTAINABLE URBAN TRANSPORT
MODES IN CAR-DEPENDENT CITIES
by
Juliane El Labani
B.A. in Geography and Planning
Paris-Sorbonne Université, Abu Dhabi, 2015
Thesis Submitted in Partial Fulfillment of the
Requirements for the Degree of
Master of Science
In the
Department of Urban and Regional Planning
Faculty of Social Sciences
Student ID: 11793740
Thesis Supervisor: Dr. Marco te Brömmelstroet
June 11
th, 2018
Copyright in this work rests with the author. Please ensure that any reproduction or re-use is done in accordance with the relevant national copyright legislation.
ATTITUDES TOWARDS MODE
CHOICE AND TRAVEL BEHAVIOR
Statement of Originality
This is to confirm that to the best of my knowledge; the content of this research is my own work. This research has not been done for any other purposes or submitted for another degree. I certify that the academic content of this thesis is the product of my own work and that all the sources have been retrieved from accredited online library platforms and/or online newspaper articles.
Abstract
Developing suitable transportation policies and/or initiatives can be a daunting and difficult task when attitudes towards mode choice and travel behavior are vaguely interpreted. Understanding the underlying values of the relationship between ‘transport geography’ and ‘social psychology’ is crucial in this research because different types of people need to be treated in different ways to be able to implement better transport policies.
How different attitudes towards mode choice are formed and what reasons this is based on is studied in this paper by using social cluster analysis to segment people based on the ‘Theory of planned Behavior’, as well as other psychological and attitudinal attributes to determine their ‘potential switchability’ and compare how their attitudes are formed and differ from each other. 5 clusters were identified which are the (1) ‘status-oriented automobilists’, (2) ‘constrained riders’, (3) ‘rational drivers’, (4) ‘guilty drivers’, (5) and ‘aspiring environmentalists’. This study showed that the clusters are largely car-dependent in Dubai because it follows a lifestyle with strong social pressures, because people are consciously and unconsciously drawn towards thinking that cars define a person’s social status.
Acknowledgements
This work was carried out between January and June 2018 in Dubai and is part of the Master program in Urban and Regional Planning at the University of Amsterdam. I am grateful for the support and optimism of my supervisor Dr. Marco te Brömmelstroet for this concerning thesis. I really appreciate the encouragement and support to include me in all meetings although I was unable to be physically present.
I would also like to thank my friends and family who kept me motivated during the research process and always encouraged me and gave me strength to continue to do better on difficult days. This thesis would not have been possible without the mental support of the loved ones around me.
Table of Contents
Page No.
List of Figures, Maps and Tables
6
Chapter 1 – Introduction
7
I. Background of Study
7
II. Theoretical Framework
9
a. Attitudes towards travel mode choices in a socially clustered environment
9
b. Transport Mode Choice and Travel Behavior
11
c. Social Status and Social Surrounding
17
d. Transport policy implementation
18
III. Conceptual Framework
19
IV. Problem Statement
20
V. Research Question(s)
20
a. Sub Questions
20
b. Expectations
20
VI. Research Location
21
Chapter 2 – Research Design
26
I. Methodology
26
II. Methods
26
I. Sub-Question 1 – Questionnaire Development
26
II. Sub-Question 2 – Follow-up interview
29
III. Sub-Question 3 – Document Analysis
30
III. Ethical Considerations
32
Chapter 3 – Results
32
I. Getting to know the audience
33
II. Establishing Number of Clusters
34
III. Cluster profiles
38
IV. Formation of attitudes
44
a. Attitudes towards public transport
45
b. Attitudes towards car driving
48
c. Attitudes towards Influence of social surrounding
49
d. Intentions to use public transport
51
e. Comparing clusters with each other
52
V. Translating this into policy
55
Chapter 4 – Discussion
61
I. Discussion
61
II.
Limitations of Research – suggestions for further research65
Chapter 5 – Conclusion
66
List of Figures
Page No.
Figure 1: The Theory of Planned Behavior
12
List of Maps
Map 1: Location of UAE
23
Map 2: Location of Dubai
23
Map 3: Emirates Living Location
24
Map 4: Emirates Living Master Plan
24
Map 5: Emirates Living – Aerial View
25
List of Tables
Table 1: Comparing theories of Reasoned Action, Planned Behavior and
16
Repeated Behavior
Table 2: Outline of survey questions
27
Table 3: Follow-up Questions – Qualitative Approach
30
Table 4: Socio-demographic distribution of respondents
33
Table 5: Descriptive analysis of attitudes
35
Table 6: Cluster profiles
38
Table 7: Socio-demographics of five clusters
42
Table 8: Descriptive analysis of remaining statements based on the
43
five clusters
Table 9: Socio-economic details of respondents
44
Table 10: Coding of follow-up interviews
45
Table 11: Most relevant quotes of most frequent codes
46
Table 12: Corresponding transport policies
57
Chapter 1 – Introduction
I.
Background of Study
Reducing car ownership while increasing public transport usage has been debated about for the past few decades especially when considering the growing environmental concerns over the depletion of fossil fuels, pollution, health issues, accident fatality rate, and the impact of heavy congestion on infrastructure. However, attitudes towards the car generally still remain positive, because for most part, it is easier and more convenient to get from point A to point B in car-dependent cities, and it is a good investment, similar to owning a property (Global Cars Brands, 2015). In addition to that, public transport seems to not always be 100% reliable and causes for people to lose trust and would eventually not use it. Filled buses, poor service quality, delays due to mechanical or weather-related issues can divert one’s entire day which could be a contributing factor for why there is an average annual increase in car ownership of 8.2% in Dubai. Dubai’s vehicle density is recognized to be the highest in the region and one of the highest in the world with 540 cars per 1,000 residents meaning that more than 1 out of 2 residents own and drive a car (Shahbandari, 2015). Extreme car-dependent cities like Dubai are adopting smart growth ideas and sustainable efforts in order to decrease their ecological foot print and resolve their severe congestion issues, yet car ownership is an aspect of personal freedom that remains a much unchangeable mindset for many across the world.
Car-ownership keeps increasing (Shahbandari, 2015), and it is generally recognized that measures to approach unsustainable patterns of trip making add a detailed understanding of people’s travel behavior and many underlying psychological factors for making the choice of transportation able and how that differs from another person, which seems to be overlooked in travel research methodology (Anable, 2005). Moreover, there exists an urgency to promote transport related strategies which can reduce the dependence of private transportat along with the driving needs of a person, by providing alternatives (Beirao & Cabral, 2007) and also to reduce the car usage and its attractive aspects (Gärling &Schuitema, 2007). The obsession with ostentatious vehicles has become a reputation in the Emirates. Collecting them has been an expensive hobby mainly confined to locals and it is debated whether car-ownership remains
one of the principal components that explains a person’s social status in Dubai or not (The Telegraph - Roberts, 2015).
It is evident that increasing public transport usage and drawing the focus away from driving and owning a car has a lot of environmental benefits. Strategies with an impact on the image of public transport are supposed to be promoted. However, this type of transport needs to become more competitive and market-oriented as well. This calls for an understanding of people’s attitudes, moral beliefs, social surroundings and their overall travel behavior to avoid making public transport less and less viable and extremely socially stigmatized by imposing harsh rules and regulations or by marketing its usage the wrong way. This would make public transport become a minority desired form of transport. It is therefore very crucial and interesting to identify different environments of market segments of people based on their attitudes and travel behaviors to be able to understand the different perceptions towards mode use, and what outcome this has for introducing sustainable public transport.
The focal point of this study is on the part and significance of individual attitudes towards sustainable transport for movement choices. While some current examinations explored the modes of transportation and decision in view of fragmenting transport users, they only utilize individual attitudes and their socio-demographical constellations as primary determinants for day by day travel (e.g. Anable, 2005). On one hand, these methodologies give a valuable and propelled reason for potential more individualized measures. Then again, they don't consider the setting subordinate part of attitudes in regard to various areas of individual travel behavior, particularly in perspective of their inspirations, world-views, and affirmations of sustainability issues. With the guide of cluster analyses, the traveler grouping can consider all the more contributing elements which influence travel behaviors and will be discussed in the following chapters. This corresponds with the appropriation of marketing and consumer behavior where it is common to establish homogenous groups of customers because then they can be targeted in the same way when they show to have similar preferences and needs (Wedel and Kamakura, 1998). The overall objective of this study is to:
1. Understand the travel behavior of individuals by grouping them with the use of cluster analysis based on their willingness to take public transport which is called the ‘potential switchability’;
2. Understand how people’s attitudes are formed and how they differ from each other in socially clustered environments;
3. And to analyze how different transport policies correspond to that.
II. Theoretical Framework
A vast variety of underlying studies have illustrated the fact that transport planning practices require a strong understanding of the connection between ‘transport geography’ and ‘social psychology’ to understand the relationship between travel behavior, personality characteristics and mode choice. This is particularly important to be able to understand some of the key components which can support travel modes such as public transportation. This chapter comprises the theoretical framework and the main topics of which the main research question consists and studies the relevant literatures to find if and how they define and operationalize attitudes, as well as mode choice in relation with social cluster analyses based on psychological factors instead of focusing on socio-demographical dispositions only. These topics consist of mobility, transport geography with the principal relationship of social psychology. I will identify potential knowledge gaps that guide my research, in order to contribute to a larger societal context of understanding people’s travel behavior and mode choice as my principal objective of this research.
a. Attitudes towards travel mode choice in socially clustered environments
Despite the efforts to implement the most suitable transportation policies that can be applied for every type of population, the reality in recent years show apparent differences. Many studies and articles show that a social cluster analysis provides a very helpful tool to analyze the data of respondents to be able to indicate that different groups of people need to be serviced in different ways to optimize the chance of influencing mode choice behavior (Anable, 2005). However, the usual practice is that the market is segmented according to socio-demographic factors and travel behavior consolidating public transport users as well as car drivers, neglecting their necessities, expectations, and beliefs which can fluctuate significantly between various segments of the market (Anable, 2005; Jenses, 1999; Quattro, 1998; Stimulus, 1999). This shows that new clusters need to be identified according to the underlying psychological aspects, incorporating attitudes, perceptions and habits (Ajzen, 1991; Fujii and Kitamura, 2003). In other words, changing the mental components may likewise change travel mode
decisions, in spite of the fact that the level of service continues to be the same (Fujii and Kitamura, 2003). Thus, to pull in more clients to use public transport, it is imperative to find out more about the psychological variables that impact mode choices and the measures expected to decrease car-dependence. Today, social clustering approaches have been established to become a way of analyzing daily travel determinants by organizing data of people into meaningful groups or taxonomies based on a set of variables that describe the key features of the observations (Malo, 2015). The segments are then translated into a group of observations, which are similar to each other and different from observations in other groups making them statistically significant. This approach is usually applied in the field of marketing according to people’s similarity in several dimensions related to a product under consideration to be able to identify what types of customers buy what products. Once the groups are recognized, the prediction is also possible regarding the way they respond in various circumstances. Policies regarding marketing and other strategies are created and their types which allow better targeted and creative policies to emerge.
In view of sustainable mobility, there exists different studies which utilize certain attitudinal factors and qualities as a resolute component to convey a starting point to measure behavioral changes with regards to transport mode choices. Hunecke et al. (2007) studied the ecological effect of individual travel behavior, in which they recognized six psychological factors as important elements for private transport usage, with the conclusion that mobility-related attitudes are better indicators for determining mode choices than values are. In regard to travel behavior changes, the important components were arranged into two categories which are (1) ‘perceived mobility necessities’, and (2) ‘perceived behavioral control’ which comprise factors that depend on subjective assessments of the behavioral scope. For travelled distances, mental factors are less essential, in light of the fact that socio-demographic determinants like age and work status are the stronger indicators. Considering the social and mental components of an individual, some research used attitudes towards certain transport modes and lifestyle attributes to be able to differentiate multiple mobility styles from each other (e.g. Anable, 2005; Götz et al., 2003). One of these ventures is called 'City:mobil' and was examined by Götz et al. (2003). They assessed different orientations of mobility and found five distinctive portability styles which incorporate (1) 'the traditional domestics', (2) 'the reckless car fans', (3) 'the status-oriented automobilists', (4) 'the traditional nature lovers', and (5) 'the ecologically resolute’. Travel behavior patterns which are related to these segments demonstrated noticeable
dissimilarities, for example for car use rates. As the depiction of travelers stays restricted to socio-demographic factors, as for example, age, sex, education, and income, the cost of looking for and building new alternatives is for the most part too high and the expected benefits related with new choices excessively undetermined. Under these circumstances, travelers reuse past experiences to make their behavior less demanding and less hazardous. This may specifically be valid if they are constrained by budget, time, or social commitments (Gärling and Axhausen, 2003). An extensive study conducted by Anable (2005) then displayed inspirations and requirements for behavioral changes, which incorporate feelings, as for example, moral norm, environmental attitudes, worldview, identity, and habits (Stern and Dietz, 1994; Harland and Wilke, 1999; Axelrod and Lehman, 1993; Verplanken et al., 1994; Aarts and Dijksterhuis 2000; Garling et al., 1998; Forward, 1994, 1998). Comparable to existing lifestyle approaches, Anable used an arrangement of 17 factors associated with attitudes towards car use, the use of alternative modes, and the environment. Her cluster analysis of approximately 600 interviews conveyed a set of four car-owning and two non-car-owning segments which are the: (1) ‘malcontented motorists’, (2) ‘complacent car addicts’, (3) ‘die hard drivers’, (4) ‘aspiring environmentalists’ (all car-owning), (5) ‘car-less crusaders’ and (6) ‘reluctant riders’ (both non-car owning). Motivations and barriers to change travel behavior and use alternative modes differed widely between the clusters. However, for all categories, some influence from environmental concerns and attitudes were found. To conclude, travel behavior needs to be carefully examined to be able to determine travel demand characteristics over time with the use of social cluster analyses. Travel behavior is a complex topic, which will be described in the following section.
b. Transport mode choice and travel behavior
Before diving into the definition of what travel behavior is and how beneficial its understanding is for social cluster analyses and transport policy makers, it is important to realize that in case of any kind of journey, there are various types of transport which are available to the public and all of them have specific advantages and disadvantages, characteristics, and costs. This phenomenon is translated by the ‘utility theory’ which represents the core theory in the transportation field as it is a measure of preferences over some set of goods and services to represent satisfaction experiences by the consumer from a good (Pardee & Phillips, 1970). This is translated as the trade-off between time, cost, comfort and convenience when choosing the
right mode of transport for a person’s own needs (Townsend, 2017). Given that a trip must be made, commuters choose the one with the lowest disutility so that basic rational decisions can be made to maximize a person’s own interests. This implies that car users settle on reasoned decisions (Everett and Watson, 1985; Golob, Horowitz, &Wachs, 1979; Gould &Golob, 1998; Hensher&Stopher, 1979; Recker & Golob, 1976; Verplanken, Aarts, Van Knippenberg, 1994), in light of the fact that their travel mode choice relies on attitudes and preferences towards accessible transport options which could rather be characterized as being negative towards public transport usage. Finally, there exists a noticeable trade-off between perceived benefits and costs when it comes to actual mode-choice behavior which has made substantial progress towards understanding how travel modes are picked, particularly when incorporating psychological properties and numerous components at various levels (Yang, 2016) of which the 'Theory of Planned Behavior' (Ajzen, 1985 &1991) might be the most widely applied theory which dates back to the late 1970s (Dobson et al., 1978; Tardiff, 1977; Tischer and Phillips, 1979).
Figure 1: The Theory of Planned Behavior Source: Ambak et al, 2016
Consequently, other research has likewise demonstrated that people travel and adjust their travel behavior based on their preferences, attitudes and perceptions (Gehlert et al., 2013; Bohte et al., 2009; Fujii and Gärling, 2003). As shown in Figure 1, the 'Theory of Planned Behavior'
(TPB) measures how human activities are guided and predicts the event of a specific behavior, provided that behaviors are volitional with a willingness to act (Steg, 2005). Individuals' expectations are the main direct psychological determinant of conduct. These intentions are seen as a summary of the considerable number of advantages and disadvantages an individual considers when purposely thinking whether he or she ought to play out a behavioral option or not (Bamberg et al., 2007). Additionally, a man must be willing to put psychological and physical effort in playing out the picked behavioral alternative. All things considered, human behavior is guided by (1) behavioral beliefs about the likely outcomes of behavior and the assessments of these results. These deliver an ideal or unfavorable attitude toward the behavior; (2) normative beliefs about the presumption of others and inspiration to follow these presumptions (Ambak et al, 2016; Bamberg et al., 2013). This leads to social norm and perceived social pressure; (3) control beliefs about the presence of components that may encourage or obstruct its execution. These offer ascent to apparent behavioral control (Ajzen, 1991). Such conceptualizations are encouraging a comprehension of the idea of the connection to the car and the degree to which people see specific obstructions to change and the intention to do so.
Intentions depend on the evaluations of attitudes and is characterized as having constructive, contrary or mixed evaluative reactions to some stimuli (issues, items, or people) which comprise the (1) cognitive factor (knowledge and perceptions of the stimuli), (2) affective factors (sentiments, feelings, values), and the (3) behavioral/conative factors (acting in light of the other two factors) (Rosenberg and Hovland, 1960). Cases of cognition incorporate the information of obtaining public transport schedules and routes, sensitivities to transportation expenses, and cases for affective attitudes are perceptions of public transport use and the sensitivity to the environment.
To make use of data regarding attitude statements, it is always important to consider determining the intensions of respondents along with what they want for the right prediction of their behavior, because desires are translated to be positive only, instead of also including negative attitudes towards a mode of transport. Attitudes are also greater indicators of mode choice as opposed to socio-demographics and travel needs (Ajzen 1987; Dawes & Smith 1985). The second component of intentions presents the social norms which reflect the person’s perceived expectations of significant reference of people who think he or she should or should
not carry out that option. These norms can either be personal or subjective. Personal norms incorporate the individual conviction that acting unquestionably is either right or wrong in view of the ethics every individual carry along. Then again, behavioral effects of social norms depend on social pressure that is the dread of social sanctions when the inspirations don't consent to these reference groups (Bamberg et al., 2007; Steg, 2005; Cialdini et al., 1991; Festinger, 1954; Masters and Smith, 1987; Schlenker, 1980). There is a great collection of literature which confirm that personal as well as subjective norms are a noteworthy indicator of the goal to utilize other transport modes more than the car. In addition to that, they add to the clarification of pro-environmental practices like recycling, energy conservation, or 'green' consumerism, regardless of whether the commitment of these practices to reduce environmental damages is regularly undetectable and objectively small.
Lastly, the third part which defines intentions is the perceived behavioral control (PBC). PBC is conceived as sometimes having a direct prescient impact on behavior in conditions when a person’s view of control coordinates the measure of real control in terms of what they can exercise in real life. This implies that higher states of PBC ought to strengthen a person's aim to perform the behavior, and the low level PBC should be less encouraged to play out the behavior (Barua, 2013). PBC is valuable in surveying a person's real control for a particular circumstance. It is likely to influence intentions and change behavior in an indirect way. Subsequently, it is thought to be an extra indicator of conduct (Ajzen, 2002). The behavioral accomplishment of PBC relies on certainty and precision of perceptions. For instance, if an individual sees low precision of perception, PBC may not be reasonable with respect to little information (Ajzen, 1991). Studies uncover that TPB depends on the nature, adequacy and formulation of the PBC construct as it has been conceptualized and operationalized in an assortment of ways (Ajzen and Madden, 1986; Beale & Manstead, 1991; McCaul, Sandgren, O'Neill, & Hinsz, 1993; Wankel and Mummery, 1993). Eagly & Chaiken (1993) established that TPB constructs demand further scrutiny. They specify that individuals either have strong or weak perceptions of behavioral control as they are pretty much inclined to perform the behavior. Thus, PBC is likely to direct the impacts of attitudes on behavior. To conclude, TPB has been applied effectively to anticipate and clarify different practices, as for example, weight reduction, voting decisions, smoking cessation, sociologies, transportation and committing traffic violations. In the social cluster analysis domain, TPB is also used to analyze how willing
people are to use public transport and how much the three above mentioned variables contribute to that.
However, travel behaviors can also be unintentional which presents a twofold nature affected by cognitive processes. It is defined to either be under a person’s volitional control where mode choice decisions are done rationally, or behavior can be uncontrollable as individuals are not constantly conscious of their behavior and indirectly adapt to positive or negative habits, as compared in Figure 1. This is particularly valid for every day travel that is executed on a repetitive basis and that therefore may turn into a normal or ongoing habit which strongly impacts a person’s travel mode choice. A person may pick an attractive travel mode rather than the ideal one which implies that rationality is limited by his/her cognitive limitations, the accessible information, restricted amount of time available to decide, the cost of research and learning, and inaccurate memory (Yang, 2016). This implies if habits are strong, the relationship between attitudes and behavior including intentions is weak which also counts the other way around and requires an attention on including a person's habitual behavior in social cluster analyses (Van Acker et al., 2010; Bamberg and Schmidt, 2001; Anable 2005; Nordlund and Westin, 2013; Haustein et al., 2009; Eriksson and Forward, 2011; Gardner and Abraham, 2007; Haustein and Hunecke, 2007). Behavior is habitual so that it is, at least in part, under the direct control of the stimulus situation. Frequency of past behavior is thus an indicator of habit strength that can be used as an independent predictor of later behavior in cluster analyses. Moreover, the reason for repeating a behavior may simply be that the intention (e.g. to drive to work) is formed repeatedly. Thus, other methods such as deliberate vs. script-based choice are proposed to classify the behavior. All in all, understanding the mechanism of how an individual’s perception is formed and why persons living in close proximity have different perceptions (e.g. how perceptions are modified by the individual’s characteristics) is crucial to the travel mode choice function and must be analyzed based on the influence of their social surroundings.
Table 1: Comparing theories of Reasoned Action, Planned Behavior and Repeated Behavior
Definition Behavior is the result of rational choices and voluntary control. Several beliefs are associated with one stimulus, because several attitudes of this stimulus are evaluated (Van Acker, Van Wee, Witlox, 2010; Nordfjaern et al, 2014). This theory specifies that beliefs about negative and positive consequences of the behavior in question lead to an overall favorable or unfavorable attitude towards that behavior, which in turn, leads to an intention to perform the behavior (Kroesen et al, 2017). The sum of all related beliefs determines the attitude towards that stimulus.
Unlike TRA, TPB is based on an expectancy-value theory which considers a third determinant of intention, namely perceived behavioral control which refers to the perceived ability to perform a behavior (Van Acker, Van Wee, Witlox, 2010). It posits that individuals form intentions based not only on their attitude toward the behavior and its outcomes, but also on their perceived ease of performing the behavior and the social pressures they feel from their surrounding social network (Thogersen, 2006, Bamberg et al., 2003; Anable, 2005).
Individuals are not constantly conscious of their behavior. Initial behavior remains the result of relevant attitudes and beliefs. But once the behavior is repeated, it becomes a habit and decision-making is no longer based on attitudes and other well-reasoned influences. There exists a trade-off between attitudes and habits in the perception of behavior. If habits are strong, the attitude-behavior relationship is weak, and vice-versa (Van Acker, Van Wee, Witlox, 2010).
Example An individual may perceive cycling as healthy and environment-friendly and because of these beliefs, the individual adopts a positive attitude toward cycling. However, this does not automatically result in a travel pattern characterized by more cycling trips, because attitudes do not directly influence behavior (Van Acker, Van Wee, Witlox, 2010). In TRA, intentions intervene in the relationship between attitudes to that stimulus which are closely related to preferences (= express how a person desires or intends to behave)
Despite a positive attitude toward cycling, an individual considers himself physically unable to cycle and commutes by car. Behavior is guided by beliefs about the likely consequences of the behavior (attitude), beliefs about the normative expectations of others (subjective norm), and beliefs about the presence of factors that facilitates or hinders performance of the behavior (PBC). This combination leads to the formation of behavioral intention which is the determinant of behavior. Another example is that someone commutes by car because he or she thinks that no public transport services are available on the route toward work. PBC is inaccurate so we must distinguish between PBC and actual behavioral control (ABC). Behavior is thus assumed to be reasoned, controlled, or planned (Bamberg et al, 2003).
Although car-users might be motivated to switch to other travel modes, habits prevent them from doing so. Switching to other travel modes necessitate learning new routines. In order to do so, someone has to search and process information about the alternative travel modes. The costs associated with this may exceed the additional benefit of a better decision so that behavior is more a matter of habits or routines. Consequently, it is logical that behavioral decisions are not always well reasoned (Van Acker, Van Wee, Witlox, 2010). This means that planning daily recurrent activities such as working, or shopping is more a matter of routines or habits than of well-reasoned behavior.
Criticism retrieved from academic literature
TRA and TPB are criticized, because both theories assume that behavior results from rational decisions, but individuals are not constantly conscious of their behavior. The influence of habits for example where behavior becomes rather uncontrollable and automatic and occur without self-instruction contradict the establishment of reasoned action (Van Acker, Van Wee, Witlox, 2010).
Subjective norms and perceived control were the two more-important social cognitive constructs of TPB for public transport use, while car habit and factors influencing the resistance to change were neglected in the model (Nordfjaern et al, 2014).
Behavior is habitual so that it is, at least in part, under the direct control of the stimulus situation. Frequency of past behavior is thus an indicator of habit strength that can be used as an independent predictor of later behavior (Bamberg et al, 2003). More importantly, whether behavioral influences are reasoned or unreasoned, they are all affected by the individual’s lifestyle.
Source: Author’s illustration c. Social status and social surrounding
As already mentioned above, individuals' attitudes can be impacted by their social surroundings. In spite of the fact that attitudes may not be specifically noticeable by different people and consequently can't themselves be sources of social impact, a person will refresh his or her attitude in light of observing the travel mode choice of his or her social peers. This influence is strongly related to the desire of increasing one’s scial status through the association with affective and symbolic benefits of owning a car. For instance, Steg (2005) directed two investigations in the Netherlands about how different intentions are related with the level of car use. The results showed that private car use was most unequivocally identified with affective and symbolic motives, and not to instrumental intentions. Especially frequent drivers, male and younger respondents valued these non-instrumental motives for car use. The perceived benefit of cars depends on the lifestyle and social-spatial relations engaged by the user (Hiscock et al., 2002). This phenomenon is supported by Dittmar’s model on the meaning of material possessions (1992) which incorporates the main idea of “to have is to be”. According to Dittmar (1992), material possessions, represent instrumental values as well as symbolic values which are twofold: (1) the expression of the self and (2) a social-categorical expression indicating one’s social position or group membership. People continuously compare their opinions, behavior and possessions with those of others and they strive to be better off. People also try to present themselves in a way that is congruent with their self-image, which implies that people may get a sense of personal identity from driving their car. For a long time, car use was predominately explained through behavior models that focus on instrumental factors, such as its speed, flexibility, and convenience (Geller, Winett, & Everett, 1982; Golob et al., 1979; Recker & Golob, 1976; Steg, 1996). For some individuals, the car seems to represent a materialistic symbol for status in terms that individuals can express themselves by means of their car, because driving is thrilling, adventurous, and pleasurable. This leads to the fact that the utility of auto travel isn't just subject to its instrumental values, which means that people who are more convinced of the symbolic advantages of driving are less inclined to reduce their car use. Nilsson and Küller (2000) revealed that individuals who are emotionally connected to their vehicle drive their car more frequently and evaluate policy measures aimed at reducing car use as less acceptable compared to those who are less emotionally attached to
their car. This phenomenon requires urgent attention when using cluster analysis to be able to understand a population’s travel behavior.
d. Translating this into transport policies
Conquering the present impediments of policy and planning measures concentrating on a behavioral change towards more environmentally friendly mobility stand out amongst the most difficult errands for travel behavior researchers. Existing clustering approaches (e.g. Anable, 2005; Hunecke et al., 2007) provide a valuable ground for addressing issues more efficiently. Cluster analyses are coupled with the theory of planned behavior, as well as other psychological factors to describe how attitudes influence behavior and vice versa. If attitudes influence behavior, then behavior can be influenced via information campaigns and promotional messages targeted at people’s attitudes. But when travel behaviors influence attitudes, policy ought to focus and center around changing individuals' behavior directly, e.g. by focusing on travel costs via pricing policies (Krösen et al, 2017).
Since there are people who may not generally drive because they need to, but also drive by choice (Handy et al., 2005), it is important to push the policies that can decrease private transport reliance, the attractiveness of car use, and in addition, the need for driving, by giving other options (Beirao and Cabral, 2007). Such approaches include the improvement of public transport and elevating a shift towards slower modes (cycling or walking). Moreover, elevating measures to decrease the attractiveness of car use is likewise an obvious instrument (Gaerling and Schuitema, 2007). This implies that policies that aim to expand public transport ought to advance and promote its image, yet, public transport frameworks need to become more market-situated and focused. This requires a change in its service quality, which must be accomplished by fully understanding travel behavior and consumer needs and desires. Hence, it is fundamental to gauge the level of service to recognize the potential qualities and shortcomings of the public transport. This provides a sign to the public transport administration when assessing elective service improvements to enhance client satisfaction and expanding the market share. However, creating exact and valid measures of service quality is a complex task since it manages the psychological aspect of the consumer which can be very unpredictable. Car users have lower perceptions of public transport than public transport users, which means that public transport is actually better than they think. Thus, one strategy to attract users could
be (both) improving public transport image and providing more information about the transport system to be able to target the market segments that are most motivated to change and willing to reduce frequency of car use (Beirao Cabral, 2007). But, this does not automatically mean that policy measures based on conclusions from instrumental-reasoned research models will be effective. Despite increased fuel prices and taxes, growing congestion and barring of cars from city centers, car use is still increasing. It is important to appreciate that changing people's attitudes will not automatically lead to behavioral change. In addition to that, if policy makers do not act on dissonance with respect to public transit (e.g. by enticing more people onto transit through lower fares and improved service), people will generally adjust their attitudes towards this mode downwards. It seems that regular use of public transport is thus necessary to uphold a positive attitude towards the use of this mode. Many other studies about acceptability of various transport policy measures conducted by Schade and Schlag in 2003 have also revealed some factors that influence people’s car use including feelings of responsibility, personal norms, perceived effectiveness, trust in the co-operative behavior of others and social value orientation. When service quality needs to be improved, it can mostly be achieved by a clear understanding of travel behavior and consumer needs and expectations. Moreover, habitual behaviors are very difficult to change. The best way of doing this would be to block or punish their execution (e.g. by physical or financial measures). In transport modelling the challenge is how to incorporate habitual choice as well as how people acquire and use knowledge of their environments and of the transportation system (Gärling & Axhausen, 2003).
III. Conceptual Framework
The conceptual framework in this research is connected to the research purpose and aim which is to enrich the knowledge for policy makers to entice more suitable transport policies based on the understanding of people’s attitudes and mode choice behavior. With this being said, the starting point for concepts used in this research is essentially translated by the ‘theory of planned behavior’ which was created by Ajzen (1971). This theory incorporates the relationships between attitudes, social norm and perceived behavioral control to be able to evaluate the motivations and constraints towards the intention to use public transport. Moreover, a special focus is made on moral norms, environmental attitudes, and symbolic functions of car use in relationship with the individual’s identity to be able to see how much of
an impact the social surrounding makes when forming attitudes towards car driving. The concept of material possessions as defining one’s social status was created by Dittmar (1992) and is incorporated in this research to be able to evaluate how much the symbolic attachments of car-use influence the ‘potential switchability’ of car drivers to use public transport. This conceptual framework depicts the idea that creating socially clustered environments based on their attitudes towards transport mode choices by using the above-mentioned variables enriches the understanding to implement better suited transport policies.
IV. Problem Statement
Whilst many cities have undergone great progress in reaching the ‘peak car’ phenomenon in which the growth of car ownership and car-use has come to an end, there exist many people whose attitudes towards car driving get influenced in socially clustered environments and eventually remain car-dependent although they want to use sustainable transport modes more.
V. Research Question(s)
The choice of research question(s) was inspired by researchers such as Jillian Anable (2005) who stated that understanding people’s travel behaviors will grant a step forward to understand their mode choice and can then be a contributor of working towards promoting better public transport use.
How do socially clustered environments influence people’s attitudes towards transport mode choice and their travel behavior and how is this taken into account by transport policies?
a. Sub-Questions
1. How important is social clustering in explaining mode choice?
2. How are the attitudes of people with different travel behaviors formed? 3. How is this (not) taken into account in policy?
The following section will explore the understanding between attitudes and travel behavior by taking the City of Dubai in the United Arab Emirates (UAE) as a study area, so that people’s preferences, attitudes and beliefs can be observed. This will be done by first applying a quantitative approach using social cluster analysis, followed by a qualitative approach in form of follow-up interviews and document analyses with the expectation that:
(1) Using attitudes towards certain modes of transport, the environment and sustainability for developing segmentation approaches enriches the analyses of sustainable travel behavior and produces useful segments for evaluating daily mobility and corresponding transport policies.
(2) In contrast, people’s travel behavior of the same segment does not always show the same attitudes towards transport mode choices.
The resulting clusters are compared in view of different individual status variables (e.g. age, gender, income, car access, and housing type) and behavioral variables (e.g. mode choice for daily travel), to characterize and compare the different user groups and highlight the differences between the segmentation approaches.
VI. Research Case Selection
Dubai is the largest city and one of the seven emirates belonging to the United Arab Emirates (UAE) in the Middle East with one side of its city facing the Persian Gulf, and the downtown set against the water (Map 1 and 2). Dubai has experienced an immense unprecedented growth over the last few decades due to great oil revenues and due to the vision and encouragement of the government to transform the city from a regional business, financial and leisure hub into a global center (Lans, 2008). After the oil discovery in 1966 major infrastructure and urban development projects were planned. The cities geographical disposition incorporates modern skyscrapers rising in clusters, artificial islands that rise from the sea and neighborhoods containing mixed-use developments for office and residents as well as new gated communities accommodating individuals and families coming in from all over the world (CBS, 2007). Further investment levels accelerated with these prestigious developments to encourage urban
growth as well as attract tourists from all around the world. Dubai’s response to housing needs and further developments has been through gated communities (Map 3, 4, and 5) of which the majority were built around the 1980s and 1990s with the highest return rates for investors. Their popularity further increased as more and more foreigners purchased housing due to the 2002 decree allowing non-residents to own freehold properties regardless of national origin (Greenhouse, 2015). However, with the occurrence of gated communities, a social and spatial stratification occurred in the city, because of a lack of deep social mixing when it comes to housing location, lifestyle and leisure activities (governingdubai.com, 2017)
Moreover, as Dubai represents one of the fastest growing cities in the world, it is experiencing a drastic surge in automobile ownership. This city is chosen for this research as an extreme case study area because although the city is undergoing sustainable growth efforts (Chaudhry, 2012), about 700 new vehicles are being registered on a daily basis and one out of two persons owns a vehicle today. By 2020, when Dubai will have more than 5 million inhabitants (today there are about 3 million), the five million car trips currently being generated daily is anticipated to increase to 20 million per day (Chaudhry, 2012). Traffic congestion has become a part of everyday life in Dubai and is a growing problem threatening the overall quality-of-life and economic prosperity of the region. Even though Dubai is the most advanced Middle Eastern city when it comes to developing public transport systems, there still exists a huge number of cars on the streets. It is therefore interesting to study this area with such an extreme impact in both fields, which is car ownership and quality of public transport. Moreover, in terms of evaluating socially clustered environments, it is very suitable to evaluate people in Dubai because this city consists of people from a vast variety of different nationalities, backgrounds, social classes that are from all around the eastern and western world which all have different experiences, values, preferences, beliefs and attitudes towards transport mode choices. There is also an uneven distribution between men and women with an overall population that consists of three males for every female living in the country (The National, 2017). Dubai is unusual in that the majority of its population is comprised of expatriates of which most are low income workers from countries such as India, Pakistan, Bangladesh, the Philippines, Iran, Egypt, Nepal, and China. There are fewer, but significant, numbers of professional workers of which the majority come from countries such as the UK, South Africa, the US, Canada, France, Germany, and Australia (Dubai-Online, 2014). This study therefore focuses on how to best describe a population based on travel behavior, socio-demographical
attributes and attitudinal behavior to be able to understand the underlying reasons of what people think of these two transport modes.
Map 1: Location of UAE Source: World Atlas
Map 2: Location of Dubai Source: Google Maps
Map 3: Emirates Living Location Source: Google Maps
Map 4: Emirates Living Master Plan Source: www.emaar.com
Map 5: Emirates Living – Aerial View Source: Google Images
Chapter 2 – Research Design
I. Methodology
From the 6 different available research designs (Bryman, 2012), a descriptive research design was proposed for this study as all of the sub-questions incorporated a determination and identification of already existing phenomena to be able to shed light through a process of data collection to describe the characteristics of the situation in Dubai more precisely than is possible without employing this method. In its essence, this research describes the behavior and attitudes of its chosen sample population towards different modes of transport to see how they are formed and what implications this has for transport policies. The research design is three-dimensional in terms that it firstly incorporates a quantitative approach via survey distribution, followed by a qualitative approach via interviews with the same respondents of the survey and lastly, a desk research in form of a document analysis of policy interventions. The interviews are composed of an ethnographic methodology which allows multiple realities and interpretations given by the respondents (Fetterman, 1989). The researcher cannot be completely objective in conducting research and acknowledges that interviewing research participants can reveal the nature of constructs, especially with a varying number of participants and their different outlooks. It is a little optimistic to attain complete neutrality in the research field, however, the aim is that each individual interviewed can have their viewpoint considered as an entity in its own right, which is not influenced by the comments of others.
II. Methods
a. Sub-Question 1 – Questionnaire Development
In spite of the efforts to understand public transport usage and eliminate unsustainable mobility patterns, a cross-sectional survey was constructed for sub-question 1 to understand and compare motivations, constraints and attitudes of the ‘potential switchability’ of different types of residents. These residents were distinguished by factors such as owning a car or not or living in a gated community in Dubai or not, which were important to distinguish because it changed
people’s travel behavior instantly due to physical constraints. To be able to create a socially clustered environment based on people’s motivations, constraints and general outlooks towards public transport modes, this survey comprised an expanded version of a psychological theory of attitude-behavior relations, namely the ‘theory of planned behavior’, as well as a representation of each of the respondent’s perceptions, identity, preferences, worldviews, exposure to its social surrounding (translated by social norm), habits and overlapping attitude statements. The cluster analysis is based on these attributes because individuals travel and adapt their travel behavior based on these specific variables.
Attitude theory was used to develop the questions because using a priori classifications to segment populations based on demographic variables or simple behavioral measures would have oversimplified the structure of mode choice behavior which would have then made it difficult to answer the second sub-question if the first sub-question didn’t include any attitude statements. It also predicts the occurrence of a particular behavior, provided that behavior is intentional. The survey included statements about the person’s behavior and involutional control to be able to extend the TPB and get a more precise overview of whether people intended to perform a behavior or whether their behavior became a form of habit that is unintentional as described in the theoretical framework.
The survey corresponds to a version of a survey created in 2003 by Anable (2005) with the focus on day trip travel to leisure attractions by segmenting 666 visitors to National Trust properties in the northwest of the UK. The structure and outline of the current survey questions are comprised below in Table 1. The respondents answered 4 questions about their travel behavior, 11 questions about identifying their attitudes towards driving a car, 7 questions about identifying their attitudes towards taking public transport and eventually 9 questions defining their socio-economic background. The second and third section of the survey were answered via Likert scale ranging from 1 to 5 with higher scores pertaining to more favorable views of the environment or ‘anti’ car/pro-public transport sentiments.
Table 2 - Outline of Survey Questions
Sections Information to collect Description
Travel Behavior
Travel information Access to a car;
Frequency of car driving per week; Kilometers traveled by car;
Frequency of taking Public Transport
Socio-psychological factors 1
Attitudes towards driving a car Car-Dependency;
Symbolic function of the car; Attachment to the car; Green Identity; Moral Norm; Habit;
Congestion as an influence; Positive attitude towards a car; Efficacy;
Enjoyment of car driving; Social status;
Socio-psychological factors 2
Attitudes towards taking Public Transport
Perceived Behavioral Control; Sacrifice to use Public Transport; Social Influence;
Social Norm;
Negative effects of the car; View of Nature;
Environmental worldview
Socio-economic demographics
Personal information Gender; Age; Nationality; Area of Residence; Level of Education; Income
Source: Author’s illustration
The questionnaire was sent out in two phases. Firstly, the survey was distributed to private contacts of my social network which consists of family members, friends, and work colleagues on Facebook and LinkedIn. The second phase is made up of a random distribution process to various online community platforms to target a great diversity of people based on their socio-demographics and geographical disposition to grant representativeness. Examples of these community platforms were (1) Dubai Marina, Springs, Meadows, JBR, Emirates Hills, (2) Market Place Jumeirah Village Circle, Springs Dubai, and (3) Meadows, Lakes, Springs Community. After the collection procedure was completed, structures within the data were identified to organize them into meaningful groups. To be able to determine the appropriate number of clusters a k-means analysis was performed in the SPSS software. This method aims to partition ‘n’ observations into ‘k-clusters’ in which each observation belongs to the cluster with the nearest mean. This is done with an initial set of centers to then modify them until the change between two iterations or repetition of combining different clusters is small enough.
After the initial cluster centers have been selected, each case was assigned to the closest cluster, based on its distance from the cluster centers. After all of the cases have been assigned to clusters, the cluster centers are recomputed, until the maximum number of iterations (10 by default) is reached. In general, for most clustering exercises somewhere between two and five clusters is ideal and easier to theorize. 5 clusters were identified of which its statistical analysis was computed properly and is documented and described in detail in the Results section. The SPSS outlook file can be accessed upon request to the author.
b. Sub-question 2 – Follow-up interviews
To be able to answer the second sub-question, it is important to analyze how the attitudes of the resulting clusters were formed and how they vary from each other and why. 8 Follow-up questions (Table 2) were sent out via email to the same respondents who filled out the survey to be able to cross-reference the data of the same context and the same people. These questions consist of what people thought about owning and driving a car and taking public transport in Dubai to enrich the knowledge of their ‘potential switchability’ and to also distinguish between reasons that were in and out of the user’s control (PT is inaccessible vs. PT is possible but low desire to drive). As attitudes were defined in the theoretical framework as having positive, negative or mixed evaluative responses to some stimuli which consist of the cognitive, affective and conative/behavioral aspects, the idea was to correlate to that theory and receive and establish a detailed description of what attitudes people had and how the attitudes of the following four categories were formed: (1) Attitudes towards public transport, (2) Attitudes towards car-driving, (3) Influence of an individual’s social surrounding to own a car, as well as in general in Dubai, and finally (4) How people respond to the intentions to use public transport. These 4 categories are important to be able to compare the feelings towards both types of transport. Moreover, instrumental, as well as symbolic factors of car ownership were evaluated to define its importance. The symbolic aspect of the follow-up questions is again based on Dittmar’s theory (1992) that materialistic possessions define a person’s status with the idea that “to have is to be”. These four categories also again correspond to the ‘theory of planned behavior’ as the respondents’ answers shed light about their attitudes, social norm, perceived behavioral control and intentions to take public transport. Moreover, a comparison of responses was made to fetch the similarities and differences between clusters to be able to strengthen the validity of the findings of the cluster analysis and to see how important social
clustering is and what exactly this says about the population in Dubai. The transcripts of the interviews can be accessed upon request to the author.
Table 3 - Follow-up Questions – Qualitative Approach
1 What are the reasons for why you drive a car (or not)?
2 How do you feel about your car driving habits (fun, stressful, annoying, etc.) and how would you evaluate the importance/attachment of the car in your life?
3 What do you think about how your attitudes towards the car evolved in the past few years or from before you moved to Dubai until now?
4 How do you think your social surrounding contributed to that? In other words, do your family, friends, and/or colleagues play a role in your decision to drive a car (or not)? If yes, why?
5 What do you think about cars as being a contributing factor to defining social status/social rank?
6 Do you think that this is a problem in Dubai (refer to question 5)? If yes, why?
7 How would you describe your possibilities to use the car less? If you are already using the car less by taking public transport, what are the reasons for doing so?
8 Are there any reasons why you don’t want to take public transport? If yes, what are they?
Source: Author’s illustration c. Sub-Question 3 – Document Analysis
The findings of Sub-question 1 and 2 brought forward multiple challenges that Dubai is exposed to in terms of how likely people are to switch to more sustainable transport modes and how their attitudes differ from each other. These challenges were tested by checking how they are incorporated in policy and/or initiatives to see how Dubai is trying to improve its transportation sector. A special focus was drawn on the influence of Dubai’s social surrounding by specifically looking at how Dubai correspondents to lowering the symbolic attachments to the car of the general public. Table 11 was then created based on the profiling procedure to evaluate the ‘potential switchability’ of clusters 1 and 5 (‘status-oriented automobilists’ and ‘aspiring environmentalists’ – profiled in results section) towards more sustainable modes of transport, as well as which constraints exist about using alternatives and what factors and policy options may be considered indicative of susceptibility to reduce car use or the main obstacles to change for each of the two cases. The information in the table were then compared to how Dubai actually implemented or is going to implement them in reality by looking at RTA’s policy documents and initiatives.
The two clusters were chosen from both extreme ends with the intention to have an easier visualization about which transport policy or initiative suits which cluster in a better way. The
policy documents that were analyzed were retrieved from Dubai’s Roads and Transport Authority (RTA) which is a governmental institution established in November 2005 and is responsible for planning, overseeing and maintaining the roads, rail and marine transportation networks in the city. RTA comprises five agencies which are the (1) Traffic and Roads, (2) Public Transport, (3) Licensing, (4) Rail, and (5) Dubai Taxi Corporation. All documents were found on RTA’s official website which was made available to the public (link mentioned in reference list). There were several criteria for including and excluding projects which consist of including documents when ‘promotional messages’ about public transport and against car driving were mentioned with the aim to increase and decrease ridership. Moreover, documents were chosen based on the intention to reduce congestion issues which automatically results in lowering car usage and what solutions are suggested in terms of promoting alternatives to the car while looking at whether the symbolic aspect of car usage is taken into account. Hard infrastructural measures were not considered in the analysis in order to have a more detailed approach of the soft transport interventions.
7 out of 17 documents were analyzed in total. The reasons are that there exist three annual reports from the years 2014, 2015 and 2016 of which the most recent one was chosen to grant accuracy of the current situation. All 4 statistical reports were excluded because they only talk about general attributes of how the number of types of public buses, metro trips, routes, and accident rate increased between 2014 and 2016, which was not the focus of this study because they don’t mention policy interventions. The case studies of the ‘Enterprise Command and Control Centre’ (EC3), ‘Dubai Water Canal’, ‘Etihad Museum Project’, and the ‘Agility state
in the Middle East’ were also excluded as these documents consist of case studies about different projects in Dubai which again do not incorporate soft transport policies. The ‘RTA corporate Safety and Environmental Sustainability Policy’ and the ‘Booklet of fines’ were included to evaluate whether RTA deals with traffic safety issues which were addressed and complained about in the first two sub-questions. However, the list of fines was retrieved from the ‘Khaleej Times’ online newspaper site because there it was translated into English. It was not always easy to find reports and data concerning Dubai on the websites belonging to the government. Information about the whole UAE was easier to find. In Dubai’s statistical portal it was common to find incomplete data, ancient data or information presented in Arabic. Private companies, magazines, newspaper articles and international institutes had plenty of information related to this paper and they are therefore used as sources in some cases. For
example, the ‘RTA Sustainability Report 2016’, ‘RTA Mobility Management Plan’, ‘UAE Vision 2021’, and ‘Smart Dubai Strategy’ were translated into English by such websites and were included with a special focus for this research because they demonstrated initiatives and more or less a summary of the most recent soft transport policies that are important to evaluate what Dubai has in mind to make its transportation sector more sustainable.
III. Ethical Considerations
Any common form of research needs to incorporate the ethical implications of the project, which will also be done here. The final selection of informants was based on their willingness to participate in this research project. Personal details were not required when filling out the surveys, however, email addresses were collected only from the participants who agreed to be interviewed in the second round. It was explained that the research is used for academic purposes and that all answers will be kept in full confidentiality. Therefore, their names are not mentioned in this paper and are coded by numbers ranging from 1-14 as there are 14 respondents respectively. Only their age, gender, car access and type of housing location were mentioned to be able to get a better insight of their personal background. Interviewees could freely choose to decline to answer any question, stop the interview, and/or redact any statements of the follow-up interviews. As there was always a low risk of harm as a consequence of the interview, a consent letter was not necessary, but a final draft of the analysis of their responses will be sent to them for their records.
Chapter 3 - Results
The results chapter of this thesis provides a detailed outcome of the research conducted in Dubai in April and May 2018. The findings are reported based upon the methodology applied to gather and analyze information. The following consists of a descriptive as well as argumentative approach of results to be able to exclude any bias of the researcher’s interpretation. The intention here is to simply represent the results with discussion and engagement between the findings and wider literature taking place in the following chapter.
I. Getting to know the audience
Table 4 - Socio-demographic distribution of respondents Total (N=135) Frequency % Gender Male 91 67.4 Female 44 32.6 Age Under 25 24 17.8 26-35 56 41.5 36-45 36 26.7 46-55 14 10.4 56 and older 5 3.7 Level of Education High School 7 5.2 Undergraduate 31 23 Postgraduate 79 58.5 PhD 9 6.7 Other 9 6.7 Monthly Income < 15,000AED 43 31.9 15,001-30,000AED 47 34.8 > 30,001AED 36 26.7 Access to a car Yes 123 91.1 No 12 8.9 Frequency of car driving Never 9 6.7 1 day 4 3.0 2 days 3 2.2 3 days 5 3.7 4 days 4 3.0 5 days 7 5.2 6 days 7 5.2 Everyday 96 71.1 Weekly kilometers traveled by car 0-150km 49 36.3 151-300km 51 37.8 > 301km 35 25.9 Frequency of taking public transport Never 59 43.7 Once a month 32 23.7 1-2 times a week 20 14.8 3-5 times a week 12 8.9A total of 521 surveys were distributed, resulting in an overall response rate of 26%. Table 3 provides further insight of the socio-demographic constellation of the total number of respondents. 67.4% of the respondents were male and 32.6% were female. The uneven distribution of gender can be explained with the above-mentioned argument that typically 1 out of 3 women live in Dubai. The majority of respondents are aged between 26 and 35 and 58.5% obtain a postgraduate degree which makes them belong to a higher educated working class. The ratio between male and female respondents is uneven with 75% of men living in gated communities, and 63.2% of men not living in gated communities. In this research, the majority of population is composed of expats from the western world as retrieved from the answers of the surveys. The reason for the unbalanced gender proportion is mainly due to the large number of foreign male workers, most of whom are middle-aged men that do not relocate with their family. This also explains why the age distribution in both cases is concentrated around the Age groups of 26-35 and 36-45. The level of education is almost similarly distributed among both resident groups.
II. Establishing Number of clusters
Everyday 12 8.9
