TECHNOLOGICAL SAVVINESS AS A MEDIATOR OF OLDER ADULTS’ REPORTED 
LIKELIHOOD OF ADOPTION AND ENDORSEMENT OF AUTONOMOUS VEHICLES 

By 

Elise Atkinson 

A THESIS 

…………………………………………. Submitted to…………………………………………. 
…..…………………..……..……. Michigan State University…..………………………..…..… 
….……………..……..……in partial fulfillment of the requirements…..……………..……..…… 
for the degree of 

Media and Information – Master of Arts 

2024 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ABSTRACT 

Autonomous vehicles are poised to transform American transportation, offering increased 

mobility, especially for vulnerable populations such as older adults. While much of the research 

has focused on the technological and safety aspects of autonomous vehicles, it is crucial to 

understand the factors influencing older adults’ adoption of this technology. This study examines 

how technological savviness, as measured by frequency of internet use, impacts older adults' 

attitudes toward autonomous vehicles. Using data from a nationally representative survey 

conducted by the Pew Research Center in November 2021, the Technology Acceptance Model 

and the Theory of Planned Behavior were employed to inform a mediation model. The analysis 

revealed that higher technological savviness is positively associated with favorable views toward 

autonomous vehicles, including increased willingness to ride in autonomous vehicles, 

perceptions of autonomous vehicles as reducing transportation stress, being acceptable for 

sharing the road with, and being used as taxis and buses. Furthermore, technological savviness 

was found to fully mediate the relationship between age and the perception of autonomous 

vehicles as beneficial for society and supporting older adults’ independence. However, no 

significant association was found between technological savviness and safety concerns, such as 

crash likelihood or hacking risks. In addition to technological savviness, demographic factors, 

particularly gender, influenced attitudes toward autonomous vehicles. These findings suggest that 

initiatives aimed at improving technological savviness, such as educational programs, user-

centered design efforts, and collaborations with advocacy groups, could foster greater acceptance 

of autonomous vehicles among older populations.

TABLE OF CONTENTS 

INTRODUCTION……….…….………………………………………………………………….. .1 

HYPOTHESES…....………………………………………………………………..……………...9 

METHODS.………...…………………………………………………………………………….13 

RESULTS……………………………………………………………………………………..….20 

DISCUSSION….…………………………………………………………………………………57 

CONCLUSION…………………………………………………………………………………..69 

BIBLIOGRAPHY………………………………………………………………………………..71 

APPENDIX………………...……………….……………………………………………………81 

iii 

INTRODUCTION 

The proportion of adults aged 65 and older in the United States has tripled over the past century 

and is projected to reach 28% of the U.S. population by 2050 (Lin & Cui, 2021). As this 

population continues to grow, it will experience increasing transportation demands, compounded 

by age-related mobility challenges. These challenges may include declining physical capabilities, 

slower reflexes, and health conditions that impair the safe operation of conventional vehicles, 

presenting significant safety concerns. Notably, drivers aged 75 and older who have adverse 

medical and physical conditions are associated with a heightened risk of crash fatalities (OECD, 

2022) with states implementing frequent in-person license renewals and mandatory vision and 

road tests have demonstrating lower rates of driving-related fatalities among older adults (Shen 

et al., 2020). Given the convergence of an aging population, growing transportation needs, and 

age-related limitations, there is a pressing need to explore innovative and accessible 

transportation alternatives, such as autonomous vehicles, to address the specific needs of this 

demographic (Eby et al., 2017). 

Transportation access remains a persistent challenge in the United States, largely due to 

infrastructure and zoning policies that prioritize personal vehicle use (Mattioli & Colleoni, 

2015). Urban sprawl, with its low-density and dispersed residential zoning, has made private 

vehicle ownership nearly essential for daily commuting (Andong & Sajor, 2017). This pattern 

contributes to longer travel times, increased traffic congestion, and greater demands on road 

infrastructure (Gössling, 2020). This car-centric approach to urban planning has often led to the 

neglect and underfunding of public transportation systems, particularly in suburban and rural 

areas, where alternatives to personal vehicles are limited (Cantilina et al., 2021, Ingram et al., 

2020). The lack of efficient, reliable, and comprehensive public transit options disproportionately 

1 

affects individuals who cannot access or operate a personal vehicle, including older adults facing 

age-related limitations (Shaheen & Wong, 2022, van Eldijk et al., 2022). This poses significant 

barriers to mobility and independence for these vulnerable populations (Ingram et al., 2020, van 

Eldijk et al., 2022). 

To address these deficiencies, autonomous vehicles present a promising solution, particularly 

given some of the more advantageous aspects of U.S. road infrastructure. The United States 

presents a uniquely favorable environment for the deployment of autonomous vehicles in part 

due to its highly structured road infrastructure (Othman, 2021). Unlike Europe, where roadways 

often evolved organically from footpaths and trade routes over centuries, U.S. roads were 

developed rapidly during the nation’s westward expansion and follow a more deliberate, grid-

like design (Rana & Hossain, 2021). This grid system, particularly in urban and suburban areas, 

reduces the complexity of intersections and roadways, making it easier for autonomous vehicles 

to navigate (Liu et al., 2023). The prevalence of standardized, perpendicular intersections, 

compared to the more irregular intersections found in Europe, enhances the predictability of 

driving environments for autonomous systems (Zhu et al., 2023). Additionally, the U.S. benefits 

from a cultural and legal framework that supports stringent traffic law enforcement and 

separation of car and pedestrian (Fliss et al., 2020). Compared to countries where roads are not 

solely reserved for cars, American roadways offer a more controlled and predictable driving 

environment (Azmi et al., 2024). This level of regulation, coupled with clearly marked lanes, 

traffic signage, and adherence to speed limits, creates a setting that is more conducive to the 

integration of autonomous vehicles (Gössling, 2020). These factors, taken together, suggest that 

the U.S. may be particularly well-suited to the early adoption and widespread use of autonomous 

transportation (Shaheen & Wong, 2022). 

2 

Autonomous vehicles hold the potential to preserve the independence of United States’s older 

adults, enhancing their overall quality of life. The ability to manage daily tasks autonomously is 

closely associated with higher levels of self-esteem, self-worth, and life satisfaction among older 

adults (Leeuwen et al., 2019). Such independence helps individuals maintain dignity and a sense 

of agency, enabling them to make decisions aligned with their personal values and preferences 

(Soósová, 2016). Furthermore, independent living affords older adults more opportunities for 

social engagement, which is vital for reducing feelings of loneliness and fostering strong social 

connections, both of which are integral to mental and emotional well-being (Suragarn et al., 

2021, Baker et al., 2018). Independence is also linked to greater physical and cognitive activity, 

which can help sustain overall health, mobility, and may lower the risk of cognitive decline, 

including conditions such as dementia (Olivari et al., 2018). In addition to these health benefits, 

maintaining independence through the use of autonomous vehicles could alleviate the financial 

strain associated with care centers and reduce the caregiving burden on families and 

communities. 

Despite the numerous benefits autonomous vehicles could offer, older adults’ receptivity to this 

technology remains a significant hurdle, as their current outlook on the technology is, on 

average, unfavorable (Zoaktafi, Gandy, & Yoon, 2023). In the existing body of literature, there is 

a consensus that older adults exhibit a significant degree of resistance towards autonomous 

vehicles (Carr et al., 2023). Studies have consistently reported a pervasive reluctance among 

older populations regarding their reported likelihood of adoption and endorsement of the 

technology (Rahman & Thill, 2023). In this study, “reported likelihood of adoption” is 

operationalized as the personal inclination to use autonomous vehicles, reflecting an individual’s 

direct interest in engaging with this technology (Zoaktafi, Gandy, & Yoon, 2023). Meanwhile, 

3 

"endorsement" refers to a broader evaluation of attitudes and perceptions surrounding the use of 

autonomous vehicles by others, rather than personal adoption (Carr et al., 2023). This captures a 

general societal viewpoint, not just individual willingness to use the technology. 

To fully realize the benefits of autonomous vehicles for older adults, it is essential to address 

existing research gaps in understanding their adoption and endorsement. A notable gap remains 

in understanding the relationship between older adults and autonomous vehicle technology. 

Studies that focus specifically on older adults often rely on small qualitative samples or fail to 

include younger cohorts for comparative analysis (Abdelrahman et al., 2020, Kadylak et al., 

2021). Furthermore, many studies that identify significant age-related differences in attitudes 

towards autonomous vehicles predominantly concentrate on implications for younger adults, 

leaving the implications for older adults underexplored. For instance, in analyses of the 2017 

American Trends Panel survey, a previous fielding of the survey explored in this study, Mesch 

and Dodel, as well as Nair and Bhat, found that age was negatively associated with support for 

autonomous vehicles, indicating that older individuals were less likely to endorse the technology 

(Mesch & Dodel, 2022, Nair & Bhat, 2021). However, in their discussions, these studies 

primarily focused on onboarding strategies for the 18-29 year-old demographic, neglecting the 

potential benefits autonomous vehicles may offer to older adults.  

Research on older adults typically explores how demographic factors, such as age, and less 

mutable traits, such as attitudes, are associated with endorsement of autonomous vehicles 

(Rahman et al., 2020, Kadylak et al., 2021). However, studies by Mesch and Dodel, as well as 

Nair and Bhat, examine key mutable factors, including frequency of internet use, with the latter 

incorporating "living a technological savviness lifestyle" as a latent variable in their path model 

(Mesch & Dodel, 2022, Nair & Bhat, 2021). These studies highlight the importance of focusing 

4 

not only on patterns of adoption and endorsement but also on mutable characteristics that can be 

shaped by public policy interventions. Despite this, the relationship between immutable 

demographic factors, such as age, and mutable characteristics, such as technological savviness, 

in relation to likelihood of adoption and endorsement remains underexplored. This gap in the 

literature points to the need for deeper investigation into how older adults perceive, adopt, and 

adapt new technologies, such as autonomous vehicles, particularly given the potential mobility 

benefits autonomous vehicles offer to this population. 

Technological savviness is defined here specifically as an individual’s ability to adapt to new 

technologies. It emphasizes a familiarity with common design patterns across technologies and 

the ability to work through challenges without external help. Meanwhile, digital/technological 

skills is more narrowly defined as the practical application of technology, emphasizing learned 

skills on specific programs and technologies (Falloon, 2020, Erstad, 2022). In stark contrast, 

digital literacy is defined by accurately evaluating information, with the inclusion of critical 

thinking abilities for assessing the validity of digital content (Mentzer, Frydenberg, & Patterson, 

2024, Harrison, 2023). This paper adopts the term technological savviness to denote a broad, 

adaptive proficiency in interacting with and applying technology, setting it apart from the more 

specific constructs of tech skills and digital literacy. 

Mutable factors play a critical role in shaping public attitudes and behaviors as they are able to 

be modified through targeted public policies (Moussaïd et al., 2013). Unlike static factors, which 

are inherent and unchangeable, mutable factors can be influenced through strategic interventions 

and policy initiatives (Arias, 2018). As previously mentioned, one such mutable factor that could 

potentially be used to increase older adults’ reported likelihood of endorsement of autonomous 

vehicles is technological savviness (Zoaktafi et al., 2023, Mesch & Dodel, 2022, Nair & Bhat, 

5 

2021). Increasing technological savviness among older adults is hypothesized to significantly 

enhance their likelihood of adopting and endorsing autonomous vehicles (Carr et al., 2023). As 

older adults become more adept at using digital technologies, their comfort and confidence in 

engaging with new technological innovations, such as autonomous vehicles, are likely to 

improve (Reimer, 2014). This correlation is supported by existing research that suggests 

technological savviness facilitates a more favorable attitude towards emerging technologies 

(Criollo-C et al., 2024).  

The primary goal of this research project is to test if technological savviness is positively 

associated with reported likelihood of adoption of autonomous vehicles. When it comes to 

adopting new technologies, older adults tend to exhibit markedly lower receptivity, a 

phenomenon influenced by the rapid pace of technological advancements (Broady et al., 2010). 

One framework for understanding receptivity, the Theory of Planned Behavior (TPB), has been 

used to analyze the reported likelihood of adoption of autonomous driving technologies in 

several previous studies (Nastjuk et3 al., 2020, Mesch & Dodel, 2022, Nair & Bhat, 2021). The 

TPB posits that an individual's intention to engage in a specific behavior is shaped by three key 

factors: attitude, subjective norms, and perceived behavioral control, accounting for behaviors 

that may not be entirely volitional (Rossmann, 2020).  

In the context of autonomous vehicle adoption, older adults' likelihood to adopt may be 

influenced by their openness to new experiences and general perceptions of technology. 

Technological savviness plays a pivotal role in shaping these attitudes, as individuals who are 

more technologically savvy tend to have more favorable views of technology (Jokisch et al., 

2020). Technologically savvy individuals are typically more comfortable and familiar with using 

technology, which leads them to perceive it as a valuable and efficient tool for daily tasks. 

6 

Consequently, they are more likely to embrace new technological advancements, including 

autonomous vehicles (Kavandi & Jaana, 2020). 

Subjective norms, referring to the perceived social pressure to adopt or avoid a particular 

behavior, play a significant role in older adults' adoption of autonomous vehicles (Rahman & 

Thill, 2023). If older adults perceive that their peers, family, or society view autonomous 

vehicles favorably, they may feel motivated to conform to these expectations (Zoaktafi & Gandy, 

2023). Tech-savvy individuals, with greater exposure to positive endorsements of technology 

through online and offline networks, are more likely to engage with content such as product 

reviews and user experiences that highlight innovation and digital adoption. This frequent 

exposure through an individual’s social circle can foster positive attitudes toward emerging 

technologies (Carr et al., 2023). In contrast, older adults with limited exposure to such 

endorsements may remain skeptical or uninformed, making subjective norms more critical in 

shaping their decisions regarding unfamiliar technologies (Zandieh & Acheampong, 2021). 

Perceived behavioral control refers to an individual's belief in their ability to successfully 

perform a given behavior. For older adults, factors such as physical abilities, access to 

autonomous vehicle services, and technological savviness can significantly influence their 

perceived behavioral control regarding autonomous vehicle adoption. Technological savviness, 

in particular, can enhance an individual's sense of control and confidence in using technology 

(Moxley et al., 2022). Technologically savvy individuals tend to exhibit higher levels of self-

efficacy, which reflects their confidence in operating digital devices, navigating software, and 

resolving technical issues (Moxley et al., 2022). This heightened sense of perceived control may, 

in turn, positively influence their intentions to adopt new technologies, such as autonomous 

vehicles, by reducing apprehensions and fostering a greater willingness to engage with such 

7 

innovations. Increased technological savviness can mitigate initial apprehensions related to the 

complexity and usability of autonomous vehicles by fostering a sense of familiarity and control 

(Zoaktafi et al., 2023).  

8 

 
HYPOTHESES 

This study proposes a path model, summarized in Figure 1, to explore how technological 

savviness mediates the relationship between age and the likelihood of adopting autonomous 

vehicles. This specific type of structural equation model allows for the simultaneous analysis of 

relationships among multiple variables (Lleras, 2005). Path models estimate parameters that 

indicate the strength and direction of these relationships, with directional arrows representing the 

influence of one variable on another (Zhang, Jing, & Xu, 2021). While path models are effective 

in modeling associative relationships, they do not directly establish causality (Naiseh et al., 

2024). The robustness of path analysis relies on several fundamental assumptions, such as the 

assumption of linear relationships, the need to account for all relevant variables without 

omission, and the requirement for precise measurement of variables with minimal error (Sharma 

& Mishra, 2021). 

Figure 1. Simplified path model of age and reported likelihood of adoption of autonomous 

vehicles mediated by technological savviness 

This figure above summarizes four hypotheses: 

9 

 
H1: There is a negative association between age and technological savviness, such that as age 

increases, technological savviness decreases. 

H2: There is a positive association between technological savviness and reported likelihood of 

adoption of autonomous vehicles, such that higher levels of technological savviness are 

associated with greater reported likelihood of adoption of autonomous vehicles. 

H3: There is a negative association between age and reported likelihood of adoption of 

autonomous vehicles, such that as age increases, the reported likelihood of adoption of 

autonomous vehicles decreases. 

H4: The relationship between age and reported likelihood of adoption of autonomous vehicles is 

mediated by technological savviness. Specifically, more technologically savvy individuals will 

have a higher reported likelihood of adoption of autonomous vehicles than less technologically 

savvy individuals. 

Messaging and Likelihood of Endorsement 

The secondary goal of this research project is to identify the associations between technological 

savviness and messages regarding the endorsement of Autonomous Vehicles. Persuasive 

communication strategies, such as ‘foot-in-the-door’ and ‘door-in-the-face’ techniques, have 

been previously successful messaging techniques for improving endorsement (Myrick et al., 

2019). The ‘foot-in-the-door’ technique, which involves securing a small initial commitment 

before requesting a larger one, could be effectively utilized to increase acceptance of 

autonomous vehicles (Garikapati & Shetiya, 2024). Conversely, the ‘door-in-the-face’ technique, 

which starts with a large, often unreasonable request followed by a more modest one, may also 

be leveraged to shift perceptions (Rana & Hossain, 2021). Both strategies capitalize on principles 

of compliance and cognitive consistency, potentially fostering a more favorable disposition 

10 

toward autonomous vehicles through incremental engagement or perceived concessions 

(Othman, 2021). The endorsement questions asked in this study, a combination of personal 

attitudes, use cases, predictions, and safety preferences, represent foot-in-the-door messages, as 

they reflect how the respondent feels about other individuals using Autonomous Vehicles 

(Garikapati & Shetiya, 2024). This study proposes a path model in Figure 2 to explore how 

technological savviness mediates the relationship between age and the likelihood of endorsing 

autonomous vehicles. 

Figure 2. Simplified path model of age and endorsement of autonomous vehicle items mediated 

by technological savviness 

This figure above summarizes four hypotheses: 

H5: There is a negative association between age and technological savviness, such that as age 

increases, technological savviness decreases. 

11 

 
H6: There is a positive association between technological savviness and endorsement of 

autonomous vehicles, such that higher levels of technological savviness are associated with 

greater likelihood of endorsement of autonomous vehicles. 

H7: There is a negative association between age and endorsement of autonomous vehicles, such 

that as age increases, the endorsement of autonomous vehicles decreases. 

H8: The relationship between age and endorsement of autonomous vehicles is mediated by 

technological savviness. Specifically, more technologically savvy individuals will have a greater 

likelihood of endorsement of autonomous vehicles than less technologically savvy individuals. 

12 

Data Source 

METHODS 

This study employs a secondary analysis of data from the American Trends Panel, conducted 

between November 1st and November 7th, 2021 (Pew Research Center, 2023). The American 

Trends Panel, designed by the Pew Research Center, is a nationally representative longitudinal 

panel of U.S. adults, aimed at collecting data on a range of social, political, and economic issues. 

The panel is primarily conducted online, with internet access provided to participants as needed, 

ensuring broad accessibility. Recruitment for the panel involves a combination of random 

sampling and targeted recruitment methods. Initially, a random sample of U.S. adults is selected 

to complete a survey and are then invited to join the panel. To ensure the panel remains 

representative of the broader U.S. population, Pew Research Center applies statistical weighting 

to the data. These weights adjust for demographic factors, internet access, and potential non-

response biases, enhancing the accuracy and generalizability of the findings. 

Technological savviness 

The 2021 Pew Research American Trends Panel measured respondents’ internet use frequency 

with options “Almost constantly”, “Several times a day,” “About once a day,” “Several times a 

week,” or “Less often”. To examine the impact of very frequent internet use, this measure was 

binarized into two groups: “Almost constantly” versus all other frequencies. This binarization 

highlights a group that represents a distinct level of digital engagement. High-frequency internet 

users, defined by their “Almost constantly” engagement, exhibit a level of digital reliance and 

familiarity that may be fundamentally different in kind, not just degree, from other users. This 

distinction in degree may correlate with higher levels of technological savviness (Lin & Yu, 

2023, Kumi-Yeboah et al., 2020).  

13 

Separating “Almost constantly” users from all others is theoretically grounded, aligning with 

principles from the Technology Acceptance Model (TAM). According to TAM, individuals who 

frequently engage with digital tools are more likely to perceive technology as both useful and 

easy to use, which may increase their openness to adopting new technologies. By isolating this 

group, researchers can investigate a cohort likely to display behavioral traits aligned with high 

technological savviness (Mesch & Dodel, 2022). In addition, the validity of internet use as a 

proxy for technological savviness can vary by age. For younger individuals, frequent internet use 

is normative, reflecting routine familiarity rather than distinguishing technological savviness 

(Yan et al., 2023). Among this group, frequent internet engagement tends to be more habitual 

and integrated (Jokisch et al., 2022). In contrast, for older adults, frequent internet use is often 

more deliberate, and as such, it can more accurately reflect technological savviness (Heponiemi 

et al., 2023). In this demographic, intensive digital engagement can indicate a greater openness to 

exploring digital solutions and a significant level of comfort with technology relative to peers 

who use the internet less often (Lin & Yu, 2023). 

Reported likelihood of adoption of Autonomous Vehicles 

The Pew Research 2021 survey of the American Trends Panel began the autonomous vehicle 

question block with the item “How much have you heard or read about driverless passenger 

vehicles?” with options “A lot, A little, Nothing at all”. Respondents who answered “Nothing at 

all” were not surveyed on any further items in the autonomous vehicle question block. 

For respondents who had some measure of prior knowledge about autonomous vehicles as 

measured in the initial item, additional items were asked that measured reported likelihood of 

adoption and endorsement. The reported likelihood of adoption measure was identified as 

individuals reporting that they “personally want[ed] to ride” autonomous vehicles. For 

14 

endorsement measures, personal attitudes, use cases, and predictions in regard to autonomous 

vehicles were collected. Personal attitude items included measures of “how comfortable” the 

respondent would feel “sharing the road with them” as well as their attitudes regarding the 

societal good of autonomous vehicles. Use case items included respondents’ attitudes toward 

autonomous vehicles performing a variety of delivery and public transport uses. Prediction items 

asked about how respondents believed autonomous vehicles would effect “the number of people 

killed or injured in traffic accidents” as well as predictions about the independence of “older 

adults and people with disabilities”, driving stress, and hacking risk. Details on the wordings of 

items and the options given are listed below. 

Adoption Item 

Reported likelihood of adoption is measured with one item: 

1.  “Would you personally want to ride in a driverless passenger vehicle, if you had the 

opportunity?” with options “Definitely want, Probably want, Probably NOT want, 

Definitely NOT want” and higher values indicating a higher desire to ride. 

Endorsement Items 

Likelihood of endorsement was measured by four items: 

Personal attitudes were measured with two items: 

1.  “If the use of driverless passenger vehicles became widespread, how comfortable would 

you feel sharing the road with them?” with the options “Extremely comfortable, Very 

comfortable, Somewhat comfortable, Not too comfortable, Not comfortable at all” and 

higher values indicating a higher comfortability sharing the road. 

15 

2.  “Do you think widespread use of driverless passenger vehicles would be a…” with 

options “Good idea for society, Bad idea for society, Not sure” and higher values 

indicating belief that driverless vehicles are better for society. 

Use cases for autonomous vehicles was measured with the item “The technology used to operate 

driverless passenger vehicles could be used for a number of purposes. Would you favor or 

oppose the use of this technology in each of the following purposes?” for topics  

-  Taxis and ride-sharing vehicles 

-  Buses for public transportation  

with options “Favor, Oppose, Not sure, No answer” and higher values indicating a higher degree 

of favorability. 

Predictions regarding autonomous vehicles were measured with two items: 

1.  “If the use of driverless passenger vehicles becomes widespread, do you think that 

would…”  

- 

“Increase the number of people killed or injured in traffic accidents, Decrease 

the number of people killed or injured in traffic accidents, Not make much 

difference” with higher values indicating a belief in a lower number of traffic 

deaths. 

2. “If the use of driverless passenger vehicles becomes widespread, do you think each of the 

following would happen?”  

-  Older adults and people with disabilities will be able to live more 

independently 

-  Getting from place to place would be less stressful 

16 

-  The computer systems in driverless passenger vehicles would be easily hacked 

in ways that put safety at risk 

with options “Definitely would happen, Probably would happen, Probably would NOT 

happen, Definitely would NOT happen” and higher values indicating a higher belief the 

scenario would happen. 

Covariates 

Items for respondents’ socio-demographics are included in the analysis. Specifically included are 

predictors such as gender (A man vs A woman, In some other way), racial/ethnic groups (White 

non-Hispanic vs Black non-Hispanic, Hispanic, Other, Asian non-Hispanic), income brackets 

(less than $30,000 vs $30,000 to less than $40,000, $40,000 to less than $50,000, $50,000 to less 

than $60,000, $60,000 to less than $70,000. $70,000 to less than $80,000, $80,000 to less than 

$90,000, $90,000 to less than $100,000, $100,000 or more), and education levels (H.S. graduate 

or less vs Associate, College Grad, Postgrad). Additionally, political affiliation (Conservative vs 

Moderately Conservative, Moderately Liberal, Liberal) was included to ensure that the findings 

would be relevant across the political spectrum, enhancing the generalizability of the results. The 

survey responses were binarized into dichotomous variables.  

Path Model 

First, R, an open-source statistical analysis software, was used to describe the sample’s frequency 

and weighted percentages. Next, demographic variables were recoded into dummy variables. 

Lavaan, an R package, was employed to estimate the indirect effects of independent variables on 

dependent variables through a mediator variable to assess the mediation hypotheses. The analysis 

involved specifying a path model where the relationships between variables were represented by 

paths that capture direct and indirect effects. The model was specified to examine the direct 

17 

effects of age and frequency of internet use, along with the forementioned covariates. The 

indirect effects were captured by regressing internet access on age categories and defining 

interaction terms (Indirect Effect ages 30-49, Indirect Effect ages 50-64, Indirect Effect ages 

65+) as products of the age coefficients and internet access. Total effects (Total Effect ages 30-

49, Total Effect ages 50-64, Total Effect ages 65+) were calculated by summing the direct and 

indirect effects for each age group.  

By applying bootstrapping techniques within Lavaan, estimates of the indirect effects and their 

corresponding confidence intervals were obtained. The model was estimated using the Weighted 

Least Squares Mean and Variance-adjusted (WLSMV) estimator, which is suited for handling the 

non-normality and ordinal nature of the dependent variables. The WLSMV estimator addresses 

missing data through a full information approach, using all available data in the estimation 

process without resorting to pairwise deletion (Enders, 2010). This method allows for the 

retention of cases with partial missing data, ensuring the maximum use of available information. 

Sampling weights, applied using the "weight" variable provided by Pew, adjust for unequal 

selection probabilities. For the endorsement model, represented by Figure 2, the error terms of 

the individual endorsement items were correlated to account for the possibility that responses to 

any single item are likely to be correlated with responses on the other. The R code for both the 

Adoption and Endorsement path models is in the appendix. 

In this study, odds ratios were calculated using R with the generalized linear model function. 

This allows access to the odds ratio for a direct relationship between predictors and the outcome. 

A sample R code for calculating odds ratios is provided in the appendix. Effect sizes quantify the 

magnitude of relationships or differences between variables, providing a measure of ‘practical’ 

significance rather than statistical significance (Pek & Flora, 2018). In this study, the large 

18 

sample size of 5,153 participants results in high statistical power, allowing for the detection of 

small effects. However, as statistical significance alone does not indicate the importance of an 

effect, effect sizes can help determine whether an effect warrants attention (Pek & Flora, 2018). 

They assist in understanding the practical implications of research findings, standardizing them 

across different studies and populations for correct comparison. For binary outcomes, a common 

effect size measure is the odds ratio, which compares the odds of an event occurring in one group 

to the odds in another group, offering a standardized measure of effect. The odds ratio is 

considered statistically significant when its 95% confidence interval does not include 1 (Pek & 

Flora, 2018). Specific odds ratio values correspond to other established benchmarks of effect 

size, such as Cohen’s d (Chen et al., 2010). The transformation formula Cohen’s d to odds ratio 

is seen below (Sánchez-Meca et al., 2003). 

log odds ratio =

𝑑𝜋

√3

Fit statistics do not convey meaningful information about model adequacy in this study as the 

model was saturated, meaning the model has zero degrees of freedom and could not be 

statistically tested for fit. In this model, the number of estimated parameters matches the number 

of unique elements in the covariance matrix, and as such fit indices such as the Comparative Fit 

Index (CFI) and Root Mean Square Error of Approximation (RMSEA) indicate a perfect fit. The 

CFI equals 1, indicating that the model fully reproduces the observed data, and the RMSEA 

equals 0, reflecting an absence of residual discrepancy between the model and the data.  

19 

 
Sample Characteristics 

RESULTS 

Table 1 summarizes the characteristics of the sample, comprising of 5,153 individuals. Of these 

approximately 8.7% of participants were aged 18-29, 32.3% aged 30-49, 29.1% aged 50-64, and 

29.6% aged 65 and older. In terms of race and ethnicity, the majority of the sample is White non-

Hispanic (63.2%), followed by Hispanic individuals (15.7%), Black non-Hispanic individuals 

(11.4%), Asian non-Hispanic individuals (5.5%), and those categorized as Other (2.8%). Gender 

distribution is nearly equal, with 46.9% identifying as men, 51.9% as women, and 0.7% 

identifying in another way. Educational attainment varies, with 21.7% holding postgraduate 

degrees, 32.2% having completed college, 30.0% holding an associate degree, and 16.1% 

possessing a high school diploma or less. The mean annual family income was $58,890. The 

political orientation of the sample includes 25.8% identifying as Conservative, 29.0% as Liberal, 

25.7% as Moderately Conservative, and 19.6% as Moderately Liberal.  

20 

 
 
Table 1 Sample Characteristics   

Variable     
Age     
           18-29     
           30-49     
           50-64     
           65+     
Gender    
           A man    
           A woman     
           In some other way    
Party    
           Conservative    
           Moderately Conservative 
           Moderately Liberal 
           Liberal    
Family income    
           Less than $30,000    
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category    
           H.S. graduate or less     
           Associate    
           College Grad    
           Postgrad  
Race-Ethnicity     
           White non-Hispanic    
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    
Note: % = percent 

Sample (5,153) 
Unweighted Frequency  Weighted % 
448 
1,663 
1,498 
1,524 

8.7 
32.3 
29.1 
29.6 

2,295 
2,816 
30 

1,589 
1,369 
1,469 
1,675 

822 
478 
433 
430 
365 
371 
267 
316 
1,437 

873 
1,506 
1,652 
1,105 

3,581 
427 
729 
165 
178 

46.9 
51.9 
0.7 

25.8 
25.7 
19.6 
29.0 

23.5 
11.1 
8.7 
8.6 
6.6 
6.1 
5.1 
5.1 
20.4 

16.1 
30.0 
32.2 
21.7 

63.2 
11.4 
15.7 
2.8 
5.5 

Table 2 presents the distribution of respondents’ internet frequency, the reported likelihood of 

adoption, and endorsement items. For internet frequency, the majority of respondents reported 

using the internet several times a day (45.4%), with 39.3% indicating almost constant use. The 

21 

  
  
  
  
  
  
  
  
  
  
 
endorsement items related to autonomous vehicle adoption show varying levels of interest: 

13.8% of respondents definitely want to ride an autonomous vehicle, while 32.3% definitely do 

not want to ride. In terms of comfort sharing the road with autonomous vehicles, 34.4% of 

respondents feel somewhat comfortable, and 26.7% feel not too comfortable. Regarding 

widespread societal use, 44.5% of respondents viewed autonomous vehicles as a bad idea for 

society. For specific applications of autonomous vehicles, 40.7% of respondents favored their 

use for taxis and ride-sharing, whereas 43.4% opposed their use for buses. When asked about 

potential impacts on traffic deaths, 39.4% believed autonomous vehicles would decrease traffic 

deaths, while 31.1% felt there would be little difference. Regarding the perceived benefits for 

older adults and transportation stress, 49.3% of respondents thought autonomous vehicles would 

likely increase independence for older adults, and 40.1% believed they would reduce 

transportation stress. 50.8% of respondents thought there was a significant risk of hacking, with 

20.1% disagreeing with this assessment. 

22 

 
 
Table 2 Distribution of Respondent’s Internet Frequency, Adoption, and Endorsement Items 

(13.8) 
(22.8) 
(32.3) 
(30.3) 

690   
1,171   
1,716   
1,532   

Unweighted Frequency 
2,334 
2,350 
273 
134 
57 

Path Model Endogenous Variables                                                                         N=5,153 
Weighted % 
Internet Frequency 
(39.3) 
           Almost constantly 
(45.4) 
           Several times a day 
(5.3) 
           About once a day 
(3.3) 
           Several times a week 
(2.4) 
           Less often 
Personally want to ride   
          Definitely want   
          Probably want   
          Probably NOT want   
          Definitely NOT want   
Sharing the road   
         Extremely comfortable   
         Very comfortable   
         Somewhat comfortable   
         Not too comfortable   
         Not comfortable at all   
Widespread Use   
          Good idea for society   
          Bad idea for society   
          Not sure   
          Missing   
Used for Taxi's and ride-sharing   
         Favor   
         Oppose   
         Not sure   
Used for Busing   
         Favor   
         Oppose   
         Not sure   
Traffic Deaths  
         Increase  
         Decrease  
         Not much different  

348  
690  
1,752  
1,451  
894  

(7.1) 
(13.6) 
(34.4) 
(26.7) 
(17.9) 

1,447   
2,290   
1,404   
12  

(26.4) 
(44.5) 
(28.9) 
(0.2) 

1,789  
2,220  
1,124  

1,380  
2,077  
1,572  

2,154  
1,695  
1,283  

(34.4) 
(43.4) 
(21.6) 

(27) 
(39.4) 
(31.1) 

(40.7) 
(33.8) 
(25) 

23 

   
   
   
   
    
    
   
   
   
   
   
   
 
Table 2 (cont’d) 

Older adults more independent   

         Definitely would happen   
         Probably would happen   
         Probably would NOT happen   
         Definitely would NOT happen   
Transportation less stressful   
        Definitely would happen   
        Probably would happen   
        Probably would NOT happen   
        Definitely would NOT happen   
Hacking security risk 
        Definitely would happen   
        Probably would happen   
        Probably would NOT happen   
        Definitely would NOT happen   
Note: % = percent 

Reported Likelihood of Adoption 

1,185  
2,604  
1,111  
218  

818  
1,976  
1,778  
537  

1,238 
2,750 
1,051 
70 

(22.4) 
(49.3) 
(22.6) 
(4.9) 

(15.7) 
(40.1) 
(33.3) 
(9.7) 

(25.3) 
(50.8) 
(20.1) 
(2.5) 

The results for the demographic of interest, adults aged 65 and older, are presented in Figure 3. 

The mediation analysis revealed that technological savviness serves as a partial mediator in the 

relationship between age and the adoption of autonomous vehicles. Specifically, these findings 

suggest that interventions aimed at enhancing technological savviness among older adults may 

alleviate the negative impact of age on their willingness to adopt autonomous vehicles. A more 

detailed breakdown of the results is provided below. 

24 

 
 
   
   
  
 
Figure 3. Simplified path model with path weights and significances of Age and Reported 

Likelihood of Adoption of Autonomous Vehicles mediated by Technological Savviness 

H1: Association between Age and Technological savviness 

H1, summarized below in Table 3, posited that increased age would be associated with a lower 

level of technological savviness. The analysis supports this hypothesis, showing significant 

negative associations between age and technological savviness across all older age groups when 

compared to the reference group of individuals ages 18-29. 

For respondents aged 30-49, the negative coefficient of b = -0.133 (p = 0.002) with an odds ratio 

of 0.622 (95% CI [0.545, 0.709]) indicates a modest but statistically significant decrease in 

technological savviness relative to those aged 18-29. This finding suggests that, even among 

those in early middle age, technological savviness is somewhat lower than in younger adults. The 

odds ratio of 0.622 corresponds to a small effect size (Cohen’s d ≈ 0.2), indicating that the 

difference in technological savviness between these two groups, while statistically significant, is 

relatively small in practical terms. 

25 

 
The effect is even more pronounced in the 50-64 age group, where the coefficient of b = -0.295 

(p < 0.001) and an odds ratio of 0.291 (95% CI [0.252, 0.337]) indicate a substantial reduction in 

technological savviness compared to the 18-29 group. This odds ratio of 0.291 represents a 

moderate to large effect (Cohen’s d ≈ 0.8), highlighting a clear and practical difference in 

technological savviness between these age groups. Individuals in this age range appear to be 

much less technologically savvy than their younger counterparts. 

For respondents aged 65 and older, the effect is strongest, with a coefficient of b = -0.462 (p < 

0.001) and an odds ratio of 0.104 (95% CI [0.087, 0.124]). This very low odds ratio suggests 

that, in comparison to younger adults, individuals in the 65 and older group show markedly 

lower levels of technological savviness. With an odds ratio of 0.104, corresponding to a very 

large effect size (Cohen’s d ≈ 0.8), the difference between this age group and younger adults is 

both statistically and practically significant, indicating a pronounced gap in technological 

savviness. 

The findings consistently support the hypothesis that technological savviness declines with age. 

Each older age group shows a significant and progressively greater reduction in technological 

savviness compared to those aged 18-29, highlighting a clear age-related trend in technology-

related skills and familiarity. 

Other Findings 

Gender was not a significant predictor of technological savviness (b = -0.022, p = 0.325), 

suggesting that both men and women may have similar levels of technological savviness. 

However, political attitudes showed significant associations with technological savviness. 

Moderate liberal attitudes (b = 0.092, p = 0.004) and liberal attitudes (b = 0.161, p < 0.001) were 

26 

positively correlated with higher internet usage, indicating that more liberal individuals are likely 

to be more technologically savvy. 

Family income results were mixed. Specifically, individuals with incomes ranging from $30,000 

to less than $40,000 (b = 0.094, p = 0.016) and those with incomes ranging from $50,000 to less 

than $60,000 (b = 0.126, p = 0.013) exhibited higher technological savviness compared to those 

earning less than $30,000. Similarly, individuals with an income of $100,000 or more 

demonstrated the highest levels of technological savviness (b = 0.136, p < 0.001), with a 1.88 

times greater likelihood. 

Education level was also a significant predictor of technological savviness. Postgraduates (b = 

0.065, p = 0.038) were more likely to be technologically savvy than those with a high school 

education or less. Additionally, individuals identifying as Black non-Hispanic (b = 0.095, p = 

0.014) and Hispanic (b = 0.248, p < 0.001) exhibited higher technological savviness compared to 

their White non-Hispanic corollaries. 

Table 3 summarizes the results of the path model relationships predicting technological savviness 

as well as the Odds ratios of these relationships. 

27 

 
 
Table 3 Mediator Regression of Technological Savviness on Age and Socio-Demographic 
Predictors 
Variable     

odds ratio  2.5% 

p 

b 

CI 

97.5% 
CI 

0.816 
0.193 

0.898 
0.325 

-0.022 
-0.277 

1.034 
1.331 
1.819 

1.201 
1.517 
2.077 

0.545 
0.252 
0.087 

0.622 
0.291 
0.104 

0.024 
0.092  ** 
0.161  *** 

-0.133  ** 
-0.295  *** 
-0.462  *** 

Age vs 18-29 
           30-49     
           50-64     
           65+     
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative   
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School 
Graduate or Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    
Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

0.094  * 
-0.013 
0.126  * 
0.055 
0.032 
0.024 
-0.018 
0.136  ** 

0.095  * 
0.248  *** 
0.085 
0.056 

1.41 
1.014 
1.449 
1.282 
1.082 
1.253 
1.122 
1.877 

1.183 
0.837 
1.201 
1.043 
0.875 
0.99 
0.888 
1.605 

-0.035 
0.034 
0.065  * 

1.394 
2.111 
1.285 
0.963 

1.198 
1.732 
1.012 
0.765 

0.911 
1.107 
1.445 

0.779 
0.97 
1.245 

0.709 
0.337 
0.124 

0.988 
0.535 

1.395 
1.73 
2.373 

1.682 
1.227 
1.748 
1.575 
1.336 
1.584 
1.414 
2.197 

1.065 
1.263 
1.679 

1.623 
2.582 
1.633 
1.21 

H2: Association Between Technological Savviness and Reported Likelihood of Adoption of 

Autonomous Vehicles 

H2, as outlined in Table 4, proposed that higher levels of technological savviness would correlate 

with a greater likelihood of adopting autonomous vehicles. The results confirm this hypothesis. 

Compared to individuals in the reference group who reported using the internet “Several times a 

28 

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
day,” “About once a day,” “Several times a week,” or “Less often” those with “Almost constant” 

internet usage, a proxy for higher technological savviness, showed a significant positive 

association with autonomous vehicle adoption. Specifically, the positive coefficient (b = 0.167, p 

= 0.021) and an odds ratio of 1.32 (95% CI [1.14, 1.53]) suggest that individuals with higher 

technological savviness are substantially more likely to adopt autonomous vehicles than those in 

the reference group. The odds ratio of 1.32 indicates a small-to-moderate effect size (Cohen's d ≈ 

0.2), meaning that individuals with higher technological savviness have about 32% higher odds 

of adopting autonomous vehicles compared to those with less frequent internet usage. 

H3: Association Between Age and Reported Likelihood of Adoption of Autonomous Vehicles 

H3 explored whether increased age would be associated with a lower reported likelihood of 

adopting autonomous vehicles. The results outlined in Table 4 reveal a significant negative 

relationship between age and the likelihood of autonomous vehicle adoption in older age groups. 

Compared to the reference group of ages 18–29, respondents aged 50–64 exhibited a notably 

lower likelihood of adopting autonomous vehicles, as shown by the coefficient (b = -0.49, p < 

0.001) and an odds ratio of 0.45 (95% CI [0.36, 0.56]). The odds ratio of 0.45 suggests a 

moderate effect size (Cohen's d ≈ 0.5), indicating that individuals aged 50–64 have about 55% 

lower odds of adopting autonomous vehicles than those aged 18-29. 

Similarly, respondents aged 65 and older showed an even lower likelihood, with a coefficient of 

b = -0.544 (p < 0.001) and an odds ratio of 0.41 (95% CI [0.32, 0.52]). The odds ratio of 0.41 

represents a large effect size (Cohen's d ≈ 0.8), meaning that individuals aged 65 and older have 

about 59% lower odds of adopting autonomous vehicles compared to those aged 18-29. These 

results indicate a clear trend of decreased adoption likelihood with increasing age. 

29 

In contrast, respondents aged 30–49 did not demonstrate a statistically significant association, 

with a coefficient of b = -0.138 (p = 0.213) and an odds ratio of 0.81 (95% CI [0.67, 0.97]). The 

odds ratio of 0.81 is close to 1, indicating no significant difference in the odds of adoption 

between this group and the 18-29 reference group, suggesting that age has less impact on 

adoption likelihood in this middle-age range. 

Other Findings 

Female participants were significantly less likely to report higher likelihood to adopt autonomous 

vehicles compared to male participants (b = -0.577, p < 0.001). Political orientation also had a 

significant impact. Moderate conservatives (b = 0.24, p = 0.027), moderate liberals (b = 0.379, p 

< 0.001), and liberals (b = 0.516, p < 0.001) were all more likely to report higher likelihood of 

adopting autonomous vehicles compared to conservatives. Results for income were mixed. While 

most income categories did not show significant effects, individuals with an income of $60,000 

to less than $70,000 (b = -0.312, p = 0.027) and $90,000 to less than $100,000 (b = -0.403, p = 

0.009) showed significant negative associations compared to individuals making less than 

$30,000. Higher education levels were significantly associated with a greater likelihood of 

adopting autonomous vehicles. College graduates (b = 0.27, p = 0.001) and postgraduates (b = 

0.374, p < 0.001) were more likely to favor autonomous vehicles than high school graduates or 

less. 

Table 4 summarizes the results of the path model relationships predicting the likelihood of 

adopting autonomous vehicles as well as the Odds ratios of these relationships. 

30 

 
 
odds ratio  2.5% CI  97.5% CI 

* 

b 

1.144 

1.324 

0.167 

0.336 
0.868 

0.240 
0.379 
0.516 

0.668 
0.362 
0.321 

-0.577  ***  0.386 
1.957 
0.355 

-0.138 
0.807 
-0.490  ***  0.449 
-0.544  ***  0.407 

Table 4 Direct Effect Regression of Autonomous Vehicle Adoption Likelihood on Age, 
Technological Savviness, and Socio-Demographic Variable 
Variable     
p 
Age vs 18-29 
           30-49     
           50-64     
           65+     
Internet vs Several times a day, About once 
a day, Several times a week, or Less often 
           Almost constantly 
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative   
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School 
Graduate or Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    
Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

-0.184 
-0.126 
-0.099 
-0.312  * 
-0.105 
0.097 
-0.403  ** 
-0.023 

0.73 
0.809 
0.836 
0.597 
0.844 
1.174 
0.518 
0.96 

0.566 
0.617 
0.640 
0.439 
0.628 
0.86 
0.365 
0.769 

1.489 
* 
***  1.871 
***  2.352 

-0.151 
-0.036 
0.123 
0.343 

0.783 
0.925 
1.226 
1.748 

0.630 
0.698 
0.867 
1.264 

** 
***  1.85 

1.022 
1.30 
1.50 

0.145 
0.270 
0.374 

1.200 
1.55 
1.941 

1.277 
1.563 

0.973 
0.556 
0.516 

1.533 

0.442 
4.871 

1.846 
2.260 
2.853 

0.939 
1.059 
1.090 
0.806 
1.132 
1.601 
0.728 
1.198 

1.594 
1.880 
2.282 

0.971 
1.225 
1.730 
2.422 

31 

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
 
 
 
H4: Mediation Effect of Technological savviness on the Association between Age and Reported 

likelihood of adoption of Autonomous Vehicles 

H4, summarized in Table 5, posited that technological savviness mediates the relationship 

between age and the likelihood of adopting autonomous vehicles. Specifically, it hypothesized 

that individuals with greater technological savviness would have a higher reported likelihood of 

adopting autonomous vehicles than those with less technological savviness. The mediation 

analysis reveals significant indirect effects. Specifically, the mediation effects for the different 

age groups show that technological savviness partially mediates the relationship between age and 

reported likelihood of adoption when compared to the reference group of ages 18–29. For ages 

30–49, the indirect effect is negative but not significant (b = -0.022, p = 0.055). For ages 50–64 

and ages 65 and over, the mediation effects are stronger and significant (b = -0.049, p = 0.031 

and b = -0.077, p = 0.027, respectively). 

The significant positive effect of technological savviness on reported likelihood of adoption, 

shown in Table 4 (b = 0.167, p = 0.021), indicates that higher levels of technological savviness 

are associated with an increased likelihood of adopting autonomous vehicles. Additionally, the 

indirect effects for the older age groups (ages 50–64 and ages 65 and over) confirm that 

technological savviness plays an important role in influencing adoption among older adults 

compared to participants ages 18–29, with the effect being strongest for the oldest group. 

The total effects of age on the reported likelihood of adopting autonomous vehicles, which 

encompass both direct and indirect pathways, support the hypothesized relationships. When 

compared to the reference group of ages 18–29, the total effect for ages 30–49 was not 

significant (b = -0.161, p = 0.148). In contrast, the total effects for the age groups 50–64 and ages 

65 and over were substantial and negative (b = -0.539, p < 0.001 and b = -0.621, p < 0.001, 

32 

respectively). These results suggest that, overall, older adults face significant age related barriers 

to adopting autonomous vehicles, with technological savviness functioning as a mitigating factor. 

Table 5 summarizes the indirect and total effects of the path model relationships between age, 

technological savviness, and the likelihood of adopting autonomous vehicles.  

Table 5 Indirect and Total Effects of the Adoption Path Model 

p 

Path 
Indirect 30-49 
Indirect 50-64 
Indirect 65+ 
Total 30-49 
Total 50-64 
Total 65+ 
Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001 

b 
-0.022 
-0.049 
-0.077 
-0.161 
-0.539 
-0.621 

* 
* 

*** 
*** 

Endorsement Items 

The results presented in Figure 4 highlight the relationships between age, technological 

savviness, and perceptions of autonomous vehicles. The analysis reveals that technological 

savviness mediates the influence of age on attitudes toward autonomous vehicles. The mediation 

analysis revealed that technological savviness serves as a mediator in the relationship between 

age and endorsement items of autonomous vehicles. In particular, technological savviness 

partially mediates the relationship between age and perceptions of autonomous vehicles as 

reducing transportation stress, being acceptable for sharing the road with, and being used as taxis 

and buses. Furthermore, technological savviness fully mediates the relationship between age and 

the perception of autonomous vehicles as beneficial for society and supporting older adults’ 

independence. These findings suggest that efforts to enhance technological savviness may 

alleviate age-related barriers and foster more positive attitudes toward the adoption of 

autonomous vehicles. A more detailed examination of these relationships is presented below. 

33 

  
  
Figure 4. Simplified path model with path weights and significances of Age and Reported 

Likelihood of Adoption of Autonomous Vehicles mediated by Technological Savviness 

H5: Association between Age and Technological savviness 

H5, summarized below in Table 6, posited that age would be negatively associated with 

technological savviness, with older adults exhibiting lower levels of technological savviness 

compared to younger individuals. The analysis supports this hypothesis, showing significant 

negative associations between age and technological savviness across all older age groups when 

compared to the reference group of individuals aged 18-29. 

For respondents aged 30-49, the coefficient of b = -0.132 (p < 0.001) and an odds ratio of 0.622 

(95% CI [0.545, 0.709]) indicate a significant reduction in technological savviness relative to 

those aged 18-29. The odds ratio of 0.622 suggests a small effect size (Cohen’s d ≈ 0.2), meaning 

that while the difference is statistically significant, it is relatively modest in practical terms. 

The negative relationship between age and technological savviness becomes more pronounced in 

the 50-64 age group, with a coefficient of b = -0.290 (p < 0.001) and an odds ratio of 0.291 (95% 

CI [0.252, 0.337]). This indicates a substantial decrease in technological savviness compared to 

34 

 
the 18-29 group. The odds ratio of 0.291 corresponds to a moderate to large effect size (Cohen’s 

d ≈ 0.8), demonstrating a clear and practical difference in technological savviness between these 

age groups. 

For respondents aged 65 and older, the coefficient of b = -0.457 (p < 0.001) and an odds ratio of 

0.104 (95% CI [0.087, 0.124]) indicate a dramatic reduction in technological savviness. The odds 

ratio of 0.104 represents a very large effect size (Cohen’s d ≈ 0.8), suggesting a pronounced gap 

in technological savviness between this age group and younger adults. These findings strongly 

support the hypothesis that technological savviness declines with age, with the largest disparities 

observed among the oldest age group. 

Other Findings 

Gender was not a significant predictor of technological savviness (b = -0.022, p = 0.325), 

suggesting no meaningful difference between men and women in their technological skills. 

However, political party affiliation was significantly associated with technological savviness. 

Respondents identifying as moderately liberal (b = 0.089, p = 0.003) and liberal (b = 0.154, p < 

0.001) were more likely to exhibit higher technological savviness compared to conservatives. 

Family income also showed significant associations with technological savviness. Specifically, 

individuals with incomes between $30,000 to $40,000 (b = 0.089, p = 0.016) and $50,000 to 

$60,000 (b = 0.133, p = 0.023) demonstrated higher levels of technological savviness than those 

with incomes under $30,000. The highest odds of technological savviness were observed among 

individuals earning $100,000 or more (b = 0.138, p = 0.004). Education level was another 

significant predictor of technological savviness. Respondents with postgraduate degrees (b = 

0.063, p = 0.016) were more likely to be technologically savvy compared to those with a high 

school education or less. Race and ethnicity were also significant predictors. Black non-Hispanic 

35 

individuals (b = 0.110, p = 0.004) and Hispanic individuals (b = 0.249, p < 0.001) exhibited 

higher odds of technological savviness compared to White non-Hispanic respondents. 

Table 6 summarizes the results of the path model relationships predicting technological savviness 

as well as the odds ratios of these relationships. 

Table 6 Endorsement Path Model Technological Savviness Mediator Regressions 

odds ratio  2.5% CI  97.5% CI 

p 

b 

** 
*** 

0.898 
0.325 

0.816 
0.193 

*** 
*** 
*** 

-0.022 
-0.272 

1.201 
1.517 
2.077 

0.622 
0.291 
0.104 

1.034 
1.331 
1.819 

0.02 
0.089 
0.154 

0.545 
0.252 
0.087 

-0.132 
-0.29 
-0.457 

Variable 
Age vs 18-29 
30-49 
50-64 
65+ 
Gender vs A man 
A woman 
In some other way 
Party vs Conservative 
Moderately Conservative 
Moderately Liberal 
Liberal 
Family income vs Less than 
$30,000 
$30,000 to less than $40,000 
$40,000 to less than $50,000 
$50,000 to less than $60,000 
$60,000 to less than $70,000 
$70,000 to less than $80,000 
$80,000 to less than $90,000 
$90,000 to less than $100,000 
$100,000 or more 
Education level category vs High 
School Graduate or Less 
Associate 
College Grad 
Postgrad 
Race-Ethnicity vs White non-
Hispanic 
0.11 
Black non-Hispanic 
0.249 
Hispanic 
0.087 
Other 
0.041 
Asian non-Hispanic 
Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

0.089 
-0.021 
0.133 
0.061 
0.031 
0.029 
-0.02 
0.138 

1.183 
0.837 
1.201 
1.043 
0.875 
0.99 
0.888 
1.605 

1.41 
1.014 
1.449 
1.282 
1.082 
1.253 
1.122 
1.877 

1.198 
1.732 
1.012 
0.765 

1.394 
2.111 
1.285 
0.963 

0.779 
0.97 
1.245 

-0.04 
0.034 
0.063 

0.911 
1.107 
1.445 

** 
** 

** 

* 

* 

0.709 
0.337 
0.124 

0.988 
0.535 

1.395 
1.73 
2.373 

1.682 
1.227 
1.748 
1.575 
1.336 
1.584 
1.414 
2.197 

1.065 
1.263 
1.679 

1.623 
2.582 
1.633 
1.21 

36 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
H6: Association between Technological savviness and Endorsement of Autonomous Vehicles 

H6, summarized in the “Almost constantly” row in Table 7-10, proposed a positive relationship 

between technological savviness and favorable views on various aspects of autonomous vehicles. 

This hypothesis was largely supported, with findings indicating that individuals with higher 

levels of technological savviness were more likely to endorse the items. 

Individuals with higher technological savviness were more likely to express the belief that 

autonomous vehicles are beneficial for society, as shown by a significant positive coefficient (b = 

0.234, p = 0.001) and an odds ratio of 1.503 (95% CI [1.281, 1.763]). This odds ratio suggests 

that those with higher technological savviness are approximately 50% more likely to view 

autonomous vehicles as beneficial to society compared to those with lower technological 

savviness. The odds ratio of 1.503 represents a moderate effect size (Cohen's d ≈ 0.5). 

In addition, individuals with higher technological savviness were more inclined to view 

autonomous vehicles as a solution for reducing travel-related stress. A positive coefficient (b = 

0.175, p = 0.009) and an odds ratio of 1.192 (95% CI [1.073, 1.324]) indicate that those with 

greater technological savviness are approximately 19% more likely to believe that autonomous 

vehicles can reduce travel-related stress. The odds ratio of 1.192 represents a small to moderate 

effect size (Cohen's d ≈ 0.3), meaning that individuals with higher technological savviness are 

somewhat more likely to perceive autonomous vehicles as a stress-reducing solution. 

Higher technological savviness was also associated with stronger endorsement of using 

autonomous vehicles as taxis (b = 0.263, p < 0.001, odds ratio = 1.491, 95% CI [1.293, 1.718]) 

and buses (b = 0.150, p = 0.037, odds ratio = 1.297, 95% CI [1.119, 1.503]). For taxis, the odds 

ratio of 1.491 means that individuals with higher technological savviness are about 49% more 

likely to support using autonomous vehicles as taxis. The odds ratio of 1.491 represents a large 

37 

effect size (Cohen's d ≈ 0.8), meaning that individuals with higher technological savviness are 

significantly more likely to endorse autonomous vehicles as taxis compared to those with lower 

savviness. Similarly, the odds ratio of 1.297 for buses indicates a 30% increased likelihood of 

endorsing the use of autonomous vehicles as buses among those with higher technological 

savviness, with a moderate effect size (Cohen’s d ≈ 0.5). 

Individuals with greater technological savviness were also more likely to believe that 

autonomous vehicles could enhance the independence of older adults and people with 

disabilities, as indicated by a positive and significant coefficient (b = 0.292, p < 0.001) and an 

odds ratio of 1.211 (95% CI [1.099, 1.335]). The odds ratio of 1.211 represents a moderate effect 

size (Cohen's d ≈ 0.5), meaning that those with higher technological savviness are about 21% 

more likely to believe that autonomous vehicles could help vulnerable populations gain more 

independence compared to those with lower technological savviness. 

Some safety-related beliefs did not show a significant association with technological savviness. 

Technological savviness was not significantly associated with the belief that autonomous 

vehicles are less likely to crash (b = 0.055, p = 0.434, odds ratio = 1.084, 95% CI [0.924, 1.271]) 

or the perception that autonomous vehicles are susceptible to hacking (b = -0.004, p = 0.958, 

odds ratio = 0.996, 95% CI [0.903, 1.098]). These results suggest that technological savviness 

alone may not strongly influence beliefs about the safety or vulnerability of autonomous vehicles 

to hacking, implying that other factors may be more relevant in shaping safety perceptions 

related to autonomous vehicle technology. 

38 

 
 
 
H7: Association between Age and Endorsement of Autonomous Vehicles 

If the use of driverless passenger vehicles becomes widespread, do you think getting from 

place to place would be less stressful? 

Summarized in Table 7, age had a significant influence on the likelihood of endorsing the belief 

that getting from place to place in autonomous vehicles would be less stressful. Compared to 

individuals aged 18-29, respondents aged 30-49 were more likely to report that autonomous 

vehicles would make getting from place to place less stressful (b = -0.391, p = 0.001), with an 

odds ratio of 0.638 (95% CI [0.547, 0.744]), corresponding to a moderate effect size (Cohen’s d 

≈ 0.5). Similarly, respondents aged 50-64 were also more likely to endorse the belief that 

autonomous vehicles would reduce stress in transportation (b = -0.318, p = 0.009), with an odds 

ratio of 0.711 (95% CI [0.602, 0.84]), corresponding to a small to medium effect size (Cohen’s d 

≈ 0.2). 

Other Findings 

Gender was a significant predictor, with women (b = -0.465, p < 0.001) less likely than men to 

view autonomous vehicles as reducing stress in transportation. Political orientation also played a 

role, with liberals (b = 0.509, p < 0.001) more likely to view autonomous vehicles as making 

getting from place to place less stressful. Income was another influential factor, with individuals 

with incomes above $60,000 more likely to report that autonomous vehicles would reduce stress. 

Those with an income between $30,000 and $39,999 were slightly less likely (b = 0.005, p = 

0.969) to view autonomous vehicles as less stressful, and those with income levels above 

$50,000 showed similar trends. Education also had a significant effect, with individuals holding 

postgraduate degrees (b = 0.271, p = 0.001) more likely to view autonomous vehicles as 

reducing transportation stress. 

39 

If the use of driverless passenger vehicles became widespread, how comfortable would you 

feel sharing the road with them? 

Summarized in Table 7, age was a significant factor in the likelihood of feeling comfortable 

sharing the road with autonomous vehicles. Compared to individuals aged 18-29, respondents 

aged 30-49 were less likely to feel comfortable sharing the road with autonomous vehicles (b = -

0.449, p < 0.001), with an odds ratio of 0.463 (95% CI [0.367, 0.582]), corresponding to a 

medium effect size (Cohen’s d ≈ 0.5). Similarly, respondents aged 50-64 were also less likely to 

endorse the belief in sharing the road with autonomous vehicles (b = -0.379, p = 0.008), with an 

odds ratio of 0.537 (95% CI [0.416, 0.691]), corresponding to a medium effect size (Cohen’s d ≈ 

0.5). 

Other Findings  

Gender was a significant predictor, with women (b = -0.465, p < 0.001) less likely than men to 

feel comfortable sharing the road with autonomous vehicles. Political orientation also played a 

significant role, with liberals (b = 0.384, p < 0.001) more likely to feel comfortable sharing the 

road with autonomous vehicles. Income also showed notable effects, with those in the $30,000–

$39,999 income range less likely to feel comfortable sharing the road with autonomous vehicles 

(b = -0.274, p = 0.08). Education also influenced the likelihood of feeling comfortable sharing 

the road with autonomous vehicles. Individuals with an associate degree (b = 0.281, p = 0.025) 

were more likely to endorse this belief, as were those with postgraduate degrees (b = 0.314, p = 

0.001). 

Table 7 summarizes the results of the path model predicting the likelihood of endorsement items: 

that autonomous vehicles reduce the stress of driving, and that respondents feel comfortable 

sharing the road with autonomous vehicles.

40 

Table 7 Direct Effect Regression of Autonomous Vehicle Endorsement on Age, Technological Savviness, and Socio-Demographic Variable   

Make driving less stressful                                                   
p 

odds ratio  2.5% CI  97.5% CI  p 

odds ratio  2.5% CI  97.5% CI 

Comfortable to share the road with                                            

0.702 
0.547 
0.602 

** 

*** 

1.192 

** 
** 

0.662 
0.95 

0.803 
0.638 
0.711 

 Variable 
Age vs 18-29 
           30-49     
           50-64     
           65+     
Internet vs Several times a day, About once a day, Several times a week, 
or Less often 
           Almost constantly 
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative 
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School Graduate or Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    
Note = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

1.106 
1.06 
1.217 
0.916 
1 
1.165 
1.049 
1.025 

0.982 
0.866 
0.867 
1.512 

1.382 
1.62 
1.558 

1.048 
1.25 
1.261 

* 
*** 
*** 

** 

1.073 

0.6 
0.575 

1.181 
1.416 
1.357 

0.924 
0.873 
1.008 
0.738 
0.805 
0.917 
0.828 
0.873 

0.891 
1.094 
1.084 

0.838 
0.71 
0.669 
1.202 

0.919 
0.744  ** 
0.84  ** 

0.821 
0.463 
0.537 

0.686 
0.367 
0.416 

0.984 
0.582 
0.691 

1.324  ** 

1.474 

1.266 

1.718 

0.731  *** 
1.542 

1.615  ** 
1.855  *** 
1.79  *** 

1.322 
1.283 
1.469 
1.133  ** 
1.238 
1.476 
1.324 
1.204 

1.231  * 
1.428  *** 
1.465  ** 

1.149 
1.051 
1.114  * 
1.896 

0.489 
1.402 

1.666 
1.849 
1.706 

0.757 
0.767 
1.034 
0.516 
0.935 
0.904 
0.909 
0.816 

1.547 
1.852 
1.541 

0.884 
1.024 
0.54 
1.497 

0.422 
0.783 

1.316 
1.504 
1.385 

0.574 
0.571 
0.788 
0.36 
0.688 
0.637 
0.644 
0.651 

1.23 
1.534 
1.238 

0.696 
0.789 
0.345 
1.097 

0.566 
2.411 

2.109 
2.278 
2.106 

0.991 
1.021 
1.348 
0.726 
1.257 
1.262 
1.264 
1.024 

1.933 
2.232 
1.912 

1.112 
1.315 
0.809 
2.016 

41 

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
The technology used to operate driverless passenger vehicles could be used for a number of 

purposes. Would you favor or oppose the use of this technology in each of the following 

purposes: Taxis and ride-sharing vehicles? 

Summarized in Table 8, age showed a significant influence on the likelihood of endorsing the 

belief that autonomous vehicles could be used as taxis. Compared to individuals aged 18-29, 

respondents aged 30-49 were less likely to support this belief (b = -0.324, p = 0.006), with an 

odds ratio of 0.587 (95% CI [0.476, 0.724]), corresponding to a moderate effect size (Cohen’s d 

≈ 0.5). Similarly, those aged 50-64 also showed a decreased likelihood of endorsing the use of 

autonomous vehicles as taxis (b = -0.313, p = 0.012), with an odds ratio of 0.596 (95% CI 

[0.475, 0.747]), corresponding to a medium effect size (Cohen’s d ≈ 0.5). This suggests that 

middle-aged respondents are somewhat less inclined to believe in autonomous vehicles being 

used as taxis compared to younger adults. 

Other Findings 

Gender was another significant predictor, with women (b = -0.54, p < 0.001) less likely than men 

to believe that autonomous vehicles can serve as taxis. Respondents identifying as a non-binary 

or other gender were even less likely to support this belief (b = -1.218, p < 0.001). Political 

orientation also demonstrated significant associations. Compared to conservatives, moderate 

liberals (b = 0.235, p = 0.007) and liberals (b = 0.421, p < 0.001) were more likely to endorse 

autonomous vehicles as taxis. Respondents with a college degree (b = 0.35, p < 0.001) and 

postgraduate degree (b = 0.274, p = 0.001) were more likely to believe in autonomous vehicles 

as viable taxi options. Race also played a role, with Black respondents being somewhat less 

likely (b = -0.221, p = 0.061) than white respondents to believe in autonomous vehicles being 

42 

used as taxis. Additionally, individuals of other racial or ethnic backgrounds were less likely (b = 

-0.406, p = 0.02) to endorse this belief. 

The technology used to operate driverless passenger vehicles could be used for a number of 

purposes. Would you favor or oppose the use of this technology in each of the following 

purposes: Buses for public transportation? 

Summarized in Table 8, age was a significant factor in the likelihood of endorsing the belief in 

using autonomous vehicles as buses. Compared to individuals aged 18-29, respondents aged 30-

49 were less likely to support this view (b = -0.352, p = 0.003), with an odds ratio of 0.538 (95% 

CI [0.433, 0.669]), corresponding to a moderate effect size (Cohen’s d ≈ 0.5). Similarly, 

respondents aged 50-64 were also less likely to endorse this belief (b = -0.313, p = 0.011), with 

an odds ratio of 0.577 (95% CI [0.456, 0.731]), corresponding to a medium effect size (Cohen’s 

d ≈ 0.5). These results suggest that older respondents are less likely to support the use of 

autonomous vehicles as buses. 

Other Findings 

Gender was a significant predictor, with women (b = -0.614, p < 0.001) less likely than men to 

believe in the use of autonomous vehicles as buses. Political orientation also showed notable 

associations. Compared to conservatives, liberals (b = 0.377, p < 0.001) were more likely to 

endorse this belief. Income also influenced endorsement, with individuals earning between 

$30,000 and $39,999 being less likely (b = -0.245, p = 0.078) to support autonomous vehicles as 

buses. Education played a significant role, as respondents with a college degree (b = 0.331, p < 

0.001) and those with postgraduate education (b = 0.383, p < 0.001) were more likely to believe 

in the viability of autonomous vehicles as buses. 

43 

Table 8 summarizes the results of the path model predicting the likelihood of endorsement items: 

that autonomous vehicles should be used for taxis, and autonomous vehicles should be used for 

buses. 

44 

Table 8 Direct Effect Regression of Autonomous Vehicle Endorsement on Age, Technological Savviness, and Socio-Demographic Variable 

 Variable 
Age vs 18-29 
           30-49     
           50-64     
           65+     
Internet vs Several times a day, About once a day, Several 
times a week, or Less often 
           Almost constantly 
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative 
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School Graduate or Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    

Should be used for taxis                                              Should be used for buses   
p 

odds ratio  2.5% CI  97.5% CI  p 

odds ratio  2.5% CI  97.5% CI 

** 
* 

*** 

*** 
*** 

** 
*** 

*** 
** 

* 

0.886 
0.587 
0.596 

0.735 
0.476 
0.475 

1.069 
0.724  ** 
0.747  * 

0.854 
0.538 
0.577 

0.706 
0.433 
0.456 

1.033 
0.669 
0.731 

1.491 

1.293 

1.718  * 

1.297 

1.119 

1.503 

0.414 
0.128 

1.41 
1.512 
2.046 

1.072 
1.046 
1.042 
0.829 
1.199 
1.161 
0.856 
1.097 

1.1 
1.742 
1.58 

0.683 
0.981 
0.525 
1.015 

0.362 
0.044 

1.148 
1.264 
1.7 

0.84 
0.805 
0.806 
0.622 
0.902 
0.855 
0.623 
0.883 

0.886 
1.458 
1.288 

0.551 
0.744 
0.365 
0.739 

0.471  *** 
0.314 

1.731 
1.809 
2.464  *** 

1.367 
1.357 
1.346 
1.104 
1.593 
1.575 
1.172 
1.363 

1.363 
2.081  *** 
1.938  *** 

0.844 
1.293 
0.747 
1.392 

0.372 
2.28 

1.128 
1.234 
1.916 

0.717 
0.763 
1.054 
0.807 
1.033 
0.861 
0.95 
1.058 

1.146 
1.76 
1.889 

0.998 
0.943 
0.778 
1.138 

0.324 
1.051 

0.908 
1.023 
1.586 

0.552 
0.576 
0.808 
0.597 
0.77 
0.623 
0.687 
0.847 

0.914 
1.467 
1.535 

0.802 
0.712 
0.537 
0.818 

0.426 
5.298 

1.401 
1.489 
2.318 

0.929 
1.005 
1.372 
1.084 
1.382 
1.183 
1.309 
1.322 

1.432 
2.111 
2.324 

1.239 
1.247 
1.113 
1.574 

Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval

45 

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
If the use of driverless passenger vehicles becomes widespread, do you think that would 

decrease the number of people killed or injured in traffic accidents? 

Summarized in Table 9, age showed some influence on the likelihood of endorsing the belief that 

autonomous vehicles decrease traffic accident injuries and fatalities. Compared to individuals 

aged 18-29, respondents aged 30-49 were more likely to support this view (b = 0.232, p = 0.07), 

with an odds ratio of 1.428 (95% CI [1.133, 1.802]), corresponding to a moderate effect size 

(Cohen’s d ≈ 0.4). While the effect size is moderate, this finding suggests that respondents in this 

age group are somewhat more likely to believe in the injury- and fatality-reduction benefits of 

autonomous vehicles. 

Other Findings 

Gender was a significant predictor, with women more likely than men to believe that 

autonomous vehicles decrease traffic accident injuries and fatalities (b = 0.478, p < 0.001). 

Political orientation was another important factor. Compared to conservatives, moderate liberals 

(b = -0.394, p < 0.001) and liberals (b = -0.578, p < 0.001) were less likely to endorse the belief 

that autonomous vehicles decrease traffic accident injuries and fatalities. Income level also 

impacted beliefs. Individuals with an income between $60,000 and $69,999 were less likely to 

endorse the belief in injury- and fatality-reduction benefits (b = -0.321, p = 0.013). Additionally, 

those earning $100,000 or more were less likely (b = -0.289, p = 0.009) to support this belief. 

Education played a role, as respondents with a postgraduate degree were less likely to endorse 

the belief that autonomous vehicles decrease traffic accident injuries and fatalities (b = -0.212, p 

= 0.01). Race was also a significant predictor. Black respondents (b = 0.277, p = 0.019) were 

more likely to believe in these safety benefits. There was also a trend for Hispanic respondents to 

endorse this benefit (b = 0.206, p = 0.237). Conversely, respondents of other racial/ethnic 

46 

backgrounds (b = -0.31, p = 0.051) and Asian respondents (b = -0.333, p = 0.066) were less 

likely to endorse the injury- and fatality-reduction benefits. 

If the use of driverless passenger vehicles becomes widespread, do you think the computer 

systems in driverless passenger vehicles would be easily hacked in ways that put safety at 

risk? 

Summarized in Table 9, age had a notable influence on the likelihood of endorsing the belief that 

autonomous vehicles will not be easily hacked in ways that put safety at risk. Compared to 

individuals aged 18-29, respondents aged 30-49 were more likely to report belief in this (b = 

0.517, p < 0.001), with an odds ratio of 1.261 (95% CI [1.092, 1.456]), corresponding to a 

moderate effect size (Cohen’s d ≈ 0.45). Similarly, those aged 50-64 also showed an increased 

likelihood of endorsing this belief (b = 0.439, p = 0.001), with an odds ratio of 1.251 (95% CI 

[1.072, 1.460]), corresponding to a moderate effect size (Cohen’s d ≈ 0.4). This suggests that 

middle-aged respondents are more inclined to believe that autonomous vehicles are not easily 

hackable in ways that compromise safety. 

Other Findings 

Gender was another significant predictor, with women (b = 0.344, p < 0.001) more likely than 

men to believe that autonomous vehicles are resistant to hacking threats that could jeopardize 

safety. Political attitudes also showed significant associations. Compared to conservatives, 

moderate conservatives (b = -0.405, p < 0.001), moderate liberals (b = -0.442, p < 0.001), and 

liberals (b = -0.559, p < 0.001) were all less likely to endorse the belief that autonomous vehicles 

are safe from hacking. Income also had some significant effects. Respondents with an income 

between $90,000 and $99,999 were less likely (b = -0.416, p = 0.016) to believe that autonomous 

vehicles are resilient to hacking threats. Those with an income of $100,000 or more also showed 

47 

reduced endorsement of this belief (b = -0.281, p = 0.027). Educational attainment showed 

mixed results: individuals with an associate degree were less likely to believe in the resilience of 

autonomous vehicles against hacking (b = -0.241, p = 0.05). However, no significant associations 

were found for college graduates or postgraduates. Race was a significant predictor, with Black 

respondents more likely (b = 0.043, p = 0.74) than their white counterparts to believe that 

autonomous vehicles are not easily hackable in ways that could compromise safety. 

Table 9 summarizes the results of the path model predicting the likelihood of endorsement items: 

that autonomous vehicles are more likely to be hacked, and autonomous vehicles decrease traffic 

accident injuries and fatalities. 

48 

Table 9 Direct Effect Regression of Autonomous Vehicle Endorsement on Age, Technological Savviness, and Socio-Demographic Variable  

 Variable 
Age vs 18-29 
           30-49     
           50-64     
           65+     
Internet vs Several times a day, About once a day, 
Several times a week, or Less often 
           Almost constantly 
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative 
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School Graduate or 
Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    

More likely to be hacked                                                             
p 

Decrease traffic accident injuries and fatalities                                             
p 

odds ratio 

odds ratio 

97.5% CI 

97.5% CI 

2.5% CI 

2.5% CI 

*** 
** 

*** 

*** 
*** 
*** 

* 
* 

1.078 
1.261 
1.251 

0.945 
1.092 
1.072 

1.23 
1.456 
1.46 

1.03 
1.428 
1.343 

0.83 
1.133 
1.048 

0.996 

0.903 

1.098 

1.084 

0.924 

1.262 
0.623 

0.853 
0.813 
0.683 

1.105 
1.196 
1.137 
0.922 
1.016 
1.401 
0.896 
0.9 

0.832 
0.981 
0.895 

1.174 
0.865 
1.018 
1.039 

1.154 
0.344 

0.742 
0.72 
0.602 

0.938 
1.005 
0.955 
0.76 
0.835 
1.127 
0.722 
0.775 

0.716 
0.866 
0.776 

1.014 
0.714 
0.804 
0.828 

1.381  *** 
1.068 

0.98 

0.917  *** 
0.776  *** 

1.301 
1.422 
1.353 
1.117  * 
1.234 
1.741 
1.11 

1.044  ** 

0.965 
1.11 

1.032  * 

1.359  * 
1.045 
1.284 
1.3 

2.231 
0.573 

0.819 
0.517 
0.377 

0.896 
1.061 
0.84 
0.583 
0.739 
0.893 
0.718 
0.621 

0.762 
0.819 
0.684 

1.571 
1.409 
0.569 
0.551 

1.926 
0.136 

0.66 
0.425 
0.303 

0.689 
0.809 
0.638 
0.421 
0.536 
0.637 
0.509 
0.487 

0.597 
0.669 
0.536 

1.252 
1.029 
0.361 
0.361 

1.28 
1.802 
1.723 

1.271 

2.588 
1.675 

1.013 
0.629 
0.467 

1.162 
1.388 
1.102 
0.8 
1.011 
1.243 
1.006 
0.791 

0.966 
0.999 
0.868 

1.967 
1.912 
0.869 
0.817 

Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

49 

 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Do you think widespread use of driverless passenger vehicles would be a good idea for 

society?  

Summarized in Table 10, age was found to influence the likelihood of endorsing the belief that 

autonomous vehicles are beneficial for society. Compared to individuals aged 18-29, respondents 

aged 50-64 were less likely to believe in the societal benefits of autonomous vehicles (b = -

0.185, p = 0.16), with an odds ratio of 0.695 (95% CI [0.534, 0.904]). The odds ratio of 0.695 

suggests a moderate effect size, indicating that individuals in this age group are somewhat less 

likely to endorse the belief that autonomous vehicles benefit society, though the difference, while 

statistically significant, is modest in practical terms (Cohen’s d ≈ 0.33). Similarly, those aged 65 

and older also showed a decline in this belief (b = -0.336, p = 0.02), with an odds ratio of 0.581 

(95% CI [0.406, 0.823]). The odds ratio of 0.581 indicates a moderate to large effect size 

(Cohen’s d ≈ 0.5), meaning that older adults in the 65+ group are significantly less likely to 

endorse the view that autonomous vehicles are beneficial for society compared to individuals 

aged 18-29. 

In addition, internet usage was a strong positive predictor of believing in the societal benefits of 

autonomous vehicles. Those who reported higher levels of internet usage were more likely to 

endorse the belief that autonomous vehicles are beneficial (b = 0.234, p = 0.001), with an odds 

ratio of 1.503 (95% CI [1.281, 1.763]). This suggests a small effect size (Cohen’s d ≈ 0.2), 

indicating that individuals with greater internet engagement are more likely to believe that 

autonomous vehicles will be beneficial for society.\ 

Other Findings 

Gender had a significant influence, with women being less likely than men to endorse the belief 

that autonomous vehicles are beneficial for society (b = -0.720, p < 0.001). Political attitudes 

50 

were another important factor, with moderate liberals (b = 0.442, p < 0.001) and liberals (b = 

0.570, p < 0.001) exhibiting significantly higher levels of endorsement for the societal benefits of 

autonomous vehicles. Respondents with an income of $100,000 or more were significantly more 

likely to endorse the belief that autonomous vehicles are beneficial for society (b = 0.118, p = 

0.032). Education also played a significant role in the belief that autonomous vehicles are 

beneficial. Respondents with a college degree exhibited significantly higher levels of 

endorsement for the societal benefits of autonomous vehicles (b = 0.327, p < 0.001). 

Postgraduates showed an even stronger association (b = 0.394, p < 0.001) suggesting that higher 

education levels are associated with greater belief in the societal benefits of autonomous 

vehicles. Race and ethnicity also had significant associations. Black respondents were 

significantly less likely to believe in the societal benefits of autonomous vehicles (b = -0.395, p = 

0.003). Individuals from other racial/ethnic backgrounds were also less likely to endorse the 

belief that autonomous vehicles benefit society (b = -0.336, p = 0.046). 

If the use of driverless passenger vehicles becomes widespread, do you think older adults 

and people with disabilities will be able to live more independently? 

Summarized in Table 10, internet usage was found to be a strong positive predictor of believing 

that autonomous vehicles can enhance the independence of older adults and people with 

disabilities. Individuals with higher levels of internet use were significantly more likely to 

endorse this belief (b = 0.292, p < 0.001), with an odds ratio of 1.211 (95% CI [1.099, 1.335]). 

This odds ratio suggests a moderate effect size (Cohen’s d ≈ 0.4), indicating that individuals 

more engaged online are more inclined to see autonomous vehicles as beneficial for promoting 

independence among these groups. 

51 

 
Other Findings 

Gender also showed a significant influence, with women being less likely than men to endorse 

the idea that autonomous vehicles can support independence for older adults and people with 

disabilities (b = -0.334, p < 0.001). Political attitudes were another important factor, with 

moderate liberals (b = 0.189, p = 0.034) and liberals (b = 0.346, p < 0.001) both significantly 

more likely than conservatives to believe that autonomous vehicles can enhance independence 

for older adults and individuals with disabilities. Higher levels of education were also associated 

with greater endorsement of autonomous vehicles as a tool for enhancing independence. 

Respondents with a college degree expressed significantly higher belief (b = 0.196, p = 0.013) in 

this potential. Postgraduates showed a similar significant positive association (b = 0.230, p = 

0.005), suggesting that educational attainment is linked to greater belief in the potential of 

autonomous vehicles to support independent living. Black respondents were significantly less 

likely to believe that autonomous vehicles can promote independence for older adults and people 

with disabilities (b = -0.424, p < 0.001). 

52 

 
Table 10 Direct Effect Regression of Autonomous Vehicle Endorsement on Age, Technological Savviness, and Socio-Demographic Variable  

 Variable 
Age vs 18-29 
           30-49     
           50-64     
           65+     
Internet vs Several times a day, About once a day, Several 
times a week, or Less often 
           Almost constantly 
Gender vs A man 
           A woman     
           In some other way    
Party vs Conservative 
           Moderately Conservative 
           Moderately Liberal 
           Liberal 
Family income vs Less than $30,000  
           $30,000 to less than $40,000    
           $40,000 to less than $50,000    
           $50,000 to less than $60,000    
           $60,000 to less than $70,000    
           $70,000 to less than $80,000    
           $80,000 to less than $90,000    
           $90,000 to less than $100,000    
           $100,000 or more    
Education level category vs High School Graduate or Less 
           Associate   
           College Grad   
           Postgrad     
Race-Ethnicity vs White non-Hispanic 
           Black non-Hispanic    
           Hispanic    
           Other     
           Asian non-Hispanic    

Good for society                                                       Increase independence for older adults 
p 

odds ratio 

odds ratio 

97.5% CI 

2.5% CI 

2.5% CI 

p 

97.5% CI 

1.205 
0.729 
0.695 

0.983 
0.575 
0.534 

1.48 
0.924 
0.904 

1.01 
0.882 
0.877 

0.888 
0.764 
0.75 

1.149 
1.019 
1.025 

1.503 

1.281 

1.763  *** 

1.211 

1.099 

1.335 

0.294 
0.853 

1.414 
2.154 
2.686 

0.957 
1.1 
1.087 
0.581 
0.94 
1.268 
0.873 
1.296 

1.323 
1.734 
1.95 

0.509 
1.055 
0.577 
1.335 

0.253 
0.393 

1.109 
1.752 
2.18 

0.721 
0.815 
0.808 
0.406 
0.678 
0.901 
0.606 
1.019 

1.034 
1.424 
1.563 

0.393 
0.788 
0.377 
0.952 

0.342  *** 
1.816 

1.8 
2.653  * 
3.316  *** 

1.266 
1.479 
1.457 
0.823 
1.296 
1.775 
1.248 
1.651 

1.687 

2.11  * 
2.432  ** 

0.655  *** 
1.407 
0.864 
1.862 

0.811 
1.028 

1.165 
1.215 
1.191 

1.191 
1.027 
0.973 
1.103 
1.057 
1.405 
1.2 
1.162 

0.987 
1.224 
1.103 

0.829 
1.031 
1.034 
0.974 

0.741 
0.628 

1.011 
1.075 
1.05 

1.008 
0.858 
0.812 
0.909 
0.866 
1.128 
0.969 
1.002 

0.85 
1.082 
0.959 

0.713 
0.86 
0.819 
0.775 

0.888 
1.662 

1.342 
1.375 
1.352 

1.406 
1.227 
1.164 
1.338 
1.287 
1.748 
1.485 
1.348 

1.146 
1.384 
1.269 

0.963 
1.235 
1.301 
1.22 

** 

*** 

*** 
*** 

* 

*** 
*** 

** 

* 

Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001, CI = Confidence Interval 

53 

 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
H8: Mediation Effect of Technological savviness Association between Age and Endorsement of 

Autonomous Vehicles 

H8, summarized in Table 11, proposed that technological savviness mitigates the negative 

perceptions older adults have toward the endorsement of autonomous vehicle-related beliefs. The 

mediation analysis supported H8 for participants aged 30-49, 50-64, and 65 and older regarding 

the belief that autonomous vehicles benefit society. Specifically, the results showed significant 

negative indirect effects for ages 30-49 (b=−0.031, p=0.022), ages 50-64 (b=−0.068, p=0.005), 

and ages 65 and older (b=−0.107, p=0.004). These findings suggest that lower technological 

savviness exacerbates negative perceptions of the societal benefits of autonomous vehicles 

among older adults. Furthermore, the total and direct effects for age 65 and older (b=−0.292, 

p=0.019) indicated full mediation, meaning that the impact of age on societal perceptions of 

autonomous vehicles is fully explained by differences in technological savviness. For 

participants aged 30-49 and 50-64, partial mediation was observed, as both the indirect and total 

effects were significant. 

A similar pattern was observed for the endorsement of autonomous vehicles as taxis, where 

technological savviness mitigated negative perceptions for ages 30-49 (b=−0.032, p=0.018), 50-

64 (b=−0.071, p=0.003), and 65 and older (b=−0.112, p=0.002). In this case, the results for all 

age groups demonstrated partial mediation, as both direct and indirect effects were significant. 

Technological savviness also mitigated negative beliefs regarding the independence-promoting 

benefits of autonomous vehicles for older adults, with significant indirect effects observed for 

ages 30-49 (b=−0.039, p=0.013), 50-64 (b=−0.085, p=0.001), and 65 and older (b=−0.134, 

p<0.001). Similar to beliefs about societal benefits, full mediation was observed for the 65 and 

older age group, while partial mediation occurred for younger age groups. The mediation 

54 

analysis for using autonomous vehicles as buses also revealed that technological savviness 

mitigates older adults’ negative perceptions, particularly for those aged 65 and older (b=−0.068, 

p=0.047) suggesting partial mediation in this age group. Regarding beliefs about the stress-

reducing benefits of autonomous vehicles, technological savviness helped mitigate skepticism 

for ages 50-64 (b=−0.051, p=0.021) and 65 and older (b=−0.08, p=0.017) with evidence of 

partial mediation for these age groups. 

Further, significant total effects demonstrated that both technological savviness and age 

contribute to the negative perceptions of societal benefits among older adults. For individuals 

aged 65 and older (b=−0.292, p=0.019), technological savviness was associated with a greater 

belief that autonomous vehicles benefit society. Additionally, for older adults, technological 

savviness mitigated negative perceptions of sharing the road with autonomous vehicles, with 

significant effects observed for ages 50-64 (b=−0.52, p<0.001) and 65 and older (b=−0.491, 

p<0.001). Support for autonomous taxis showed significant negative effects for ages 50-64 

(b=−0.395, p=0.001) and for ages 65 and older (b=−0.425, p<0.001). For using autonomous 

vehicles as buses, ages 50-64 (b=−0.396, p=0.001) and 65 and older (b=−0.382, p=0.001) also 

showed significant negative effects. 

Table 11 summarizes the indirect and total effects of the path model relationships between age, 

technological savviness, and the likelihood of endorsement items for autonomous vehicles.  

55 

 
 Table 11 Indirect and Total Effects of the Endorsement Path Models 

Good for 
society                                                       

Decrease 
More likely 
Make driving 
traffic 
to be 
less stressful                                                   
hacked                                                             
accident 
injuries and 
fatalities                                             
b 

Increase 
Good to share 
the road with                                              
independence 
for older 
adults 

p  b 

p 

p 

p 

b 

b 

b 

Path 

b 

p 

Should be used 
for taxis                                            

Should be used 
for buses   

p 

b 

p 

b 

p 

Indirect 
30-49 
Indirect 
50-64 
Indirect 
65+ 
Total 30-
49 
Total 50-
64 
Total 65+ 

-0.031  * 

-0.023 

0.001 

-0.007 

-0.032  * 

-0.039  * 

-0.032  * 

-0.02 

-0.068  ** 

-0.051  * 

0.001 

-0.016 

-0.071  ** 

-0.085  ** 

-0.071  ** 

-0.043 

-0.107  ** 

-0.08  * 

0.002 

-0.025 

-0.112  ** 

-0.134  *** 

-0.112  ** 

-0.068  * 

0.087 

-0.221 

0.194 

0.018 

-0.131 

0.042 

-0.112 

-0.113 

-0.224 

-0.441  *** 

0.519  *** 

0.216 

-0.52  *** 

-0.161 

-0.395  ** 

-0.396  ** 

-0.292  * 

-0.398  ** 

0.441  ** 

0.162 

-0.491  *** 

-0.263  * 

-0.425  *** 

-0.382  ** 

Note: b = coefficient, * p < 0.05, ** p < 0.01, *** p < 0.001 

56 

 
 
 
 
 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
DISCUSSION 

Although research on older adults’ engagement with technology and their attitudes toward 

autonomous vehicles is growing, few studies comprehensively examine how demographic and 

socioeconomic factors influence these patterns. Furthermore, limited attention has been given to 

the mediating role of technological savviness in shaping older adults’ adoption of autonomous 

vehicles. To address these gaps, this study analyzed survey data to investigate age-related 

disparities in technological savviness, their impact on autonomous vehicle adoption, and the 

broader implications for promoting inclusive technological adoption. 

Older adults were significantly less likely to exhibit technological savviness than their younger 

counterparts, consistent with prior research documenting the digital divide (Heponiemi et al., 

2023). In this context, the digital divide refers to the gap between individuals or groups with 

different levels of access to, knowledge of, and skills in using digital technologies, particularly 

information and communication technologies (Rogers et al., 2017). For older adults, the digital 

divide typically manifests as a lower level of technological savviness compared to younger 

populations, which can lead to challenges in engaging with emerging technologies like 

autonomous vehicles (Heponiemi et al., 2023). Respondents aged 65 and older reported the 

lowest levels of technological savviness. The findings suggest that age-related barriers, including 

lower exposure to digital environments and reduced confidence in using technology, persist as 

significant challenges for older adults. 

Age was a significant predictor of autonomous vehicle adoption, with older adults exhibiting a 

lower likelihood of adoption and of endorsing all items. These results suggest that older adults 

face unique challenges in engaging with autonomous vehicle technology, likely driven by 

reduced technological savviness and trust in automation. Interestingly, no significant differences 

57 

were observed for the 30-49 age group, indicating that middle-aged adults may be more 

adaptable to new technologies or that other factors, such as familiarity with autonomous vehicle 

technology, play a greater role for this demographic. Further research, however, is needed to 

explore these dynamics and identify strategies to enhance autonomous vehicle adoption across 

all age groups. 

The findings from the mediation analysis reveal that technological savviness plays a significant 

role in shaping older adults' perceptions of autonomous vehicles through both full and partial 

mediation effects. These results emphasize that technological savviness can act as a key 

moderating factor in reducing skepticism and fostering more positive attitudes toward 

autonomous vehicles, particularly for older adults. This suggests that, for older adults, a lack of 

familiarity with technology exacerbates negative beliefs, and increasing technological 

competence can substantially improve their views of autonomous vehicles. 

In the case of adults of age 65 and over, the mediation path revealed that technological savviness 

completely explains their negative perceptions of the societal benefits of autonomous vehicles. 

This finding indicates full mediation, meaning that the relationship between age and skepticism 

about societal benefits is fully accounted for by the individual’s level of technological savviness. 

For adults aged 30-49 and 50-64, partial mediation was observed, as technological savviness 

moderated, but did not entirely explain, their attitudes toward autonomous vehicles. This 

suggests that while technological savviness is a crucial factor in shaping attitudes, other 

influences, such as personal experiences, prior exposure to technology, and generational 

differences in technological adoption, may have also played significant roles in how these 

individuals perceive autonomous vehicles. 

58 

For individuals aged 65 and older, technological savviness completely explains their beliefs 

about the societal advantages of autonomous vehicles, including their potential to increase 

mobility, reduce reliance on human-driven transportation, and enhance independence (Graham-

Rowe et al., 2011). This full mediation effect suggests that for older adults, the perception of 

autonomous vehicles as beneficial to society is primarily shaped by their technological 

savviness. Technologically savvy older adults are more likely to see autonomous vehicles as a 

positive force for change, contributing to goals such as reducing traffic accidents, supporting 

people with disabilities, and providing more accessible transportation options. On the other hand, 

older adults with lower technological savviness may struggle to recognize these benefits, 

focusing instead on perceived risks and uncertainties associated with adopting new technologies 

(Graham-Rowe et al., 2011, Lin et al., 2017). 

Additionally, in the case of adults 65 and above, technological savviness fully mediated their 

perceptions of autonomous vehicles as supporting their, and other older adults’, and disabled 

individuals’ independence. This suggests that older adults who are more technologically 

proficient are more likely to see autonomous vehicles as a tool to maintain mobility and 

independence. For participants aged 30-49 and 50-64, partial mediation was observed, indicating 

that while technological savviness was a significant factor in shaping perceptions, other elements 

such as health status, family support, and current independence may have also influenced how 

they viewed autonomous vehicles in this context. The results suggest that older adults who 

embrace technology are more inclined to view autonomous vehicles as a way to enhance their 

autonomy, whereas others may not fully recognize their potential in this area. 

The mediation effect for the perception of autonomy and independence highlights the role of 

technological competence in alleviating fears of loss of control. Autonomous vehicles, which 

59 

offer an alternative to driving, are particularly relevant for older adults who may face challenges 

in maintaining independence due to physical limitations or declining health. For older adults who 

are comfortable with technology, autonomous vehicles may be seen as a means of maintaining 

mobility and self-sufficiency. Conversely, those less familiar with technology may find the idea 

of relinquishing control to a machine daunting, leading to a reluctance to embrace autonomous 

vehicles as a tool for preserving independence (AARP, 2017). This underscores the importance 

of technological comfort in shaping older adults' perceptions of autonomy and control. 

The presence of both full and partial mediation is noteworthy because it underscores the varying 

degrees to which technological competence influences the acceptance of autonomous vehicles 

depending on age and technological familiarity. For older adults, the mediation effect is stronger, 

possibly due to a higher level of resistance to new technologies and the potential emotional and 

cognitive challenges associated with learning new systems. Younger adults, by contrast, may 

have already developed a greater degree of comfort and trust in technology, which makes the 

mediation effect less pronounced. 

Technological savviness was also found to mitigate older adults’ negative perceptions of 

autonomous vehicles as buses and taxis. In this case, partial mediation was observed for the 65 

and older age group, suggesting that while technological savviness played a role, additional 

factors such as accessibility, public transportation needs, and trust in autonomous systems may 

have also influenced their views. For adults aged 30-49 and 50-64, similar patterns were 

observed, where technological competence was a contributing factor, but concerns may have 

remained about the feasibility and safety of using autonomous vehicles for public transportation 

(Venkatesh & Bala, 2008). Technologically savvy older adults may be more willing to embrace 

the idea of autonomous vehicles in public transportation settings, as they may have greater 

60 

confidence in the technology’s reliability and safety. However, individuals with lower 

technological competence may perceive autonomous vehicles on the road or in public transport 

as unpredictable or dangerous, which reinforces their reluctance to share the road with these 

vehicles (Graham-Rowe et al., 2011, Lin et al., 2017). These partial mediation effects underscore 

the importance of technological comfort in shaping perceptions but also highlight that other 

factors, such as concerns about safety, social influences, and prior transportation experiences, 

should be considered in policies aimed at increasing acceptance of autonomous vehicles. 

For adults aged 50-64 and 65 and older, technological savviness partially mediated their 

perceptions of autonomous vehicles as reducing transportation stress. These results suggest that 

while technological savviness helped mitigate skepticism about the stress-reducing potential of 

autonomous vehicles, other factors such as current transportation experiences, lifestyle, and 

health concerns also played a role in shaping their attitudes. The perception of autonomous 

vehicles as reducing transportation stress may be influenced by an individual’s previous 

experiences with technology and their general comfort with modern systems. Older adults with 

higher technological savviness are likely to view autonomous vehicles as a more efficient, safer 

alternative to driving themselves, particularly in managing stress-inducing factors such as traffic 

and navigation (Venkatesh & Bala, 2008, Nielsen et al., 2019). Technologically savvy individuals 

may appreciate the reduction in stress and cognitive load associated with driving, such as having 

to navigate unfamiliar roads or deal with traffic congestion. In contrast, individuals with lower 

technological savviness may remain more skeptical, potentially perceiving autonomous vehicles 

as an unfamiliar and risky technology, regardless of their potential benefits (Nielsen et al., 2019). 

 The full mediation effect for the perception of autonomous vehicles as beneficial for society and 

supporting independence among older adults can be understood in light of the unique challenges 

61 

faced by this demographic. Older adults, particularly those aged 65 and older, often experience 

reduced mobility due to physical health issues, and autonomous vehicles offer a potential 

solution to maintaining independence without relying on others for transportation. For 

technologically savvy individuals, the benefits of autonomous vehicles may be more apparent, as 

they are better equipped to understand and trust the technology. However, those with less 

technological experience may perceive autonomous vehicles as a threat to their autonomy, as 

they may feel uncomfortable with the idea of relying on a machine for transportation (AARP, 

2017, Lin et al., 2017). 

Similarly, the partial mediation observed in other areas, such as transportation stress, road 

sharing, and public transit use, may reflect a more complex interplay of factors. While 

technological savviness is certainly important, perceptions of autonomy, control, and safety also 

play crucial roles in shaping older adults' attitudes toward autonomous vehicles. These factors 

may be influenced by personal experiences, societal attitudes toward new technologies, and 

broader concerns about the risks and benefits of autonomous systems (Graham-Rowe et al., 

2011, Nielsen et al., 2019). 

This aligns with existing research, which suggests that individuals with higher technological 

literacy are generally more receptive to new technologies, as they are better able to understand 

and trust the technological mechanisms behind them (Venkatesh & Bala, 2008). Technologically 

savvy individuals are not only more likely to use autonomous vehicles themselves but also 

recognize their broader societal potential. Moreover, the positive association between 

technological savviness and support for autonomous vehicles as taxis and buses highlights the 

role of convenience and familiarity with technological innovations in fostering acceptance 

(Nielsen et al., 2019). 

62 

The lack of a significant association between technological savviness and safety concerns, 

particularly regarding hacking and the likelihood of crashes, warrants further investigation. It is 

possible that even highly technologically savvy individuals are wary of autonomous vehicles. In 

this context, perceived safety may be influenced by societal narratives, media portrayals of 

accidents involving autonomous vehicles, or personal experiences with traditional transportation 

modes. Furthermore, the complexity of issues like cybersecurity and the vulnerability of 

autonomous vehicles to potential hacking might require more than just technical knowledge to 

overcome. These concerns may stem from a broader societal unease about the risks associated 

with autonomous technology, which is not easily mitigated by individual technological 

savviness. 

To facilitate the acceptance of autonomous vehicles among older adults, efforts should focus on 

improving technological savviness and addressing broader concerns regarding safety, reliability, 

and societal impact. Targeted education, free trial runs, and training programs aimed at 

enhancing technological competence could help mitigate skepticism, particularly among older 

adults who may have had limited exposure to modern technologies. Creating opportunities for 

older adults to interact with autonomous vehicle technology in a low-pressure environment could 

help build confidence and reduce fears of obsolescence (Venkatesh & Bala, 2008, Nielsen et al., 

2019). 

Additionally, understanding the factors that contribute to the partial mediation effects for 

transportation stress, road sharing, and public transit use could help in designing more 

comprehensive approaches to increasing acceptance. These interventions might not only focus on 

technological training but also address broader concerns, such as promoting the safety and 

reliability of autonomous vehicles, improving public perceptions of shared transportation, and 

63 

emphasizing the personal control and convenience that autonomous vehicles can offer, especially 

for older adults (Davis, 1989, Venkatesh et al., 2003). 

These findings can be understood in the context of both psychological and social factors that 

influence technology adoption. Older adults often have a more entrenched relationship with 

traditional modes of transportation, particularly driving, and may view autonomous vehicles as a 

challenge to their sense of autonomy and independence. Technological savviness, in this case, 

serves as a buffer against these anxieties, allowing older adults to view autonomous vehicles 

more positively as tools that can enhance mobility and independence rather than threats to their 

personal control. This suggests that fostering familiarity and comfort with new technologies is 

key to overcoming cognitive and emotional barriers to acceptance. 

In contrast, younger adults, who are generally more accustomed to technological innovation, 

may approach autonomous vehicles with greater trust in the technology, which could explain 

why the mediation effect is less pronounced for them. For these individuals, other factors, such 

as social influence, cost, or perceived utility, may play a more prominent role in shaping their 

views on autonomous vehicles. 

The reluctance to adopt shared autonomous transportation systems, such as autonomous buses, 

may stem from older adults' concerns about safety and the perceived loss of control in public 

spaces. Unlike personal vehicles, which offer a level of privacy and control, autonomous buses 

involve shared spaces that may be seen as less predictable or secure. While technological 

savviness may help alleviate some of these concerns by increasing trust in the technology, it may 

not completely override the intrinsic discomfort associated with sharing public transportation 

with others, particularly in a new and unfamiliar format. 

64 

Women consistently emerged as significant negative predictors across autonomous vehicle items. 

However, since gender was not found to be a significant factor in predicting tech savviness, 

increasing tech savviness would likely have a similar positive effect on both older men and 

women in terms of their attitudes toward autonomous vehicles (Hill & Johnson, 2018). This is 

key as autonomous vehicles hold significant promise for enhancing mobility and independence 

among older adult women, a demographic disproportionately affected by osteoporosis and 

associated mobility challenges. Osteoporosis, characterized by decreased bone density and 

increased fracture risk, affects women at higher rates than men, particularly post-menopause 

(Zhang et al., 2023). This condition can render traditional modes of transportation, such as 

driving or public transit, daunting due to concerns about physical stability and safety. 

Compounding this, older women typically outlive men, often leading to increased social isolation 

and reliance on others for transportation (Shekelle et al., 2024). 

The results also suggest that political ideology may play a significant role in internet usage and 

was a key predictor in autonomous vehicle items, with moderate liberals and liberals 

demonstrating higher levels of internet usage compared to conservatives. This aligns with 

existing research that suggests liberal individuals are often more open to new technologies and 

digital platforms (Dube & Schmitt, 2021). The connection between political ideology and digital 

engagement could reflect broader cultural and ideological differences in media consumption, 

with liberal individuals perhaps more likely to embrace digital technology and new forms of 

media (Foster & Andrews, 2020). Respondents with more liberal ideologies, including moderate-

liberal and liberal individuals, were significantly more likely to endorse the societal benefits of 

autonomous vehicles and express support for their use as taxis and buses. These findings align 

with prior research showing that individuals with liberal political views are generally more open 

65 

to progressive technologies (Smith & Green, 2020). In contrast, conservative respondents 

exhibited more skepticism about the societal benefits and were less willing to share autonomous 

vehicles with others. These ideological differences underscore the need to tailor messaging 

around autonomous vehicles to resonate with individuals’ political leanings and values (Chen & 

Pohl, 2021). For example, messaging that stresses individual choice may be more appropriate for 

conservative audiences while messaging that stresses societal welfare may be more appropriate 

for liberal audiences. 

Given the findings that technological savviness influences attitudes toward autonomous vehicles, 

interventions aimed at improving digital literacy could have a significant impact on adoption. 

Policymakers and community organizations should consider launching digital literacy initiatives 

specifically targeted at older adults, emphasizing the practical benefits of autonomous vehicles 

and their potential to enhance independence (Czaja et al., 2006). These programs should not only 

focus on basic digital skills but also on building confidence in navigating new technologies, 

which may reduce resistance to adopting autonomous vehicles (Charness & Boot, 2009). 

Furthermore, car manufacturers and developers of autonomous vehicle technology could 

collaborate with advocacy groups for older adults to create accessible, user-friendly interfaces 

that reduce perceived complexity (Fisk et al., 2009). Incorporating older adults' feedback into the 

design and testing phases could help ensure that autonomous vehicles meet their specific needs, 

making them feel more comfortable and confident in using the technology. 

Given that frequent internet use was used as a proxy for technological savviness, additional 

policy positions can be introduced to further address the factors that contribute to technological 

adoption, especially among older adults. First, increasing broadband access in underserved 

communities will help older adults stay connected and access vital technology, such as 

66 

autonomous vehicles (Helsper, 2012). Supporting peer mentorship programs can also reduce 

technology-related anxieties and provide older adults with hands-on assistance (Nguyen et al., 

2016). Additionally, subsidizing the cost of internet-enabled devices and internet access will 

lower financial barriers, making these technologies more accessible (Vaportzis et al., 2017). 

Limitations 

This study has several limitations that warrant consideration. First, the path analysis limits the 

ability to establish causal relationships among variables. Although the findings provide insight 

into associations between age, technological savviness, and attitudes toward autonomous 

vehicles, the directionality of these relationships cannot be definitively determined. Furthermore, 

the analysis is subject to potential confounding due to the interplay of age, period, and cohort 

effects. Age effects, which reflect biological and psychological changes over the lifespan, may 

explain older adults’ reduced openness to autonomous vehicle technology. Period effects, such as 

the influence of current sociopolitical and cultural contexts, may also shape attitudes, particularly 

if autonomous vehicles are associated with broader societal shifts like climate change advocacy 

or technological innovation. Similarly, cohort effects, rooted in the historical experiences of 

different generational groups, may account for older adults’ nostalgic attitudes toward traditional 

vehicles or skepticism of emerging technologies. These overlapping effects complicate the 

interpretation of results and underscore the need for longitudinal studies to disentangle these 

influences (Ryder, 1965, Huang & Lee, 2017). Second, the use of self-reported data introduces 

the potential for response biases. Social desirability bias may have influenced participants to 

provide answers they perceived as more acceptable, while recall bias could affect the accuracy of 

responses regarding past technology use. Additionally, self-reported measures rely on subjective 

interpretation, which may vary across individuals and demographic groups, potentially limiting 

67 

the generalizability of findings (Tourangeau, Rips, & Rasinski, 2000, Schwarz, 1999). Lastly, 

while the study draws from a diverse sample, it does not account for all potential contextual 

factors that could influence attitudes toward autonomous vehicles, such as regional differences in 

autonomous vehicle availability, transportation infrastructure, or exposure to autonomous 

vehicle-related messaging (Binns & Carver, 2021). Future research should consider these factors 

to provide a more nuanced understanding of the barriers to autonomous vehicle adoption, 

particularly among older adults. Despite these limitations, this study offers valuable 

contributions to the understanding of how demographic and psychosocial factors influence 

autonomous vehicle adoption and highlights areas for further investigation, including 

interventions to enhance technological savviness and trust in autonomous vehicles among older 

populations. 

68 

 
 
CONCLUSION 

Understanding the interplay between technology usage and acceptance of autonomous vehicles 

among older adults has significant implications for the market. Automakers and technology 

companies may need to adjust their marketing and product development strategies to align with 

the preferences and requirements of this demographic. Addressing specific concerns and interests 

of older adults, such as enhancing mobility, accessibility, and convenience, can render 

autonomous vehicles more attractive to them. Furthermore, older adults' attitudes toward 

autonomous vehicles are strongly shaped by their perceived societal benefits and practical uses, 

such as autonomous vehicles functioning as taxis and buses. These vehicles offer the potential to 

increase mobility and independence for older adults, particularly women, who often face greater 

mobility challenges due to conditions like osteoporosis. As such, emphasizing the broader 

societal benefits of autonomous vehicles, such as enhanced mobility for vulnerable populations, 

will likely resonate with older adults, particularly those with higher technological savviness. 

This study demonstrates that age is a significant predictor of technological savviness, with older 

adults generally showing lower levels of technological engagement, which in turn impacts their 

attitudes toward autonomous vehicles. Younger generations, with higher technological savviness, 

were more likely to view autonomous vehicles positively. This finding suggests that automakers 

should focus on enhancing older adults’ comfort with technology, providing user-friendly 

interfaces and tutorials to improve technological literacy and confidence. 

Political ideology also emerged as a factor in the adoption of autonomous vehicles, with liberals 

and moderate liberals exhibiting more favorable views toward autonomous vehicles compared to 

conservatives. This suggests that marketing efforts may need to be tailored to different political 

orientations to appeal to a broader range of potential consumers. Significant racial and ethnic 

69 

differences were observed in internet usage and technological savviness, with Black and 

Hispanic individuals demonstrating higher levels of internet engagement compared to White 

individuals. This suggests the importance of culturally relevant marketing strategies and outreach 

efforts to ensure that autonomous vehicle adoption is inclusive and accessible across different 

demographic groups. 

Finally, while technological savviness positively influenced comfort with autonomous driving 

and the belief in the societal benefits of autonomous vehicles, concerns about safety, remained 

significant barriers unable to be accessed by tech savviness. Therefore, automakers should also 

focus on addressing these specific concerns through targeted messaging that emphasizes the 

robust safety features and security measures of autonomous vehicles. Collaborating with 

organizations focused on aging, advocacy groups, and gerontology experts can provide essential 

insights into the needs and concerns of older adults, thereby informing more effective marketing 

and product development strategies. 

70 

 
BIBLIOGRAPHY 

AARP. (2017). Autonomous vehicles: Implications for older adults and people with disabilities. 
AARP Public Policy Institute. 

Abdelrahman, N., Haque, R., Polverento, M. E., Wendling, A., Goetz, C., & Arnetz, B. B. 
(2020). Brain Health: Attitudes towards Technology Adoption in Older Adults. Healthcare, 9(1), 
23. https://doi.org/10.3390/healthcare9010023 

Anderson, L. A., Goodman, R. A., Holtzman, D., Posner, S. F., & Northridge, M. E. (2012). 
Aging in the United States: Opportunities and challenges for public health. American Journal of 
Public Health, 102(3), 393–395. https://doi.org/10.2105/ajph.2011.300617 

Andong, R. F., & Sajor, E. (2017). Urban sprawl, public transport, and increasing CO2 
emissions: The case of Metro Manila, Philippines. Environment, Development and 
Sustainability, 19, 99-123. Springer. 

Anstey, K. J., Wood, J., & Lord, S. (2005). The Role of Cognition in the Prediction of Older 
Drivers’ Motor Vehicle Crash Involvement. Age and Ageing, 34(5), 529-534. 
https://doi.org/10.1093/ageing/afi158 

Arias, E. (2018). How does media influence social norms? Experimental evidence on the role of 
common knowledge. Political Science Research and Methods, 7(3), 561-578. 

Azmi, M., Misdi, N. S., Johar, M., & Octora, Y. (2024). Systematic review of road traffic 
regulation compliance. In Innovative Technologies for Enhancing Experiences and Engagement 
(pp. 1-12). Springer. 

Baker, S., Warburton, J., Waycott, J., Batchelor, F., Hoang, T., Dow, B., Ozanne, E., & Vetere, F. 
(2018). Combatting social isolation and increasing social participation of older adults through the 
use of technology: A systematic review of existing evidence. Australasian Journal on Ageing, 
37(3), 184–193. https://doi.org/10.1111/ajag.12572 

Binns, A., Lin, D., & Xie, J. (2020). Perceptions of autonomous vehicle technology: A study of 
older adults. Technology in Society, 62, 101290. https://doi.org/10.1016/j.techsoc.2020.101290 

Binns, C., & Carver, K. (2021). Contextual factors influencing autonomous vehicle acceptance: 
Regional variation and infrastructure challenges. Transportation Research Part F: Traffic 
Psychology and Behaviour, 82, 132-145. https://doi.org/10.1016/j.trf.2021.06.004 

Blažič, B. J., & Blažič, A. J. (2019). Overcoming the digital divide with a modern approach to 
learning digital skills for the elderly adults. Education and Information Technologies, 25(1), 259–
279. https://doi.org/10.1007/s10639-019-09961-9 

71 

 
 
 
 
 
 
 
 
 
 
 
Broady, T. R., Chan, A., & Caputi, P. (2010). Comparison of older and younger adults’ attitudes 
towards and abilities with computers: Implications for training and learning. British Journal of 
Educational Technology, 41(3), 473–485. https://doi.org/10.1111/j.1467-8535.2008.00914.x 

Cantilina, K., Daly, S. R., Reed, M. P., & Hampshire, R. C. (2021). Approaches and barriers to 
addressing equity in transportation: Experiences of transportation practitioners. Transportation 
Research Record, 2675(10), 972-985. SAGE. 

Carr, D., Devos, H., Akinwuntan, A., & Babulal, G. (2023). Addressing the complex driving 
needs of an aging population: The rise of autonomous vehicles. Innovation in Aging, 
7(Supplement_1), 1165-1166. 

Charness, N., & Boot, W. R. (2009). Technology, gaming, and aging. The Gerontologist, 49(3), 
318-328. https://doi.org/10.1093/geront/gnp046 

Chen, H., Cohen, P., & Chen, S. (2010). How Big is a Big Odds Ratio? Interpreting the 
Magnitudes of Odds Ratios in Epidemiological Studies. Communications in Statistics - 
Simulation and Computation, 39(4), 860–864. https://doi.org/10.1080/03610911003650383 

Chen, L., & Pohl, M. (2021). The role of political orientation in shaping attitudes towards 
innovation: A study of autonomous vehicle perceptions. Journal of Technology and Politics, 8(3), 
67-83. 

Criollo-C, S., González-Rodríguez, M., Guerrero-Arias, A., Urquiza-Aguiar, L. F., & Luján-
Mora, S. (2024). A review of emerging technologies and their acceptance in higher education. 
Education Sciences, 14(1), 10. 

Czaja, S. J., Charness, N., Fisk, A. D., Hertzog, C., Nair, S. N., & Rogers, W. A. (2006). Factors 
predicting the use of technology: Findings from the Center for Research and Education on Aging 
and Technology Enhancement (CREATE). The Journals of Gerontology: Series B, Psychological 
Sciences and Social Sciences, 61(6), 401-407. https://doi.org/10.1093/geronb/61.6.P401 

Dahlquist, E. M., Reese, S. E., & Powers, M. L. (2019). Older adults’ acceptance of autonomous 
vehicle technology: The role of trust and safety concerns. Transportation Research Part F: Traffic 
Psychology and Behaviour, 65, 84-97. https://doi.org/10.1016/j.trf.2019.06.004 

Davies, R. (2011). Understanding Technology Literacy: A framework for evaluating educational 
technology integration. TechTrends, 55(5), 45–52. https://doi.org/10.1007/s11528-011-0527-3 

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of 
information technology. Management Information Systems Quarterly, 13(3), 319. 
https://doi.org/10.2307/249008 

Dodel, M., & Mesch, G. S. (2020). Perceptions about the impact of automation in the workplace. 
Information, Communication & Society, 23(5), 665–680. 
https://doi.org/10.1080/1369118x.2020.1716043 

72 

 
 
 
 
 
 
 
 
 
 
 
Dube, L., & Schmitt, B. (2021). Political ideology and the acceptance of technology: A review of 
liberal vs conservative tech adoption. Journal of Political Technology, 16(1), 23-37. 

Eby, D. W., Molnar, L. J., & St. Louis, R. M. (2017). Transportation and aging: An updated 
research agenda to advance safe mobility. The Gerontologist, 59(2), 215-222. 
https://doi.org/10.1093/geront/gnx120 

Enders, C. K. (2010). Applied Missing Data Analysis. Guilford Press. 

Erstad, O. (2022). Key factors in digital literacy in learning and education: A systematic review. 
Educational Technology & Society. 

Erwin, K., & Mohammed, S. (2022). Digital literacy skills instruction and increased skills 
proficiency. International Journal of Technology in Education and Science, 6(2), 323-332. 

Fagan, C., & Jenkins, T. (2021). Closing the digital divide: Affordable technology access for 
low-income individuals. Journal of Public Policy and Technology, 25(3), 200-213. 

Falloon, G. (2020). From digital literacy to digital competence: The teacher digital competency 
(TDC) framework. Educational Technology Research and Development, 68, 2449-2472. 
https://doi.org/10.1007/s11423-020-09798-6 

Filippi, F. (2024). Visions, paradigms, and anomalies of urban transport. Future Transportation, 
4(3), 938-967. MDPI. 

Fisk, A. D., Rogers, W. A., Charness, N., Czaja, S. J., & Sharit, J. (2009). Designing for older 
adults: Principles and creative human factors approaches (2nd ed.). CRC Press. 

Fleming, P., Qu, L., & Walmsley, J. (2018). Policy and technology adoption: Subsidies for 
autonomous vehicle services for low-income older adults. Transportation Research Part A: Policy 
and Practice, 117, 12-20. https://doi.org/10.1016/j.tra.2018.07.017 

Fliss, M. D., Baumgartner, F., Delamater, P., Marshall, S., Poole, C., & Robinson, W. (2020). Re-
prioritizing traffic stops to reduce motor vehicle crash outcomes and racial disparities. Injury 
Epidemiology, 7, 3. 

Foster, K. J., & Andrews, M. A. (2020). Political media consumption and technology adoption: 
Understanding the liberal-conservative divide. Media Studies Journal, 14(2), 45-58. 

Frik, A. (2019). Privacy and security threat models and mitigation strategies of older adults. 
USENIX. https://www.usenix.org/conference/soups2019/presentation/frik 

Garikapati, D., & Shetiya, S. S. (2024). Autonomous vehicles: Evolution of artificial intelligence 
and the current industry landscape. Big Data and Cognitive Computing, 8(4), 42. 

73 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Gössling, S. (2020). Why cities need to take road space from cars - and how this could be done. 
Journal of Urban Design, 25(4), 443-448. Taylor & Francis. 

Graham-Rowe, E., Mullan, B., & Parker, D. (2011). Public attitudes to driverless cars: A study of 
the UK public's views. Transportation Research Part F: Traffic Psychology and Behaviour, 14(5), 
248-261. 

Guzman, A., & Nussbaum, M. (2022). Teachers’ digital competencies in higher education: A 
systematic review. Educational Technology Journal. 

Hargittai, E., Piper, A. M., & Morris, M. R. (2018). From internet access to internet skills: digital 
inequality among older adults. Universal Access in the Information Society, 18(4), 881–890. 
https://doi.org/10.1007/s10209-018-0617-5 

Harrison, G. (2023). Tools for assessing teacher digital literacy: A review. Journal of Educational 
Technology. 

Heath, M. C., Krahmer, B. E., & Kearsley, J. L. (2017). Promoting access to digital technology 
for older adults: The role of government policy. Policy & Internet, 9(3), 321-339. 
https://doi.org/10.1002/poi3.159 

Heponiemi, T., Kainiemi, E., Virtanen, L., Saukkonen, P., Sainio, P., Koponen, P., & Koskinen, S. 
(2023). Predicting internet use and digital competence among older adults using performance 
tests of visual, physical, and cognitive functioning: Longitudinal population-based study. Journal 
of Medical Internet Research, 25, e42287. https://doi.org/10.2196/42287 

Hill, C. E., & Johnson, S. T. (2018). Gender and technology adoption: A study of older women 
and their digital engagement. International Journal of Gender and Technology, 11(3), 76-89. 

Horrigan, J. (2020, May 30). Technology and the internet: Impact on American lifelong learners. 
Pew Research Center: Internet, Science & Tech. 
https://www.pewresearch.org/internet/2016/03/22/adults-with-tech-access-tools-are-more-likely-
to-be-lifelong-learners-and-rely-on-the-internet-to-pursue-knowledge/ 

Horrigan, J. B. (2016). Digital readiness gaps. Pew Research Center. 

Huang, J., & Lee, J. (2017). Understanding the generational divide in attitudes toward 
technology adoption: Age effects, cohort effects, and the role of perceived innovation 
characteristics. Journal of Technology in Human Services, 35(4), 243-264. 
https://doi.org/10.1080/15228835.2017.1391609 

Hwang, W., Woon, C. M., & Siau, K. (2018). A framework for the acceptance of autonomous 
vehicles among older adults. International Journal of Human-Computer Interaction, 34(2), 114-
128. https://doi.org/10.1080/10447318.2017.1371952 

74 

 
 
 
 
 
 
 
 
 
 
 
 
Ingram, M., Leih, R., Adkins, A., Sonmez, E., & Yetman, E. (2020). Health disparities, 
transportation equity and complete streets: A case study of a policy development process through 
the lens of critical race theory. Journal of Urban Health, 97, 876-886. Springer. 

Jokisch, M. R., Scheling, L., Doh, M., & Wahl, H. W. (2022). Contrasting internet adoption in 
early and advanced old age: Does internet self-efficacy matter? The Journals of Gerontology: 
Series B, 77(2), 312-320. https://doi.org/10.1093/geronb/gbab122 

Jokisch, M., Schmidt, L., Doh, M., Marquard, M., & Wahl, H. (2020). The role of internet self-
efficacy, innovativeness and technology Autonomous Vehicleoidance in breadth of internet use: 
Comparing older technology experts and non-experts. Computers in Human Behavior, 111, 
106408. https://doi.org/10.1016/j.chb.2020.106408 

Kadylak, T., & Cotten, S. R. (2020). United States older adults’ willingness to use emerging 
technologies. Information, Communication & Society, 23(5), 736–750. 
https://doi.org/10.1080/1369118x.2020.1713848 

Kavandi, H., & Jaana, M. (2020). Factors that affect health information technology adoption by 
seniors: A systematic review. Health & Social Care in the Community, 28(6), 1827–1842. 
https://doi.org/10.1111/hsc.13011 

Kline, R. B. (2015). Principles and practice of structural equation modeling (4th ed.). The 
Guilford Press. 

Kumi-Yeboah, A., Sallar, A. W., Kiramba, L. K., & Kim, Y. (2020). Exploring the use of digital 
technologies from the perspective of diverse learners in online learning environments. Online 
Learning, 24(4), 42-63. 

Kurichi, J. E., Pezzin, L. E., Streim, J. E., Kwong, P. L., Ling, N., Bogner, H. R., Xie, D., & 
Hennessy, S. (2017). Perceived barriers to healthcare and receipt of recommended medical care 
among elderly Medicare beneficiaries. Archives of Gerontology and Geriatrics, 72, 45–51. 
https://doi.org/10.1016/j.archger.2017.05.007 

Lee, K. H., Zink, J. D., & Lee, J. (2016). Intergenerational technology transfer: Older adults 
learning from younger adults. Educational Gerontology, 42(4), 255-264. 
https://doi.org/10.1080/03601277.2016.1155519 

Lin, D., & Cui, J. (2021). Transport and mobility needs for an ageing society from a policy 
perspective. International Journal of Environmental Research and Public Health, 18(22), 11802. 
https://doi.org/10.3390/ijerph182211802 

Lin, P., Abney, K., & Bekey, G. A. (2017). Autonomes driving and public acceptance: The role of 
attitudes toward technology and government regulation. Journal of Transportation Research Part 
C: Emerging Technologies, 45, 295-312. 

75 

 
 
 
 
 
 
 
 
 
 
 
Lin, Y., & Yu, Z. (2023). Extending technology acceptance model to higher-education students’ 
use of digital academic reading tools on computers. International Journal of Educational 
Technology in Higher Education, 20, 34. https://doi.org/10.1186/s41239-023-00349-1 

Liu, H., Yang, M., Guan, C., Chen, Y. S., Keith, M., You, M., & Menendez, M. (2023). Urban 
infrastructure design principles for connected and autonomous vehicles: A case study of Oxford, 
UK. Computational Urban Science, 3, 34. 

Liu, S., & Harrison, A. (2020). The role of culturally tailored digital literacy programs in 
reducing technology gaps. International Journal of Social Inclusion, 12(2), 35-48. 

Lleras C. (2005). Encyclopedia of social measurement. Elsevier, 42(10), 42–5629. 
https://doi.org/10.5860/choice.42-5629 

Marottoli, R. A., Williams, C. S., & Cooney, L. M. (1994). Driving Cessation and Changes in 
Travel Patterns in Older Adults. The Journals of Gerontology: Series A, 49(5), M283-M290. 
https://doi.org/10.1093/gerona/49.5.M283 

Mattioli, G., & Colleoni, M. (2015). Transport disadvantage, car dependence and urban form. In 
Understanding Mobilities for Designing Contemporary Cities (pp. 171-190). Springer. 

Mattioli, G., Roberts, C., Steinberger, J., & Brown, A. (2020). The political economy of car 
dependence: A systems of provision approach. Energy Research & Social Science, 66, 101486. 
https://doi.org/10.1016/j.erss.2020.101486 

Mentzer, K., Frydenberg, M., & Patterson, A. (2024). Are technological savviness students tech 
literate? Digital and data literacy skills of first-year college students. Information Systems 
Education Journal, 22(3), 4-24. 

Mesch, G. S., & Dodel, M. (2022). The acceptance of driverless cars: the roles of perceived 
outcomes and technology usefulness. American Behavioral Scientist, 67(14), 1736–1754. 
https://doi.org/10.1177/00027642221127250 

Mitzner, T. L., Chen, T. L., & Rogers, W. A. (2010). The effects of an intergenerational 
computing program on older adults' technology attitudes and skills. Gerontology, 56(4), 446-453. 
https://doi.org/10.1159/000288635 

Mjelde, J. W. (2012, July 1). The value of public transportation for improving the quality of life 
for the rural elderly. https://rosap.ntl.bts.gov/view/dot/24602 

Moussaïd, M., Kämmer, J. E., Analytis, P. P., & Neth, H. (2013). Social influence and the 
collective dynamics of opinion formation. PLoS ONE, 8(11), e78433. 

Moxley, J., Sharit, J., & Czaja, S. J. (2022). The factors Influencing older adults’ decisions 
surrounding adoption of Technology: Quantitative Experimental study. JMIR Aging, 5(4), 
e39890. https://doi.org/10.2196/39890 

76 

 
 
 
 
 
 
 
 
 
 
 
 
Myrick, J. G., Ahern, L., Shao, R., & Conlin, J. (2019). Technology name and celebrity 
endorsement effects of autonomous vehicle promotional messages: Mechanisms and moderators. 
Science Communication, 41(1), 38-65. 

Nair, G. S., & Bhat, C. R. (2021). Sharing the road with autonomous vehicles: Perceived safety 
and regulatory preferences. Transportation Research Part C: Emerging Technologies, 122, 
102885. https://doi.org/10.1016/j.trc.2020.102885 

Nastjuk, I., Herrenkind, B., Marrone, M., Brendel, A., & Kolbe, L. M. (2020). What drives the 
acceptance of autonomous driving? An investigation of acceptance factors from an end-user’s 
perspective. Technological Forecasting and Social Change, 161, 120319. 
https://doi.org/10.1016/j.techfore.2020.120319 

Nielsen, A. T., Wang, J., & Pedersen, M. (2019). The role of technological innovation in the 
acceptance of autonomous vehicles. International Journal of Automotive Technology and 
Management, 19(1), 57-70. 

Nielsen, J., et al. (2019). Understanding the role of perceived convenience and technological 
familiarity in the acceptance of autonomous vehicles. Journal of Technology in Society, 47(3), 
58-73. 

Nielsen, T. A., Palmer, K., & Reimer, B. (2019). Older Adults' Willingness to Use Autonomous 
Vehicles: Factors Influencing Adoption and Public Policy Implications. Transportation Research 
Part F: Traffic Psychology and Behaviour, 62, 254-266. https://doi.org/10.1016/j.trf.2019.03.010 
OECD. (2022). Ageing and transport: Mobility needs and safety issues. OECD iLibrary. 
https://doi.org/10.1787/9789264249455-en 

Olivari, B., Baumgart, M., Lock, S., Whiting, C. G., Taylor, C. A., Iskander, J. K., Thorpe, P., & 
McGuire, L. C. (2018). CDC Grand Rounds: Promoting Well-Being and Independence in Older 
Adults. Morbidity and Mortality Weekly Report, 67(37), 1036–1039. 
https://doi.org/10.15585/mmwr.mm6737a4 

Ortman, J., Velkoff, V., & Hogan, H. (2021, October 8). An aging nation: the older population in 
the United States. US Census Bureau. 
https://www.census.gov/library/publications/2014/demo/p25-1140.html 

Othman, K. (2021). Impact of autonomous vehicles on the physical infrastructure: Changes and 
challenges. Designs, 5(3), 40. 

Parnell, P. R., & Joseph, T. (2019). Income disparities and internet adoption among older adults: 
A study of the digital divide. Journal of Digital Inclusion, 8(1), 10-22. 

Pek, J., & Flora, D. B. (2018). Reporting effect sizes in original psychological research: A 
discussion and tutorial. Psychological Methods, 23(2), 208–225. 

77 

 
 
 
 
 
 
 
 
 
 
 
 
Pew Research Center. (2023, September 22). American Trends Panel Datasets | Pew Research 
Center. https://www.pewresearch.org/american-trends-panel-datasets/ 

Portz, J. D., Bayliss, E. A., Bull, S., Boxer, R. S., Bekelman, D. B., Gleason, K., & Czaja, S. J. 
(2019). Using the Technology Acceptance Model to explore user experience, intent to use, and 
use behavior of a patient portal among older adults with multiple chronic conditions: Descriptive 
Qualitative study. Journal of Medical Internet Research, 21(4), e11604. 
https://doi.org/10.2196/11604 

Rahman, M. M., & Thill, J. C. (2023). What drives people’s willingness to adopt autonomous 
vehicles? A review of internal and external factors. Sustainability, 15(15), 11541. 

Rana, M. M., & Hossain, K. (2021). Connected and autonomous vehicles and infrastructures: A 
literature review. International Journal of Pavement Research and Technology, 16, 264-284. 

Reimer, B. (2014). Driver assistance systems and the transition to automated vehicles: A path to 
increase older adult safety and mobility? Public Policy & Aging Report, 24(1), 27-31. 

Rogers, W. A., Mitzner, T. L., Boot, W. R., Charness, N. H., & Czaja, S. J. (2017). Understanding 
individual and age-related differences in technology adoption. Innovation in Aging, 1(suppl_1), 
1026. https://doi.org/10.1093/geroni/igx004 

Rossmann, C. (2020). Theories of reasoned action and planned behavior. The International 
Encyclopedia of Media Psychology, 1–14. https://doi.org/10.1002/9781119011071.iemp0108 
Ryder, N. B. (1965). The cohort as a concept in the study of social change. American 
Sociological Review, 30(6), 847-861. https://doi.org/10.2307/2090964 

Sánchez-Meca, J., Marín-Martínez, F., & Chacón-Moscoso, S. (2003). Effect-size indices for 
dichotomized outcomes in meta-analysis. Psychological Methods, 8(4), 448-467. 

Schwarz, N. (1999). Self-reports: How the questions shape the answers. American Psychologist, 
54(2), 93-105. https://doi.org/10.1037/0003-066X.54.2.93 

Shaheen, S., & Wong, S. (2022). The future of public transit and shared mobility: Policy actions 
and research options for COVID-19 recovery. In Pandemic in the Metropolis (pp. 313-331). 
Springer. 

Sharma, I., & Mishra, S. (2021). Modeling consumers’ likelihood to adopt autonomous vehicles 
based on their peer network. Transportation Research Part D: Transport and Environment, 93, 
102929. 

Shekelle, P. G., et al. (2024). Gender differences in aging and health: A review of challenges in 
elderly women. American Journal of Public Health, 114(2), 145-154. 

Shekelle, P. G., Miake-Lye, I. M., Begashaw, M. M., Booth, M. S., Myers, B., Lowery, N., & 
Shrank, W. H. (2024). Interventions to Reduce Loneliness in Community-Living Older Adults: a 

78 

 
 
 
 
 
 
 
 
 
 
 
 
Systematic Review and Meta-analysis. Journal of General Internal Medicine, 39(4), 1015-1028. 
doi:10.1007/s11606-023-08517-5 

Shen, S., Ratnapradipa, K. L., Pervall, G. C., Sweeney, M., & Zhu, M. (2020). Driver license 
renewal laws and older adults’ daily driving, United States, 2003–2017. The Journals of 
Gerontology: Series B, 75(10), 2268–2277. https://doi.org/10.1093/geronb/gbz162 

Smith, M. L., & Green, R. D. (2020). Political ideologies and autonomous vehicles: A divide in 
societal acceptance. Journal of Political Technology, 18(1), 12-28. 

Soósová, M. S. (2016). Determinants of quality of life in the elderly. Central European Journal of 
Nursing and Midwifery, 7(3), 484–493. https://doi.org/10.15452/cejnm.2016.07.0019 

Suragarn, U., Hain, D., & Pfaff, G. (2021). Approaches to enhance social connection in older 
adults: an integrative review of literature. Aging and Health Research, 1(3), 100029. 
https://doi.org/10.1016/j.ahr.2021.100029 

Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The psychology of survey response. 
Cambridge University Press. 

Tsertsidis, A., Kolkowska, E., & Hedström, K. (2019). Factors influencing seniors’ acceptance of 
technology for ageing in place in the post-implementation stage: A literature review. 
International Journal of Medical Informatics, 129, 324–333. 
https://doi.org/10.1016/j.ijmedinf.2019.06.027 

Valliant, J. C. D., Burris, M., Czebotar, K., Stafford, P. B., Giroux, S., Babb, A., Waldman, K. B., 
& Knudsen, D. C. (2021). Navigating food insecurity as a rural older adult: the importance of 
congregate meal sites, social networks and transportation services. Journal of Hunger & 
Environmental Nutrition, 17(5), 593–614. https://doi.org/10.1080/19320248.2021.1977208 

Valor, J. (2003). Internet usage as an indicator of technological proficiency. Journal of 
Information Technology, 18(4), 255-269. 

Valor, J., & Sieber, S. (2003). Uses and attitudes of young people toward technology and mobile 
telephony. Social Science Research Network. https://doi.org/10.2139/ssrn.459222 

van Eldijk, J., Gil, J., & Marcus, L. (2022). Disentangling barrier effects of transport 
infrastructure: Synthesising research for the practice of impact assessment. European Transport 
Research Review, 14(1), 1-15. Springer. 

Venkatesh, V., & Bala, H. (2008). Technology Acceptance Model 3 and a Research Agenda on 
Interventions. Decision Sciences, 39(2), 273-315. https://doi.org/10.1111/j.1540-
5915.2008.00192.x 

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of 
information technology: Toward a unified view. MIS quarterly, 425-478. 

79 

 
 
 
 
 
 
 
 
 
 
 
 
Wassmer, R. W. (2008). Causes of Urban Sprawl in the United States: Auto reliance as compared 
to natural evolution, flight from blight, and local revenue reliance. Journal of Policy Analysis and 
Management, 27(3), 536–555. https://doi.org/10.1002/pam.20355 

Williams, E. M., & Hill, K. D. (2022). Bridging the digital divide: How Black and Hispanic 
communities are adopting technology. Journal of Technology and Society, 19(4), 145-160. 

Yan, Y., Deng, Y., Igartua, J. J., & Song, X. (2023). Does internet use promote subjective well-
being? Evidence from different age groups based on CGSS 2017 data. International Journal of 
Environmental Research and Public Health, 20(4), 2897. https://doi.org/10.3390/ijerph20042897 

Zandieh, R., & Acheampong, R. A. (2021). Attitudes toward four levels of self-driving 
technology among older adults in Greater Manchester, UK. Frontiers in Psychology, 12, 682973. 

Zhang, X., Wang, Z., Zhang, D., Ye, D., Zhou, Y., & Qin, J. (2023). “The prevalence and 
treatment rate trends of osteoporosis in postmenopausal women.” PLOS ONE, 18(9), e0290289. 
doi:10.1371/journal.pone.0290289 

Zhu, C., Susskind, J., Giampieri, M., O’Neil, H. B., & Berger, A. M. (2023). Optimizing 
sustainable suburban expansion with autonomous mobility through a parametric design 
framework. Land, 12(9), 1786. 

Zoaktafi, M., Gandy, T., & Yoon, J. S. (2023). Are rural older adults ready to adopt autonomous 
vehicles (AV)? Predictors of attitudes toward Autonomous Vehicle in rural areas. Innovation in 
Aging, 7(Supplement_1), 957. 

80 

 
 
 
 
 
 
 
 
 
Adoption Path Model 

APPENDIX 

model <- ' 
# Direct effect 
#Exogenous Binary Variable ~ Endogenous Variables 
wantyes ~ c2*age2 + c3*age3 + c4*age4 + b*Internetc + Woman + otherg + 
moderateconservative + moderateliberal + liberal + thirtyk + fourtyk + fiftyk + sixtyk + 
seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + PostGrad + Black + 
Hispanic + otherr + Asian 
# Mediator 
Internetc ~ a2*age2 + a3*age3 + a4*age4 + Woman + otherg + moderateconservative + 
moderateliberal + liberal + thirtyk + fourtyk + fiftyk + sixtyk + seventyk + eightyk + 
ninetyk + hundredk + assoc + collegeGrad + PostGrad + Black + Hispanic + otherr + 
Asian 
# Indirect effect 
ab2 := a2*b 
ab3 := a3*b 
ab4 := a4*b 
# Total effect 
total2 := c2 + ab2 
total3 := c3 + ab3 
total4 := c4 + ab4/ 
' 
# Estimate the model 
fit <- sem( 

model, data = cars,  
#Estimator Weighted Least Squares Mean and Variance-adjusted 
estimator = "WLSMV",  
sampling.weights = "weight",  
#Binary Variable marked as Categorical 
ordered = c("wantyes")) 
# Summary of the model 
summary(fit, fit.measures = TRUE, standardized = TRUE) 

Endorsement Path Model 
model <- ' 
# Direct effects with unique paths for each outcome 
societyyes ~ c2_society*age2 + c3_society*age3 + c4_society*age4 + 
b_society*Internetc + Woman + otherg + moderateconservative + moderateliberal + 
liberal + thirtyk + fourtyk + fiftyk + sixtyk + seventyk + eightyk + ninetyk + hundredk + 
assoc + collegeGrad + PostGrad + Black + Hispanic + otherr + Asian 

stressfulyes ~ c2_stress*age2 + c3_stress*age3 + c4_stress*age4 + b_stress*Internetc + 
Woman + otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk 
+ fiftyk + sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + 
PostGrad + Black + Hispanic + otherr + Asian 

81 

 
 
hackyes ~ c2_hack*age2 + c3_hack*age3 + c4_hack*age4 + b_hack*Internetc + Woman 
+ otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk + fiftyk + 
sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + PostGrad + 
Black + Hispanic + otherr + Asian 

crashyes ~ c2_crash*age2 + c3_crash*age3 + c4_crash*age4 + b_crash*Internetc + 
Woman + otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk 
+ fiftyk + sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + 
PostGrad + Black + Hispanic + otherr + Asian 

sharingyes ~ c2_share*age2 + c3_share*age3 + c4_share*age4 + b_share*Internetc + 
Woman + otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk 
+ fiftyk + sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + 
PostGrad + Black + Hispanic + otherr + Asian 

olderyes ~ c2_older*age2 + c3_older*age3 + c4_older*age4 + b_older*Internetc + 
Woman + otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk 
+ fiftyk + sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + 
PostGrad + Black + Hispanic + otherr + Asian 

taxiyes ~ c2_taxi*age2 + c3_taxi*age3 + c4_taxi*age4 + b_taxi*Internetc + Woman + 
otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk + fiftyk + 
sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + PostGrad + 
Black + Hispanic + otherr + Asian 

busyes ~ c2_bus*age2 + c3_bus*age3 + c4_bus*age4 + b_bus*Internetc + Woman + 
otherg + moderateconservative + moderateliberal + liberal + thirtyk + fourtyk + fiftyk + 
sixtyk + seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + PostGrad + 
Black + Hispanic + otherr + Asian 

# Mediator 
Internetc ~ a2*age2 + a3*age3 + a4*age4 + Woman + otherg + moderateconservative + 
moderateliberal + liberal + thirtyk + fourtyk + fiftyk + sixtyk + seventyk + eightyk + 
ninetyk + hundredk + assoc + collegeGrad + PostGrad + Black + Hispanic + otherr + 
Asian 

# Correlate error terms 
societyyes ~~ stressfulyes 
societyyes ~~ hackyes 
societyyes ~~ crashyes 
stressfulyes ~~ hackyes 
stressfulyes ~~ crashyes 
hackyes ~~ crashyes 

# Indirect effects 
ab2_society := a2 * b_society 

82 

 
 
 
 
 
 
 
 
ab3_society := a3 * b_society 
ab4_society := a4 * b_society 

ab2_stress := a2 * b_stress 
ab3_stress := a3 * b_stress 
ab4_stress := a4 * b_stress 

ab2_hack := a2 * b_hack 
ab3_hack := a3 * b_hack 
ab4_hack := a4 * b_hack 

ab2_crash := a2 * b_crash 
ab3_crash := a3 * b_crash 
ab4_crash := a4 * b_crash 

ab2_share := a2 * b_share 
ab3_share := a3 * b_share 
ab4_share := a4 * b_share 

ab2_older := a2 * b_older 
ab3_older := a3 * b_older 
ab4_older := a4 * b_older 

ab2_taxi := a2 * b_taxi 
ab3_taxi := a3 * b_taxi 
ab4_taxi := a4 * b_taxi 

ab2_bus := a2 * b_bus 
ab3_bus := a3 * b_bus 
ab4_bus := a4 * b_bus 

# Total effects 
total2_society := c2_society + ab2_society 
total3_society := c3_society + ab3_society 
total4_society := c4_society + ab4_society 

total2_stress := c2_stress + ab2_stress 
total3_stress := c3_stress + ab3_stress 
total4_stress := c4_stress + ab4_stress 

total2_hack := c2_hack + ab2_hack 
total3_hack := c3_hack + ab3_hack 
total4_hack := c4_hack + ab4_hack 

total2_crash := c2_crash + ab2_crash 
total3_crash := c3_crash + ab3_crash 

83 

 
 
 
 
 
 
 
 
 
 
 
total4_crash := c4_crash + ab4_crash 

total2_share := c2_share + ab2_share 
total3_share := c3_share + ab3_share 
total4_share := c4_share + ab4_share 

total2_older := c2_older + ab2_older 
total3_older := c3_older + ab3_older 
total4_older := c4_older + ab4_older 

total2_taxi := c2_taxi + ab2_taxi 
total3_taxi := c3_taxi + ab3_taxi 
total4_taxi := c4_taxi + ab4_taxi 

total2_bus := c2_bus + ab2_bus 
total3_bus := c3_bus + ab3_bus 
total4_bus := c4_bus + ab4_bus 
' 

# Estimate the model 
fit <- sem( 
  model, data = cars,  
  estimator = "WLSMV",  
  sampling.weights = "weight",  
  ordered = c("societyyes", "stressfulyes", "hackyes", "crashyes", "sharingyes", 
"olderyes", "taxiyes", "busyes") 
) 

# Summary of the model 
options(max.print = 10000) 
summary(fit, fit.measures = TRUE, standardized = TRUE) 

Odds Ratio 

# Fit logistic regression separately for direct paths 
logit_model <- glm(wantyes ~ age2 + age3 + age4 + Internetc + Woman + otherg + 
moderateconservative + moderateliberal + liberal + thirtyk + fourtyk + fiftyk + sixtyk + 
seventyk + eightyk + ninetyk + hundredk + assoc + collegeGrad + PostGrad + Black + 
Hispanic + otherr + Asian, data = cars, family = binomial, weights = weight) 

# Calculate odds ratios by exponentiating the coefficients 
odds_ratios <- exp(coef(logit_model)) 
odds_ratios 

# Confidence intervals for odds ratios 
confint_odds_ratios <- exp(confint(logit_model)) 
confint_odds_ratios 

84