*2.2. Data and Methods*

Data used in this study were collected via a face-to-face survey which was distributed to Londoners by the fieldwork conducted from June to August 2018. We utilised a random sampling method [39], similar to the approach used by Cao [40], and Cao and Hickman [41–43]. Prior to each interview, participants were asked if they were permanent residents in London and aged 17 or over. The targeted sample comprised those aged 17 or over since it is only license holders from this cohort who are legally permitted to drive. Only those who fulfilled these requirements were accepted to participate in this study and complete the survey. With a view to filling the knowledge gap in the literature regarding public attitudes towards the forthcoming ban on the internal combustion engine (ICE) vehicles (i.e., diesel and petrol), the survey sought to: (1) Collect insightful information about the Londoners' level of awareness about the harm caused by diesel vehicle emissions; (2) establish their views and attitudes regarding the forthcoming ban; and (3) discover whether they would be willing to shift to EVs. We argue that undertaking these investigations using a bottom-up approach is important because driving diesel vehicles can be cost-e ffective. As such, diesel vehicles use slightly less fuel than petrol ones, which o ffer cheaper running costs. Thus, diesel vehicles might be perceived as a more convenient option. The survey was piloted to make sure that all the questions were clearly worded

without ambiguity and to avoid problems related to a low response rate. It was checked by a small number of colleagues and peers (n = 18) at the University of Westminster and University College London. As a result, 139 valid responses were collected. The survey consisted of eleven multiple-choice questions, while the twelfth was an open question designed to allow respondents to provide more detailed comments, which have been used for qualitative analysis in this research.

This section outlines the survey questions that were distributed to Londoners: (1) Since poor air quality a ffects the quality of life referring to the state of wellbeing in terms of health, comfort, and happiness [44], the first survey question sought to ascertain the Londoners' levels of satisfaction in respect to air quality. This question sought to discover to what extent they agreed with the following statement: "I am satisfied with the air quality in London". (2) The second question in the survey aimed at identifying their level of awareness about the harm caused by diesel vehicles since it is argued that the level of awareness influences people's attitudes towards environmentally-friendly products [45]. (3) The next survey question was consequently designed to find out how Londoners would feel about purchasing a diesel vehicle having been informed about some of the health problems that diesel vehicles cause. Respondents were asked: "Emissions from diesel vehicles can cause premature deaths and many diseases such as asthma and lung cancer. Would you, as a consequence of the statement above, still be willing to purchase a diesel vehicle?" (4) In response to the UK government's announcement that ICE vehicles will be banned from the beginning of 2040, the fourth survey question involved a hypothetical scenario whereby respondents were invited to envisage that the governmen<sup>t</sup> wished to enforce a ban on diesel vehicles from the start of 2019. The purpose of this survey question was to determine how quickly the Mayor of London should introduce the policy. (5) The fifth question in the survey asked respondents about the type of vehicle that they owned (i.e., diesel, petrol, EV, etc.) so as to provide an insight into the proportion of car users in addition to the vehicle types. (6) The sixth survey question was essential to gain insights into how the diesel and petrol car drivers would act if the governmen<sup>t</sup> incentivises EVs or bans ICE vehicles completely. Therefore, only respondents who represent the car-driving cohort were required to complete this question. Respondents were asked to what extent they agreed with the following statement: "I would replace my diesel/petrol vehicle with an electric one." (7) The purpose of the seventh survey question was to ascertain whether participants had become sensitised towards the environment after they had been exposed to information about the potential harm caused by diesel vehicles. Respondents were asked about the kind of vehicle they would purchase if they were going to purchase one today. Lastly, the eighth, ninth, tenth, and eleventh questions in the survey sought to obtain information about the participants' characteristics in terms of gender, age, highest educational attainment, and main transport mode for work [41,43,46–48], all of which are presented in Table 1 below. The final survey question was an open question that asked respondents to provide any suggestions and recommendations on how to improve London's air quality in relation to the diesel vehicle emissions.


\* "Others" is comprised of working from home and/or not currently employed.

The present study utilises a non-parametric statistical testing method known as the Kruskal–Wallis H [49]. The test static, H, is calculated by the following formula:

$$H = \frac{12}{N(N-1)} \sum\_{i=1}^{k} \frac{R\_i^2}{n\_i} - 3(N+1) \tag{1}$$

where:



The Kruskal–Wallis H Test (KWT) is a rank-based approach to the one-way ANOVA, which determines whether there are statistically significant differences between three or more groups for a dependent variable [50]. The KWT assumes that: (1) The sample population is drawn at random; (2) the observations of the samples are independent; and (3) the scale of measurement of the dependent variable should be at least ordinal [50]. This non-parametric statistical test has been used by many researchers [51–53], most commonly in behavioural sciences [54]. It is utilised in this study with the aim of gaining greater insight into the factors that may influence the public acceptance of the ban on diesel vehicles.

Perhaps a minor limitation of this method is that it indicates if there is a significant difference between groups but does not indicate which groups are different. Therefore, a qualitative analysis was conducted to overcome this limitation. In this paper, a mixture of quantitative and qualitative approaches is used [47]. The purpose of doing this is to minimise the disadvantages of each individual approach while maximising the benefits of both approaches combined [55].
