Investigating how the public acceptance of autonomous vehicles evolve with the changes in the level of knowledge: A demographic analysis

Abstract

Autonomous vehicles (AVs) are expected to provide various advantages such as improved mobility, increased comfort, and reduced traffic accidents. However, the deployment of AVs is contingent upon the public’s attitude, which in turn may affect their consequences. The relationship between public knowledge about AVs and attitude has been debated. While some studies indicate a positive correlation between knowledge and optimism, others suggest a negative association. The present study aims to evaluate the association between public knowledge, attitude, and AVs in the US. A questionnaire survey was conducted between June and November 2022, yielding 5778 complete responses from various regions of the US. Data were analyzed to estimate the public attitude and level of knowledge by region. Findings revealed a negative shift in public attitude towards AVs with increased knowledge. Specifically, a 1% increase in knowledge was associated with a 0.65% reduction in interest, a 0.68% decrease in trust, and a decline of $2466 USD in willingness to pay for AVs, as well as a 0.56% increase in concerns about traveling in AVs. In addition, further analysis was conducted to understand how the public attitude of different demographic groups evolves with the level of knowledge. Furthermore, the results of this study were discussed in light of the introduction of the automobiles showing a lot of similarities in the public attitude towards the introduction of both the automobiles (more than 100 years ago) and AVs (now). In addition, the results were discussed in light of the theory of diffusion of innovation and the Gartner Hype Cycle which are used to explain how the public reacts to innovations.

Keywords:

• autonomous vehicles
• knowledge levels
• public acceptance
• regression analysis
• demographic analysis
• the introduction of automobiles
• theory of diffusion of innovation

1. Introduction

In recent years, autonomous vehicles (AVs) or self-driving cars have emerged as a potential solution for various transportation issues, including traffic safety (Othman, 2021a). Many argue that AVs have the potential to reduce traffic accidents by 90% by eliminating human error, which is a leading cause of accidents. While safety is a significant advantage of AVs, they also offer other benefits such as optimizing traffic flow to minimize energy consumption and emissions (Bansal et al., 2016; Othman, 2022b, 2022c), improving accessibility for people with disabilities or limited transportation options, and increasing productivity by allowing passengers to engage in other activities during travel time (Antov et al., 2012; Bansal et al., 2016; Othman, 2022a, 2022b, 2022c). However, there are also potential risks associated with AVs, such as an increase in vehicle kilometers traveled due to empty trips, leading to increased congestion on transportation networks (Berrada & Leurent, 2017; Othman, 2020). While there has been significant discussion about the technological aspects and implications of AVs, less attention has been paid to public attitudes towards AVs and the factors influencing those attitudes (Newcomb, 2012; Othman, 2021b). The public’s resistance to new technology, as seen historically with the introduction of the automobile, may present a significant barrier to the deployment of AVs (Ladd, 2008; Norton, 2019; Winton, 1930). As such, understanding public attitudes towards AVs is crucial to ensure their successful deployment.

Numerous studies have examined how people feel about autonomous vehicles (AVs) in different countries and over time (Eden et al., 2017; Janatabadi & Ermagun, 2022; Jardim et al., 2013; Jing et al., 2020; Luo et al., 2022; Rezaei & Caulfield, 2020; Wu et al., 2019; Yuen, Chua, et al., 2020; Yuen, Wong, et al., 2020). While these studies have contributed to understanding public attitudes towards AVs, they have typically focused on attitudes in a specific location. They also often explore how demographics impact attitudes towards AVs, with some studies suggesting that people with more knowledge about the technology tend to be more optimistic (Anderson et al., 2014; Piao et al., 2016; Richardson & Davies, 2018; Wintersberger et al., 2016). However, the public’s perception of AVs is not static and can change over time and across countries. For instance, studies by the American Automobile Association (AAA) in 2016 and 2019 show that the public’s fear of AVs increased from 63% in 2016 to 71% in 2019, even as awareness of the technology grew (American Automobile Association, 2018). Similarly, other studies have found that as knowledge of AVs increases over time, public opinion tends to become more negative (Lienert, 2018; Othman, 2021b, 2023). This has led to debates about the relationship between knowledge and attitudes towards AVs. While some studies have found that those with more knowledge about AVs tend to be more optimistic (Charness et al., 2018; Chikaraishi et al., 2020; Kaye et al., 2021; Nordhoff et al., 2019; Pettigrew et al., 2019), others have found the opposite (Lienert, 2018; Othman, 2021b, 2023). To better understand this relationship, this study will conduct a detailed analysis of the relationship between knowledge of AVs and public attitudes towards them specifically for residents of the United States. In addition, previous studies have shown that the level of knowledge of AVs is a major factor that affects their attitude. In general, these studies explored the impact of the public’s misconceptions about AVs or some of their functions. In general, the results show that the majority of the public has some misconceptions about AVs. On the other side, the results suggest that people who have misconceptions about AVs are more positive towards them than people who do not have misconceptions about AVs (Du et al., 2022; Lipson & Kurman, 2016; Liu et al., 2022; Nikitas et al., 2019; Smith, 2014). Thus, these studies concluded that people who are more wrong about AVs are more positive to adopt the technology (Du et al., 2022; Liu et al., 2022). Thus, these results indicate that the level of knowledge about AVs will affect the public’s opinions and their purchasing behaviour. Thus, it is crucial to understand the relationship between the level of knowledge about AVs and the public’s opinions.

This study aimed to investigate the public’s level of awareness and attitude towards autonomous vehicles (AVs) in the United States. To achieve this, a questionnaire survey was conducted in 2022, focusing on different states in the US. The survey covered various aspects, such as the public’s level of interest, trust, concern, and willingness to pay more for AVs. The study also sought to establish the relationship between the public’s level of knowledge about AVs and their attitude towards them. The paper is structured into four sections, with the first providing an introduction and background information on the topic. The second section outlines the methodology used, while the third section analyzes the data and draws conclusions about the relationship between knowledge and attitude towards AVs. Finally, the fourth section presents the main conclusions of the study. Overall, this research provides valuable insights into the public’s attitudes towards AVs, as well as the factors that shape their perception of this emerging technology.

2. Theory of the diffusion of innovation

The Diffusion of Innovation (DOI) theory endeavors to elucidate the mechanisms, reasons, and speed at which novel ideas and technological advancements proliferate. The theory was brought to prominence by Everett Rogers, a communication studies scholar, through his publication, Diffusion of Innovations, which first appeared in 1962 (E. M. Rogers, 2003). Rogers posits that diffusion occurs through the communication and transmission of an innovation across a social system’s participants over time. The DOI theory draws on diverse sources and encompasses a range of disciplines. The DOI theory has identified five main elements that affect the deployment of new innovations. These elements are the innovation itself or the technology, adopters, communication channels, time, and a social system (E. M. Rogers & Cartano, 1962). In addition, the theory explains that the process of diffusion occurs through a five-step decision-making process as summarized in Figure 1 (M. F. Rogers, 1974). These steps are the knowledge or awareness of the innovation, persuasion, decision, implementation, and confirmation. In addition, the theory shows that the public knowledge or awareness of the technology is a critical factor for the success of innovation at the early stages as this is the first step in the decision-making process. In addition, any innovation passes through these five steps in order to succeed so failure in the first stage (knowledge) means that the public will not proceed to the following stages, resulting in diffusion failure. Thus, failure of innovation usually occurs with high levels of public knowledge about the failures or issues of the innovation. In the initial stage of the diffusion process, the individual is exposed to an innovation but lacks adequate information about it, given the low penetration rate of the technology. At this stage, the individual has not yet developed a desire to seek further information about the innovation (Newell et al., 2001). As a result, the public knowledge about the innovation is critical for the success of any innovation. Currently, AVs are not really deployed in the market so the success of the technology mainly depends on the level of public knowledge of the technology, its benefits, and issues. Thus, it is critical to analyze the impact of the level of knowledge of AVs on the public’s attitude in order to understand the future of deployment of AVs.

Figure 1. Illustration of the five-step decision-making process for the DOI (showing the state of AVs on it).

The DOI theory showed that the adoption of an innovation follows an S curve when plotted over time. This S-curve is called the S-curve of innovation. The S-curve represents the market share of emerging technologies, while the adoption rate curve demonstrates the pace of adoption as presented in Figure 2. This curve identifies five categories of adopters, including innovators, early adopters, early majority, late majority, and laggards. Innovators are willing to embrace new technology despite the risk, possessing high social status and financial resources to support adoption. Early adopters also have high social status, educational levels, and financial security. The early majority has average social status, relying on early adopters for technology education. The late majority is skeptical and adopts technology only after the average participant. Laggards are the least likely to adopt new technology (Fisher & Pry, 1971; Kinnunen, 1996). In general, previous innovations followed the S-curve and it is also important to understand how similar technologies were deployed. Especially, previous studies have shown that the public usually resists new technologies in the early stage. For example, 100 years ago, the public was hostile towards cars and there was a large debate about the introduction of Automobiles (Ladd, 2008; Norton, 2019; Winton, 1930).

Figure 2. Illustration of the innovation adoption curve and market share (the S-curve) showing the state of AVs on the curve.

3. The introduction of automobiles and lessons learned

The introduction of automobiles in the late 19th and early 20th centuries was a major technological breakthrough that revolutionized transportation and had a profound impact on society. In the early days of the automobile, there was a great deal of skepticism and resistance from the public. Some people saw the automobile as a dangerous and unnecessary luxury, while others worried about the impact it would have on traditional forms of transportation, such as horses and trains. There were also concerns about the safety of automobiles, as early models had a reputation for being unreliable and prone to accidents (Geels, 2005). In fact, some cities even passed laws limiting the speed of cars to as little as 4 miles per hour. In general, there was a lot of resistance and skepticism towards this new technology. Some of the reasons for this resistance included (Lupkin, 2001; Mattioli et al., 2020; Mercure, 2015; Parsons & Vigar, 2018):

1. Fear of the unknown: The automobile was a new and unfamiliar technology, and many people were afraid of what it might bring. Some worried about the safety of cars, while others were concerned about the impact they would have on society.

2. Opposition from established industries: The automobile posed a threat to established industries such as horse-drawn carriage makers and public transportation companies. These industries lobbied against the automobile and sought to restrict its use.

3. Infrastructure challenges: The early automobile lacked a comprehensive infrastructure, such as paved roads and gas stations, which made it difficult to travel long distances.

4. Cost: In the early days, cars were expensive and were seen as a luxury item that only the wealthy could afford. This made them inaccessible to many people.

Despite these concerns, the popularity of the automobile continued to grow. As cars became more reliable and affordable, they gradually replaced horses and other forms of transportation. As more and more people began to own cars, they became an essential part of modern life and transformed the way people lived and worked (Mercure, 2015). Today, the automobile is an essential part of modern life, and it is difficult to imagine a world without it. The rise of the automobile in the early 20th century brought about significant changes to society, including changes in the way people lived, worked, and traveled. One major effect of the automobile was the growth of suburbanization. As cars became more affordable and reliable, people could travel greater distances more quickly and easily, allowing them to live farther away from the city center. This led to the development of suburbs, where people could enjoy the benefits of both city and country living. The automobile also had a profound impact on the economy (Lupkin, 2001; Mattioli et al., 2020). The rise of the automobile industry created jobs and stimulated economic growth, and the widespread use of cars created new opportunities for businesses such as gas stations, repair shops, and motels.

One of the key factors that helped to break down public resistance was the development of mass production techniques that made cars cheaper and more widely available. The Ford Motor Company, in particular, was instrumental in making cars more affordable through the use of assembly line production (Parsons & Vigar, 2018). As more and more people began to own cars, they became an essential part of modern life and transformed the way people lived and worked. The automobile made it possible for people to travel long distances quickly and easily, which in turn led to the growth of suburbs and the development of a car-centric culture. It also stimulated economic growth and created new opportunities for businesses such as gas stations, repair shops, and motels (Rip & Kemp, 1998). In general, the acceptance of automobiles by people has been a gradual process that has taken place over many decades. There are several key reasons why people have come to embrace cars as an essential part of modern life such as (Geels, 2005; Lupkin, 2001; Mattioli et al., 2020; Mercure, 2015; Parsons & Vigar, 2018):

1. Convenience and flexibility: One of the main reasons why people have accepted automobiles is that they offer unparalleled convenience and flexibility. Cars allow people to travel quickly and easily from one place to another, without having to rely on public transportation or other modes of travel. This convenience is especially important for people who live in suburban or rural areas where public transportation is less accessible.

2. Increased mobility: The widespread use of cars has dramatically increased people’s mobility, making it possible for them to travel farther and more frequently than ever before. This has had a profound impact on many aspects of society, from the growth of suburbs to the expansion of global trade.

3. Independence and control: For many people, owning a car represents a sense of independence and control over their lives. Cars provide a sense of freedom and autonomy, allowing people to go where they want, when they want, without having to rely on others.

4. Economic benefits: The automobile industry has created millions of jobs and contributed significantly to economic growth around the world. The production, sale, and maintenance of cars have provided a significant source of income for many people, as well as driving the development of related industries such as gas stations, repair shops, and motels.

5. Technological innovation: The ongoing innovation and development of automobile technology have made cars safer, more reliable, and more efficient than ever before. This has helped to improve the public perception of cars and has encouraged people to embrace them as a part of modern life.

6. Cultural acceptance: Finally, over time, the automobile has become deeply ingrained in many cultures around the world. Cars have been featured in movies, music, and literature, and they have become symbols of freedom, independence, and adventure. This cultural acceptance has played a significant role in shaping public perception of cars and has helped to ensure their continued popularity.

Thus, this section shows that the public resistance to the introduction of automobiles was driven by a variety of factors, including fear of the unknown, opposition from established industries, infrastructure challenges, and cost. However, over time, cars became more reliable and affordable, and their practicality and convenience eventually won over public opinion. Today, the automobile is an essential part of modern life, and it is difficult to imagine a world without it.

In light of the DOI theory, private cars followed the S-curve as it was competing with horses as shown in Figure 3. In the early stage of the invention of the automobiles, the technology was not reliable and so the number of cars in the market share was not really increasing as shown between 1900 to 1915. In the same period, we can see that the number of horses used in transportation increased by more than 10 M horses jumping from 22 M to more than 32 M horses. Then, starting from 1915, the level of trust in automobiles started to increase resulting in some increase in the adoption of the technology (cars) and a decrease in the number of horses. Starting from 1945 to 1966, the number of cars in the market started to significantly increase with an average annual increase of 2.5 M cars (25 M cars every 10 years). During this period, cars have already proven their reliability, benefits, and their ability to improve the lives of the public. In addition, it can be observed that the number of cars is significantly higher than the absolute maximum number of horses in their peak, indicating that the cars have reached a wider range of people because of their benefits (these people were not previously served by the horses).

Figure 3. The S-curve for the adoption of cars with the decline in the number of horses in the United States (based on the data from (Foster & Rosenzweig, 2010; Geels, 2005; Mercure, 2015; Rip & Kemp, 1998)).

Thus, for AVs, it is anticipated that AVs will follow a similar pattern and that currently AVs is the very first stage of adoption. Thus, there will be some resistance till the technology proves its reliability and ability to offer new benefits similar to what happened when cars were first deployed. Given the previous discussion, it can be anticipated that AVs will follow a similar pattern as shown in Figure 4. The figure shows the share of horses, cars, and AVs used in the transportation field in the past and for the future. The future data are shown using the dotted lines, while the solid lines were used to represent historical data. The S-curve for the AVs was built based on the data of the horses vs. cars that is previously discussed. Thus, the figure shows that AVs are still in the public resistance phase which was the same phase cars passed through when they were first developed. It is also anticipated that this phase will last for few more years, until the public realizes the technology and its benefits to enter the second phase which is the mature of technology. Then, AVs will enter the widespread phase during which AVs will be the main mode of transportation replacing the human driven vehicles. it must be also mentioned that AVs will reach wider number of people resulting in higher adoption when compared with human-driven vehicles, similar to what happened with cars that reached higher market shares than horses. For example, it is anticipated that AVs will be able to transport kids, children, the elder, the disabled, etc. In addition, people will be able to make longer trips in AVs (when compared to human-driven cars) as they can spend their trips in other activities such as sleeping, working, reading, etc. thus, AVs will cause an increase in the average trip length similar to what happened with the introduction of cars that allowed people to travel longer compared to horses. Thus, it can be concluded that the introduction and growth of AVs might be similar to the introduction of cars as both of them have similar benefits and issues.

Figure 4. Market share of horses, cars, and AVs over the years in the past and future.

4. Methodology

To investigate the relationship between the public’s knowledge of autonomous vehicles AVs and their attitudes towards the technology, a questionnaire survey was designed and conducted on US residents using the SurveyMonkey platform from June to November 2022. Prior to the survey, a pilot survey was conducted in two stages to ensure that the survey was consistent with the study objectives and was easy for participants to understand and complete. The first stage involved testing the survey questions with eight researchers working in the field of public opinion on AVs and the survey was subsequently revised based on their feedback. The second stage involved sending the survey to 28 individuals of the public in the US to ensure that the survey was fair understandable easy to navigate and easy to complete. All 28 respondents agreed that the survey met these criteria.

At this stage of the research project, a survey was disseminated to the general populace throughout the United States, yielding a total of 5778 completed responses. The survey was comprised of three distinct sections. The primary section aimed to provide participants with pertinent background information pertaining to the study, as well as general introductory knowledge surrounding autonomous vehicles (AVs). The second section focused on gathering demographic data from respondents, including age, gender, income, and pre-existing familiarity with AVs. Specifically, respondents were requested to indicate their respective age range, gender, and income category that best reflected their demographic, in addition to rating their level of knowledge regarding AVs on a 5-point Likert scale. The third and final section of the survey focused on gauging public attitudes towards AVs, with respondents asked to rate their levels of interest, trust, and concern towards AVs on a 5-point Likert scale. Furthermore, respondents were requested to disclose the extent to which they were willing to incur additional costs to acquire an AV. The questions of the questionnaire survey and their types are summarized in Table 1. To explore the association between levels of awareness and public attitudes towards AVs, analyses will be conducted on a state-by-state basis.

Table 1. Summary of the survey questions and their types

Question	Type
section 1
Introduction
section 2
How would you describe your gender?	Multiple Choice
How old are you?	Multiple Choice
What is your total annual household income?	Multiple Choice
Prior to taking part in this survey, to what extent have you been exposed to information regarding self-driving cars, on a scale of 1 to 5, where 1 represents no prior knowledge and 5 represents a high level of knowledge?	Ranking
Which state or U.S. territory do you currently reside in?	Dropdown List
section 3
Would you be open to the idea of traveling in autonomous vehicles?	Ranking
Do you share the belief that autonomous vehicles have the potential to enhance traffic safety in comparison to conventional vehicles?	Ranking
how concerned would you be about riding in a self-driving car?	Ranking
What is the maximum amount of additional money, in USD ($), that you would be willing to pay to have a self-driving vehicle? Please enter 0 if you are not willing to pay extra for a self-driving car.	Ranking

The study collected responses from participants across the United States, resulting in a total of 5778 completed surveys and the demographic properties of the respondents are summarized in Table 2. However, the number of responses varied considerably across different states, with some states receiving 240 responses and others receiving no responses at all. In order to prevent the underrepresentation of any particular state and to ensure adequate representation across all regions, the analysis will be conducted at a regional level rather than at a state level. To achieve this, the United States was divided into nine distinct regions: New England, Middle Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, and Pacific. The included states for each region are depicted in Figure 5.

Figure 5. Map showing the states included in every region used in the analysis.

Table 2. Summary of the characteristics of the survey respondents

Character	Number of responses (N)	Percentage (%)
Gender
Male	2773	48%
Female	3005	52%
Age
18–29	1271	22%
30–44	1791	31%
45–60	1329	23%
>60	1387	24%
Education
High School or lower	1907	33%
Bachelor	2600	45%
Master’s degree	924	16%
Higher than master’s degree	347	6%
Annual household income
<25,000	1213	21%
25,000–50,000	1214	21%
50,000–100,000	2022	35%
100,000–200,000	1156	20%
>200,000	173	3%

Due to the varying number of responses received from each state, the data analysis was performed at the regional level. To ensure a representative sample that accurately reflects the spatial distribution of the US population, the number of survey responses received from each region was calculated and compared with the 2022 US Census data (United States Census Bureau, 2022). The results are presented in Figure 6, which displays the number of survey responses and the corresponding percentages of responses from each region alongside the percentage of the US population residing in those regions. As shown in Figure 6, the number of responses collected from each region varied. However, the results indicate that the survey sample accurately represents the entire US population, as the percentage of responses received from each region is comparable to the percentage of the US population residing in those regions, with a maximum error of 0.15%. Consequently, the survey sample is deemed to be representative of the overall US population.

Figure 6. Percentage of responses received in comparison to the percentage of the population that lives in the region as per the US censuses (United States Census Bureau, 2022).

5. Demographic analysis

In this study, the main objective is to evaluate the public’s opinion about AVs. Given the number of responses received, the analysis can be conducted for specific groups that have the same characteristics to minimize the impact of the other demographic parameters. Thus, in this section, the opinions of respondents that have the same demographic properties are summarized and analyzed as shown in Table 3. The values shown in every row show the average opinions of people that have the same gender, are within the same age category, and have similar educational levels. The cells are highlighted from green to red where the green cells highlight the groups with the highest levels of interest, trust, and willingness to pay extra for AVs, while the red cells are for the groups with the lowest inter, trust, and openness to pay for AVs. On the other hand, for the concern data, the cells are highlighted so that the green cells represent the groups with the lowest concern levels, while the red cells are for the groups that have the highest concern levels. This can help in understanding how the different groups of people perceive AVs. It should be noted that the number of responses received from respondents that have an age between 18–29 years and have an educational level that is higher than master’s degree is small across the different groups (less than 3% of the total responses) and very small when compared with the responses received from the other groups. Thus, this category was excluded from the analysis and is highlighted by NA in the table. In addition, this small number of responses for this group is acceptable because the number of people who get an educational degree that is Higher than master’s degree and has 18–29 years is very small and represents less than 2% of the population which is consistent with the results of this study (De Brey et al., 2021), indicating that the survey sample is consistent with the overall population of the US. In addition, the results are summarized in Figures 7, 8, 9, and 10 in order to make the comparison between the different groups easier to be visualized and interpreted. The results can show the difference in the attitude between the different groups. In general, male respondents have a more positive attitude towards AVs than female respondents across the different groups (that have similar age and educational properties). In addition, younger respondents are the most open to adopting AVs than older respondents who have the same gender across the different groups that have different educational levels. Furthermore, the results show that respondents with higher educational levels are the most willing to adopt AVs compared. These results are consistent with previous studies that investigated the public’s opinion towards AVs (Alsghan et al., 2022; Mara & Meyer, 2022; Yuen et al., 2022).

Figure 7. Average level of interest in traveling or riding an AV for the different demographic groups.

Figure 8. Average level of trust in AVs for the different demographic groups.

Figure 9. Average level of concern in traveling or riding in an AV for the different demographic groups.

Figure 10. Average willingness to pay extra to own an AV for the different demographic groups.

Table 3. Average opinions for respondents with different demographic characteristics

6. Evaluating the relationship between the level of knowledge and the willingness to adopt AVs

6.1. Understanding how the public opinions evolve with the level of knowledge

The primary objective of this research is to investigate the association between public awareness of Autonomous Vehicle (AV) technology and public attitudes towards AVs. To achieve this goal, the study undertook a regional analysis of the average level of knowledge and public attitude parameters. The findings of this study are presented and summarized in Table 4. The table depicts the average level of knowledge of respondents from each region, with the colour code representing the regions with low knowledge levels in red and high knowledge levels in green. In addition, the table illustrates the average level of interest in AVs among respondents from each region, with the color code that highlight regions with low interest levels in red and high interest levels in green. Similarly, the table shows the average level of concern of respondents from different regions regarding riding or driving AVs. The table is color-coded to represent regions with low concern levels in green and high concern levels in red. Furthermore, the table summarizes the average level of trust in AVs among respondents from various regions, with a color code indicating high trust levels in green color and low trust levels in red. Lastly, the table presents the average willingness of respondents from different regions to pay extra for an AV. The table is coded to represent regions with low willingness to pay in red and high willingness to pay in green.

Table 4. Summary of the average level of knowledge and the average public attitude parameters across the nine regions

The presented data indicates that the degree of familiarity with autonomous vehicles (AVs) in various regions is related to the attitudes towards AVs expressed by the respondents. Specifically, regions with greater familiarity with AVs are found to have a more pessimistic outlook, evidenced by a lower level of interest, trust, and willingness to pay for AVs, coupled with a higher level of concern about riding in an AV. Conversely, regions with less familiarity with AVs exhibit a greater degree of optimism, as indicated by a higher level of trust, interest, and willingness to pay for AVs. To facilitate comparison, Table 4 summarizes the average level of knowledge and the corresponding average public attitude parameters across the nine regions investigated in this study with color codes where green cells show the highest levels of knowledge and best public attitude and the red color shows the regions or cells that have the lowest levels of knowledge or the lowest levels of openness towards AVs.

According to the table the regions that have the greatest level of familiarity with AVs are the East North Central the West South Central and the East South Central Surprisingly these regions also exhibit the lowest levels of trust interest and willingness to pay more for an AV and rank amongst the regions with the greatest level of concern regarding AVs Conversely the regions with the least prior knowledge about AVs are Mountain New England and South Atlantic and these regions demonstrate the highest levels of trust interest and willingness to pay more for an AV Additionally respondents from these regions exhibit the lowest level of concern about traveling in an AV Consequently the findings reveal an inverse relationship between the level of familiarity with AVs and public attitudes towards AVs This is due to the fact that the increase in the level of knowledge about AVs is associated with a negative shift in public attitudes towards AVs This outcome aligns with the findings of other studies such as those by Othman 2021 (Othman, 2021b) as well as the surveys conducted by AAA in 2016 (American Automobile Association, 2018) and 2019 (American Automobile Association, 2018) which suggest that an increase in knowledge about AVs tends to decrease public interest in AVs

6.2. Developing the relationships between the level of knowledge and public attitude

In order to further understand the impact of prior knowledge on the public attitude towards AVs, it is important to quantify this relationship. To achieve this, this section focuses on the relationship between the average level of prior knowledge about AVs and the public attitude parameters, namely interest, trust, concern, and willingness to pay extra. Figure 11 depicts this relationship, with the x-axis representing the average level of prior knowledge and the y-axis representing the respective public attitude parameter. The figure shows an inverse relationship between the level of prior knowledge and interest in AVs, as summarized by Equation 1(1) $\mathit{ALOI}=5.16-0.6531\ast \mathit{ALOK}$(1) and a high coefficient of determination (R2) of 0.88. This suggests that a 1% increase in knowledge about AVs results in a 0.653% decrease in interest in traveling in an AV. On the other side, Figure 11 provides an overview of the relationship between the average level of knowledge about Autonomous Vehicles (AVs) and the average level of concern about traveling in AVs. The figure depicts a direct relationship between the two parameters, exhibiting a moderate correlation (R2 = 0.5). Equation 2(2) $\mathit{ALOC}=1.69+0.5577\ast \mathit{ALOK}$(2) summarizes the model that describes this relationship, suggesting that an increase of 1% in knowledge about AVs results in a 0.56% increase in the level of concern about traveling in an AV. Furthermore, the figure demonstrates the inverse relationship between the average level of prior knowledge about AVs and the average level of trust in AVs, displaying a strong correlation (R2 = 0.72). Equation 3(3) $\mathit{ALOT}=5.43-0.684\ast \mathit{ALOK}$(3) summarizes the model that describes this relationship, indicating that a 1% increase in knowledge about AVs is associated with a 0.684% decrease in the level of trust in AVs. Moreover, the figure portrays the relationship between the average level of prior knowledge about AVs and the average level of willingness to pay extra to buy an AV, which exhibits an inverse relationship with a strong correlation (R2 = 0.895). The model that summarizes this relationship is shown in Equation 4(4) $\mathit{AWTPE}\left(\mathrm{\$}\right)=15981-2466.4\ast \mathit{ALOK}$(4) , which suggests that a 1% increase in knowledge about AVs results in a reduction of 2466.4 USD in the willingness to pay extra to buy an AV. Overall, these findings imply that the augmentation of the level of knowledge about AVs generates a negative shift in public attitudes towards them concerning interest, trust, concern, and willingness to pay extra to buy an AV.

(1) $\mathit{ALOI}=5.16-0.6531\ast \mathit{ALOK}$(1)

(2) $\mathit{ALOC}=1.69+0.5577\ast \mathit{ALOK}$(2)

(3) $\mathit{ALOT}=5.43-0.684\ast \mathit{ALOK}$(3)

(4) $\mathit{AWTPE}\left(\mathrm{\$}\right)=15981-2466.4\ast \mathit{ALOK}$(4)

Figure 11. Relationship between the average level of prior knowledge and the public attitude parameters.

Where: $\mathit{ALOI}$ is the average level of interest, $\mathit{ALOC}$ is the average level of concern, $\mathit{ALOT}$ is the average level of trust, $\mathit{AWTPE}$ is the average willingness to pay extra for AVs, and $\mathit{ALOK}$ is the average level of knowledge.

7. Developing the relationships between the level of knowledge and public attitude based on the demographic properties

While the previous section shows the relationship between the knowledge level and the public’s opinion and how the public attitude evolves with the level of knowledge, it is also important to evaluate this relationship for respondents with different demographic properties. Consequently, this section focuses on evaluating the relationship between the level of knowledge and the attitude for people with different demographics. In addition, the analysis will compare how the level of acceptance evolves for the different groups in order to understand and reveal the groups whose opinions are sensitive to the changes in the knowledge levels.

7.1. Gender analysis

This subsection focuses on drawing the relationship between the level of awareness of the public and the attitude for male and female respondents. Thus, the level of knowledge and public opinions (in terms of the average levels of trust, interest, concern, and willingness to pay extra for AVs) for male and female respondents were estimated across the nine regions. Then, four figures were developed to show the relationship between the public opinion in the y-axis and the level of knowledge as the independent factor on the x-axis for male and female respondents. These figures are Figures 12, 13, 14, and 15. The results show that male respondents are more open towards AVs than female respondents at the different levels of knowledge; however, there is a negative or an inverse relationship between the level of knowledge and the public attitude. The results show that the level of interest, trust, and willingness to pay for AVs decreases with the increase in the level of knowledge of AVs for both male and female respondents. In addition, males are more open towards AVs. However, the difference (or the gap) in the attitude between male and female respondents increases with the increase in the level of knowledge, indicating that the opinions of female respondents are more sensitive to the level of knowledge than the opinions of male respondents. Thus, it can be stated that the drop in the opinions of the public is higher for female respondents than male respondents. In order to quantify the relationships between the knowledge and attitude, Equation 5(5) $\mathit{AL}{I}_{\mathit{male}}=-0.5619\ast \mathit{AL}{K}_{\mathit{male}}+5.0393$(5) to Equation 8(8) $\mathit{AWT}{P}_{\mathit{male}}=-1799.2\ast \mathit{AL}{K}_{\mathit{male}}+15165$(8) show the relationships between the level of knowledge and public attitude for male respondents, while Equation 9(9) $\mathit{AL}{I}_{\mathit{female}}=-0.699\ast \mathit{AL}{K}_{\mathit{female}}+5.1144$(9) to Equation 12(12) $\mathit{AWT}{P}_{\mathit{female}}=-2788.4\ast \mathit{AL}{K}_{\mathit{female}}+15590$(12) show these relationships for female respondents.

Figure 12. Relationship between the average knowledge level and the average level of interest in AVs for male and female respondents.

Figure 13. Relationship between the average knowledge level and the average level of trust in AVs for male and female respondents.

Figure 14. Relationship between the average knowledge level and the average level of concern of AVs for male and female respondents.

Figure 15. Relationship between the average knowledge level and the average willingness to pay extra for AVs for male and female respondents.

For male respondents:

(5) $\mathit{AL}{I}_{\mathit{male}}=-0.5619\ast \mathit{AL}{K}_{\mathit{male}}+5.0393$(5)

(6) $\mathit{AL}{T}_{\mathit{male}}=-0.6047\ast \mathit{AL}{K}_{\mathit{male}}+5.351$(6)

(7) $\mathit{AL}{C}_{\mathit{male}}=0.4656\ast \mathit{AL}{K}_{\mathit{male}}+1.8191$(7)

(8) $\mathit{AWT}{P}_{\mathit{male}}=-1799.2\ast \mathit{AL}{K}_{\mathit{male}}+15165$(8)

For female respondents:

(9) $\mathit{AL}{I}_{\mathit{female}}=-0.699\ast \mathit{AL}{K}_{\mathit{female}}+5.1144$(9)

(10) $\mathit{AL}{T}_{\mathit{female}}=-0.7146\ast \mathit{AL}{K}_{\mathit{female}}+5.3331$(10)

(11) $\mathit{AL}{C}_{\mathit{female}}=0.6084\ast \mathit{AL}{K}_{\mathit{female}}+1.705$(11)

(12) $\mathit{AWT}{P}_{\mathit{female}}=-2788.4\ast \mathit{AL}{K}_{\mathit{female}}+15590$(12)

Where: $\mathit{ALI}$ is the average level of knowledge, $\mathit{ALT}$ is the average level of trust, $\mathit{ALC}$ is the average level of concern, $\mathit{AWTP}$ is the average willingness to pay extra for AVs ($), $\mathit{ALK}$ is the average level of knowledge.

7.2. Age analysis

Similar to the gender analysis, this subsection focuses on evaluating how the public attitude evolves with the level of knowledge for respondents with different age ranges. Thus, in this subsection, the average level of knowledge and attitude were calculated for respondents within the different age groups across the nine regions of the US. Then, the relationships between the level of knowledge on the x-axis and the attitude parameters on the y-axis were drawn for respondents within different age categories. These relationships are summarized in Figures 16, 17, 18, and 19 with the linear modeling representing the relationships and the resulting R2 for the different age groups. The results show that younger people are the most willing to adopt AVs and the level of openness to adopt AVs decreases with the increase in the age of the respondents. In addition, the increase in the level of knowledge is subjected with a decrease in the openness level to adopt AVs across the different age classes. Furthermore, the opinions of the younger groups are the most sensitive to the level of knowledge because the decrease in the openness to AVs with the level of knowledge is the highest for respondents within the age group of 18–30 years, followed by respondents who have an age of 30–45, followed by 45–60 years old respondents, and finally respondents whose age is higher than 60 years are the least sensitive (however, they have the lowest level of openness to AVs). In addition, the results show that the difference in the level of openness towards AVs decreases with the increase in the level of knowledge of the different age groups. The final equations representing the relationships between the level of knowledge and public attitude of the respondents within the different age groups are summarized in Equation 13(13) $\mathit{AL}{I}_{18-30}=-0.7338\ast \mathit{AL}{K}_{18-30}+5.6281$(13) to Equation 28(28) $\mathit{AWT}{P}_{>60}=-1377.2\ast \mathit{AL}{K}_{>60}+11012$(28) .

Figure 16. Relationship between the average knowledge level and the average level of interest in AVs for people within the different age groups.

Figure 17. Relationship between the average knowledge level and the average level of trust in AVs for people within the different age groups.

Figure 18. Relationship between the average knowledge level and the average level of concern in AVs for people within the different age groups.

Figure 19. Relationship between the average knowledge level and the average willingness to pay extra for AVs for people within the different age groups.

Level of interest models across the different age groups:

(13) $\mathit{AL}{I}_{18-30}=-0.7338\ast \mathit{AL}{K}_{18-30}+5.6281$(13)

(14) $\mathit{AL}{I}_{30-45}=-0.6716\ast \mathit{AL}{K}_{30-45}+5.3004$(14)

(15) $\mathit{AL}{I}_{45-60}=-0.6185\ast \mathit{AL}{K}_{45-60}+4.9644$(15)

(16) $\mathit{AL}{I}_{>60}=-0.4955\ast \mathit{AL}{K}_{>60}+4.4589$(16)

Level of trust models across the different age groups:

(17) $\mathit{AL}{T}_{18-30}=-0.6924\ast \mathit{AL}{K}_{18-30}+5.6256$(17)

(18) $\mathit{AL}{T}_{30-45}=-0.6897\ast \mathit{AL}{K}_{30-45}+5.5294$(18)

(19) $\mathit{AL}{T}_{45-60}=-0.6612\ast \mathit{AL}{K}_{45-60}+5.2688$(19)

(20) $\mathit{AL}{T}_{>60}=-0.6323\ast \mathit{AL}{K}_{>60}+5.0806$(20)

Level of trust models across the different age groups:

(21) $\mathit{AL}{C}_{18-30}=0.6719\ast \mathit{AL}{K}_{18-30}+1.1105$(21)

(22) $\mathit{AL}{C}_{30-45}=0.5863\ast \mathit{AL}{K}_{30-45}+1.5167$(22)

(23) $\mathit{AL}{C}_{45-60}=0.5148\ast \mathit{AL}{K}_{45-60}+1.9018$(23)

(24) $\mathit{AL}{C}_{>60}=0.3808\ast \mathit{AL}{K}_{>60}+2.425$(24)

Level of trust models across the different age groups:

(25) $\mathit{AWT}{P}_{18-30}=-2950.2\ast \mathit{AL}{K}_{18-30}+19124$(25)

(26) $\mathit{AWT}{P}_{30-45}=-2577.8\ast \mathit{AL}{K}_{30-45}+16974$(26)

(27) $\mathit{AWT}{P}_{45-60}=-2233\ast \mathit{AL}{K}_{45-60}+14617$(27)

(28) $\mathit{AWT}{P}_{>60}=-1377.2\ast \mathit{AL}{K}_{>60}+11012$(28)

Where: $\mathit{ALI}$ is the average level of knowledge, $\mathit{ALT}$ is the average level of trust, $\mathit{ALC}$ is the average level of concern, $\mathit{AWTP}$ is the average willingness to pay extra for AVs ($), $\mathit{ALK}$ is the average level of knowledge.

7.3. Educational levels

This subsection focuses on investigating the relationship between the level of knowledge and the public attitude for people with different educational levels. Thus, the average level of knowledge and the average public attitude parameters were estimated for respondents with different educational levels across the nine regions of the US. Then, the relationship between the average level of knowledge on the x-axis and the public attitude parameters on the y-axis were drawn and analyzed in order to evaluate how the public’s openness to AVs evolves with the level of knowledge for people with different educational levels. Figures 20, 21, 22, and 23 summarize the relationships between the public attitude and the level of knowledge of respondents who attained different educational levels. In addition, the figures show the linear models that summarize these relationships with the resulting R2 values. In general, the results show that the level of openness to adopt AVs increases with the increase in the educational level attained. In addition, the public attitude towards AVs has an inverse relationship with the level of knowledge for people with different educational levels because the four public attitude parameters move in the negative direction with the increase in the average level of knowledge. In addition, the results show that the opinions of respondents who have bachelor’s degrees are the most sensitive to the level of knowledge as the shift in their attitude with the change in the level of knowledge is the highest in terms of the different public attitude parameters. On the other side, the changes in the opinions of respondents who attained master’s degrees are the least affected by the level of knowledge as the changes in their attitude are minor (with the change in the level of knowledge) when compared to other respondents who have different educational levels, expect for the level of concern. For the average level of concern, the results show that respondents who have a master’s degree are the most affected by the change in the level of knowledge followed by respondents who have bachelor’s degrees, respondents who have a high school degree or lower, and finally respondents who have an educational degree that is higher than master’s degree. Furthermore, it can be noted that the opinions of respondents who have higher school degrees or lower and respondents who have bachelor’s degrees are similar, especially at high levels of knowledge. At low levels of knowledge, respondents who have bachelor’s degrees have better opinions towards AVs than respondents who attained high school or have lower degrees. However, this difference decreases with the increase in the level of knowledge and at high levels of knowledge both respondents who have bachelor’s degrees and respondents who attained high school degrees or lower have identical opinions in terms of the different attitude parameters. On the other hand, the opinions of respondents who have master’s degrees or higher are much more positive towards AVs across the different levels of knowledge. The final linear models that represent the relationships between the level of knowledge and public attitude for respondents with different educational levels are summarized in Equation 29(29) $\mathit{AL}{I}_{\mathit{HSorL}}=-0.5766\ast \mathit{AL}{K}_{\mathit{HSorL}}+4.7043$(29) to Equation 44(44) $\mathit{AWT}{P}_{>\mathit{MSC}}=-1477.6\ast \mathit{AL}{K}_{>\mathit{MSC}}+19649$(44) .

Figure 20. Relationship between the average knowledge level and the average level of interest in AVs for people within the different age groups.

Figure 21. Relationship between the average knowledge level and the average level of trust in AVs for people within the different age groups.

Figure 22. Relationship between the average knowledge level and the average level of concern in AVs for people within the different age groups.

Figure 23. Relationship between the average knowledge level and the average willingness to pay extra for AVs for people within the different age groups.

Level of interest models for participants with different educational levels:

(29) $\mathit{AL}{I}_{\mathit{HSorL}}=-0.5766\ast \mathit{AL}{K}_{\mathit{HSorL}}+4.7043$(29)

(30) $\mathit{AL}{I}_B=-0.7581\ast \mathit{AL}{K}_B+5.3758$(30)

(31) $\mathit{AL}{I}_{\mathit{MSC}}=-0.3841\ast \mathit{AL}{K}_{\mathit{MSC}}+4.617$(31)

(32) $\mathit{AL}{I}_{>\mathit{MSC}}=-0.5458\ast \mathit{AL}{K}_{>\mathit{MSC}}+5.7329$(32)

Level of trust models for participants with different educational levels:

(33) $\mathit{AL}{T}_{\mathit{HSorL}}=-0.5986\ast \mathit{AL}{K}_{\mathit{HSorL}}+4.9421$(33)

(34) $\mathit{AL}{T}_B=-0.7749\ast \mathit{AL}{K}_B+5.599$(34)

(35) $\mathit{AL}{T}_{\mathit{MSC}}=-0.4642\ast \mathit{AL}{K}_{\mathit{MSC}}+5.0607$(35)

(36) $\mathit{AL}{T}_{>\mathit{MSC}}=-0.5284\ast \mathit{AL}{K}_{>\mathit{MSC}}+5.838$(36)

Level of trust models for participants with different educational levels:

(37) $\mathit{AL}{C}_{\mathit{HSorL}}=0.4637\ast \mathit{AL}{K}_{\mathit{HSorL}}+2.1812$(37)

(38) $\mathit{AL}{C}_B=0.6002\ast \mathit{AL}{K}_B+1.6385$(38)

(39) $\mathit{AL}{C}_{\mathit{MSC}}=0.6588\ast \mathit{AL}{K}_{\mathit{MSC}}+1.1529$(39)

(40) $\mathit{AL}{C}_{>\mathit{MSC}}=0.3326\ast \mathit{AL}{K}_{>\mathit{MSC}}+1.3211$(40)

Level of trust models for participants with different educational levels:

(41) $\mathit{AWT}{P}_{\mathit{HSorL}}=-1960.8\ast \mathit{AL}{K}_{\mathit{HSorL}}+12787$(41)

(42) $\mathit{AWT}{P}_B=-3212.7\ast \mathit{AL}{K}_B+17494$(42)

(43) $\mathit{AWT}{P}_{\mathit{MSC}}=-460.39\ast \mathit{AL}{K}_{\mathit{MSC}}+11988$(43)

(44) $\mathit{AWT}{P}_{>\mathit{MSC}}=-1477.6\ast \mathit{AL}{K}_{>\mathit{MSC}}+19649$(44)

Where: $\mathit{ALI}$ is the average level of knowledge, $\mathit{ALT}$ is the average level of trust, $\mathit{ALC}$ is the average level of concern, $\mathit{AWTP}$ is the average willingness to pay extra for AVs ($), $\mathit{ALK}$ is the average level of knowledge, $\mathit{HSorL}$ is high school or lower, $\mathit{B}$ is bachelor degree, $\mathit{MSC}$ is masters degree, and $>\mathit{MSC}$ is higher than master’s degree.

7.4. Analysis by household income

This subsection focuses on investigating and analyzing the relationship between the level of knowledge and the public openness to adopt AVs based on the annual household income of the respondents. Thus, the average level of knowledge and the average public attitude parameters were estimated for respondents who have different levels of household incomes across the nine regions used in the analysis. In addition, the relationships between the level of knowledge on the x-axis and the public attitude parameters on the y-axis were drawn to understand how the public openness to adopt AVs evolves with the change in the level of knowledge for people who have different annual household incomes. Figures 24, 25, 26, and 27 summarize the relationships between the public attitude and the level of knowledge for respondents with different household income levels. In addition, the figures show the linear models that summarize these relationships with the resulting R2 values. The general trend shows that respondents with low and intermediate household incomes (<25, 25–50, and 50–100) have similar levels of openness toward AVs across the different levels of knowledge. On the other hand, at higher household incomes, the level of openness to adopt AVs increases with the increase in the household income. In addition, the results show that the level of acceptance of AVs decreases with the increase in the level of knowledge about AVs across the different annual household income groups. Furthermore, the results show that respondents who have low household incomes have similar levels of openness towards AVs. For example, respondents who have a household income lower than 25K ($) and a household income 25-50K ($) have identical levels of opinion towards AVs across the different levels of knowledge. In addition, respondents who have a household income of 50-100K (USD), when compared to respondents who have a low annual household income of (<25K and 25 to 50K USD), have better opinions about AVs at low levels of knowledge. On the other hand, at high levels of knowledge about AVs, the opinions of respondents with a household income of 50-100K (USD) dropped to a level that worse than respondents who have low household income that is lower than 25K or 25-50K ($). On the other side, respondents who live with a household income that is higher than 100K ($) are more (100-200K $ and > 200K $) open to adopt AVs across the different knowledge levels. In general, the opinions of respondents who have a household income of 50-100K ($) are the most sensitive to the level of knowledge and the shift in their attitude as a result of the increase in their knowledge level, when compared to respondents with low household incomes (<25K and 25-50K $), shows that their opinions drop from higher openness towards AVs at low levels of knowledge to lower openness towards AVs at high levels of knowledge. On the other hand, respondents with low household incomes (<25K $ and 25-50K $) are the least sensitive to the change in the level of knowledge. For respondents with high household incomes (100-200K $ and > 200K $), the results show that respondents with a household income of 100-200K $ are more sensitive to the level of knowledge as the results show that these respondents have higher drops in their opinions with the increase in the level of knowledge. In general, respondents who have an annual household income that is higher than 200K ($) are the least sensitive to the changes in the level of knowledge across the different household groups. Thus, it can be stated that respondents with high household incomes, when compared to the other household income groups, are the most willing to adopt AVs and their opinions are the least sensitive (or the least affected by) to the changes in the level of knowledge about AVs. The final linear models that represent the relationships between the level of knowledge and public attitude for respondents with different annual household income levels are summarized in Equations 45(45) $\mathit{AL}{I}_{<25}=-0.5815\ast \mathit{AL}{K}_{<25}+4.8364$(45) to Equation 64(64) $\mathit{AWT}{P}_{>200}=-846.53\ast \mathit{AL}{K}_{>200}+14372$(64) .

Figure 24. Relationship between the average knowledge level and the average level of interest in AVs for people with different annual household incomes.

Figure 25. Relationship between the average knowledge level and the average trust of interest in AVs for people with different annual household incomes.

Figure 26. Relationship between the average knowledge level and the average level of concern in traveling in AVs for people with different annual household incomes.

Figure 27. Relationship between the average knowledge level and the average willingness to pay extra for people with different annual household incomes.

Level of interest models for participants with different annual household income levels:

(45) $\mathit{AL}{I}_{<25}=-0.5815\ast \mathit{AL}{K}_{<25}+4.8364$(45)

(46) $\mathit{AL}{I}_{25-50}=-0.5811\ast \mathit{AL}{K}_{25-50}+4.834$(46)

(47) $\mathit{AL}{I}_{50-100}=-0.6785\ast \mathit{AL}{K}_{50-100}+5.1578$(47)

(48) $\mathit{AL}{I}_{100-200}=-0.6753\ast \mathit{AL}{K}_{100-200}+5.3242$(48)

(49) $\mathit{AL}{I}_{>200}=-0.5508\ast \mathit{AL}{K}_{>200}+5.0436$(49)

Level of trust models for participants with different annual household income levels:

(50) $\mathit{AL}{T}_{<25}=-0.6078\ast \mathit{AL}{K}_{<25}+5.0893$(50)

(51) $\mathit{AL}{T}_{25-50}=-0.6211\ast \mathit{AL}{K}_{25-50}+5.1295$(51)

(52) $\mathit{AL}{T}_{50-100}=-0.7274\ast \mathit{AL}{K}_{50-100}+5.4817$(52)

(53) $\mathit{AL}{T}_{100-200}=-0.6852\ast \mathit{AL}{K}_{100-200}+5.527$(53)

(54) $\mathit{AL}{T}_{>200}=-0.601\ast \mathit{AL}{K}_{>200}+5.3809$(54)

Level of trust models for participants with different annual household income levels:

(55) $\mathit{AL}{C}_{<25}=0.4844\ast \mathit{AL}{K}_{<25}+2.008$(55)

(56) $\mathit{AL}{C}_{25-50}=0.4961\ast \mathit{AL}{K}_{25-50}+1.973$(56)

(57) $\mathit{AL}{C}_{50-100}=0.5794\ast \mathit{AL}{K}_{50-100}+1.6942$(57)

(58) $\mathit{AL}{C}_{100-200}=0.5872\ast \mathit{AL}{K}_{100-200}+1.5039$(58)

(59) $\mathit{AL}{C}_{>200}=0.4914\ast \mathit{AL}{K}_{>200}+1.6994$(59)

Level of trust models for participants with different annual household income levels:

(60) $\mathit{AWT}{P}_{<25}=-1277.1\ast \mathit{AL}{K}_{<25}+10664$(60)

(61) $\mathit{AWT}{P}_{25-50}=-1365.6\ast \mathit{AL}{K}_{25-50}+10918$(61)

(62) $\mathit{AWT}{P}_{50-100}=-2937.3\ast \mathit{AL}{K}_{50-100}+16152$(62)

(63) $\mathit{AWT}{P}_{100-200}=-2757.8\ast \mathit{AL}{K}_{100-200}+18516$(63)

(64) $\mathit{AWT}{P}_{>200}=-846.53\ast \mathit{AL}{K}_{>200}+14372$(64)

Where: $\mathit{ALI}$ is the average level of knowledge, $\mathit{ALT}$ is the average level of trust, $\mathit{ALC}$ is the average level of concern, $\mathit{AWTP}$ is the average willingness to pay extra for AVs ($), $\mathit{ALK}$ is the average level of knowledge.

8. Discussion

This study focuses on analyzing and understanding how the level of openness towards AVs is affected by the level of knowledge for the public in the US. The results show an inverse relationship between the level of knowledge and the willingness to adopt AVs. These results are consistent with the results of previous studies that highlighted that while the level of knowledge about AVs increases, the willingness to adopt or travel in AVs decreases (Lienert, 2018; Othman, 2021b, 2023). Thus, it can be concluded that the information delivered to the public are mostly negative and causes a negative shift in their opinions. Thus, people with high levels of knowledge about AVs are less willing to adopt the technology. These results are also consistent with previous studies that tried to understand and evaluate the impact of the public’s misconception about AVs on their attitude towards this technology. These studies show that people who have higher levels of misconception about AVs are the most optimistic to adopt AVs (Du et al., 2022; Lipson & Kurman, 2016; Liu et al., 2022; Nikitas et al., 2019; Smith, 2014). On the other hand, the results show that people with lower miscaptions about AVs are more skeptical about AVs. These results indicate that people who have low levels of knowledge about AVs are the most optimistic to adopt and travel using this technology, which is consistent with the results of this study.

The comparison between the introduction of AVs and the introduction of automobiles (more than 100 years ago) shows that the two technologies are facing the same resistance at their early stages. For example, in the early days of the automobile, there was a great deal of skepticism and resistance from the public. Some people saw the automobile as a dangerous and unnecessary luxury, while others worried about the impact it would have on traditional forms of transportation, such as horses and trains. There were also concerns about the safety of automobiles, as early models had a reputation for being unreliable and prone to accidents. Similarly, now with the introduction of AVs, there is a great level of skepticism and fear of AVs. In addition, some people cannot leave their lives in the hands of machines to decide what are the correct actions. While the two technologies were introduced in different ears, they are facing similar issues and similar resistance. This resistance can be explained using the Gartner Hype Cycle which is used to explain how the public reacts to innovative or new technologies (Tarkovskiy, 2013). The cycle shows that any innovation passes through five phases during its life cycle. These phases are called “Technology Trigger”, “Peak of Inflated Expectations”, “Trough of Disillusionment”, “Slope of Enlightenment”, and “Plateau of Productivity” (Chaffey & Ellis-Chadwick, 2019). The first stage, known as the Technology Trigger, occurs when a potential breakthrough in technology generates early proof-of-concept stories and media interest that trigger significant publicity, despite the absence of usable products and unproven commercial viability (Sodhi et al., 2022). In the second stage, the Peak of Inflated Expectations, early publicity produces a number of success stories, but also numerous failures, and while some companies take action, most do not. The third stage, the Trough of Disillusionment, marks a decline in interest as experiments and implementations fail to deliver, and producers of the technology either shake out or fail, with investment continuing only if the surviving providers improve their products to the satisfaction of early adopters (Kondo et al., 2022). In the fourth stage, the Slope of Enlightenment, more instances of how the technology can benefit the enterprise start to crystallize and become more widely understood, with second- and third-generation products appearing from technology providers, and more enterprises funding pilots, while conservative companies remain cautious (Stratopoulos & Wang, 2022). Finally, in the fifth stage, the Plateau of Productivity, mainstream adoption starts to take off, criteria for assessing provider viability become more clearly defined, and the technology’s broad market applicability and relevance are clearly paying off (Mason & Manzotti, 2009). If the technology has more than a niche market, then it will continue to grow (Chaffey & Ellis-Chadwick, 2019; Dehghanimadvar et al., 2022; Tarkovskiy, 2013). Thus, looking at the introduction of AVs shows that the technology is following this curve. Initially, the self-driving car concept was accompanied by optimistic news regarding its potential benefits (Dehghanimadvar et al., 2022). Subsequently, the technology entered the second stage of the technology adoption life cycle, known as the Peak of Inflated Expectations, during which there was an abundance of positive news, articles, reports, and research papers discussing the benefits and implications of autonomous vehicles (Liljaniemi et al., 2023; Tarkovskiy, 2013). This phase was characterized by large investments from both the industry and research sectors (Chaffey & Ellis-Chadwick, 2019). However, in recent years, the technology has entered the third and most critical phase of the life cycle, the Trough of Disillusionment, as self-driving cars have been involved in a significant number of accidents, revealing the imperfections of the technology and changing public perception towards it. The five phases of the Gartner Hype Cycle are summarized in Figure 28 and the figure also shows the current state of AVs technology. Therefore, it is reasonable to expect that the negative shift in public perception towards AVs will persist for a considerable period until the technology reaches the “Slope of Enlightenment” phase. Currently, the technology is in a critical stage that will ultimately determine its acceptability. Furthermore, according to the technology adoption life cycle framework, it is anticipated that there will be further instances of experimentation failures, accompanied by negative news and a further decline in public opinion towards AVs in the near future.

Figure 28. the Gartner Hype Cycle showing the current state of AVs.

9. Conclusions

The potential of autonomous vehicles (AVs) to enhance traffic safety by eliminating human error, which is responsible for 90% of traffic accidents, is widely recognized. However, the realization of these benefits depends heavily on public acceptance and opinion of this emerging technology. Prior research has produced inconsistent results regarding the impact of knowledge level on public attitudes towards AVs, with some studies suggesting a positive relationship between knowledge and attitudes and others indicating a negative relationship. Despite these discrepancies, no research has specifically explored the relationship between prior knowledge and public attitudes towards AVs. In response, this study aims to address this gap in the literature by examining the relationship between knowledge level and public attitudes towards AVs in the US. To this end, a survey questionnaire was designed and distributed to the US public, generating a total of 5778 complete responses from all nine main regions of the US. The study then calculated the average levels of knowledge, trust, interest, concern, and willingness to pay extra for AVs in each region, and subsequently examined the relationship between knowledge level and public attitude parameters to gain a better understanding of how prior knowledge influences public attitudes towards AVs. The main conclusions of this study can be summarized as follows:

• According to the study, there exists an inverse relationship between prior knowledge and public attitude towards autonomous vehicles (AVs). The study found that regions with higher knowledge about AVs were more pessimistic about them, whereas regions with lower knowledge were more optimistic. Regression analysis suggested that a 1% increase in knowledge was linked to a 0.65% decrease in interest, 0.68% decrease in trust, $2466 decrease in willingness to pay for AVs, and a 0.56% increase in concern about traveling in AVs.

• These results indicate that people become increasingly negative towards AVs as their knowledge about them increases, which may be due to the negative news surrounding AV accidents that tends to increase fear of AVs.

• In addition, the analysis was conducted for respondents with different demographic properties. The results show that female respondents are more sensitive to the level of knowledge as the drop of their opinions is higher than male respondents with the increase in the level of knowledge about AVs. Furthermore, younger respondents are the most willing to adopt AVs. Moreover, respondents who have educational levels that are high school or lower or bachelor’s degree have similar attitude towards AVs, especially at high levels of knowledge. On the other hand, respondents with higher educational levels have higher levels of openness to adopt AVs across the different levels of knowledge. Finally, respondents with high household incomes are the most willing to adopt AVs and their opinions are the least sensitive (or the least affected by) to the changes in the level of knowledge about AVs. In addition, respondents with low and intermediate have similar levels of openness towards AVs.

• Analysing the results of this study and comparing it to the history of the introduction of the automobiles shows that the two technologies have similar resistance. In addition, analysing the results in light of both the theory of DOI and the Gartner Hype Cycle shows that AV technology is in the Trough of Disillusionment phase. Therefore, it is anticipated that the negative shift in public perception towards AVs will persist in the future (in the next few years) until the technology reaches the “Slope of Enlightenment” phase. Thus, it can be concluded that, according to the technology adoption life cycle framework, it is anticipated that there will be further instances of experimentation failures, accompanied by negative news and a further decline in public opinion towards AVs in the near future.

10. Study limitations and recommendations for future research

The analysis conducted in this study was mainly based on the average values of the levels of knowledge and the average public attitude parameters which might result in some estimation error. Moreover, this study was focusing on capturing the opinions of respondents from the US, so all the results are limited to this area. Thus, it is recommended to replicate this study in other countries and compare the results with the results presented in this study (as a benchmark). In addition, this study was conducted at a specific point in time. However, the theory of DOI and the hype cycle have shown that the public attitude changes over time and currently AV technology is in the Trough of Disillusionment phase. Thus, it is anticipated that the positive attitudes towards AVs will keep declining in the next few years till a specific point and the opinions will start to be positive once again. Thus, it is highly recommended to replicate this study on a regular basis in order to keep tracking the public attitude towards AVs and to find the time when AVs enter the next phase which is called the “Slope of Enlightenment”. Furthermore, this study compared the introduction of automobiles and the introduction of AVs showing that both of them have had similar patterns and public perception. Thus, it is highlight recommended to review these two cases and review the law and regulatory history with the case of the introduction of the automobiles and provide some regulatory or law recommendations for AVs in light of the introduction of the automobiles.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Kareem Othman

Kareem Othman received his B.S and M.Sc degrees from the civil engineering department, faculty of engineering, Cairo University. His graduation project received the “Best Transportation Engineering Graduation Project” award in Egypt (Engineer Syndicate Competition) by the Egyptian Society of Engineering, The Egyptian Engineers Syndicate in 2016. In 2017, He received the “Excellence Award” from The Egyptian Engineers Syndicate. He did his Ph.D. at the University of Toronto, Canada, during which he received multiple distinguished awards such as the CAA award. His main research interests include public transit, autonomous vehicles, multimodal arterial control, and asphalt pavement design. This study is part of a bigger research project that focuses on understanding the public attitude towards autonomous vehicles and the factors that influence the interest of the public in autonomous vehicles.