Content uploaded by Nirajan Shiwakoti
Author content
All content in this area was uploaded by Nirajan Shiwakoti on Sep 01, 2023
Content may be subject to copyright.
CORRECTED PROOF
Case Studies on Transport Policy xxx (xxxx)101072
Contents lists available at ScienceDirect
Case Studies on Transport Policy
journal homepage:www.elsevier.com/locate/cstp
Analysis of Australian public acceptance of fully automated vehicles by
extending technology acceptance model
Yilun Chen a,⁎,Shah Khalid Khan a,Nirajan Shiwakoti a,Peter Stasinopoulos a,Kayvan Aghabayk b
aSchool of Engineering,RMIT University,Melbourne,Australia
bSchool of Civil Engineering,College of Engineering,University of Tehran,Tehran,Iran
ARTICLE INFO
Keywords:
Intelligent Transportation System
Driverless car
Transport policy
Smart mobility
User behaviour
ABSTRACT
There has been an increasing trend in using user acceptance models to explore the public acceptance of auto-
mated vehicles (AVs)in different countries.Most of the previous studies have analysed perceived usefulness,be-
havioural attitude,subjective norms as well as perceived ease of use,but other important factors,such as trust
and data privacy,have not been adequately considered.Likewise,public perceptions of fully AVs are limited in
the literature as most studies focus on different levels of automation.This study aims to assess the behavioural in-
tention to use fully AVs in Australia by extending the Technology Acceptance Model (TAM)that includes data
privacy and trust in the TAM constructs.Based on a survey of 809 adult respondents from Australia,the model
was evaluated with Structural Equation Modelling.
The research revealed perceived trust and perceived data privacy is the first and second most important vari-
able affecting the attitude,followed by perceived ease of use and usefulness.Perceived data privacy was discov-
ered to positively impact attitude,perceived trust,perceived usefulness,perceived ease of use as well as behav-
ioural intentions.The perceived trust mediated perceived data privacy on the attitudes in this study.Addition-
ally,the two major variables in the proposed model –perceived trust and data privacy-affect attitudes of AVs sig-
nificantly,with total effects being 0.637 and 0.604,respectively.Attitude is the most significant variable that
correlates with behavioural intentions,which leads to acceptance of fully AVs.Multigroup analysis showed the
gender,age and income related differences regarding the public acceptance of AVs.Several theoretical and prac-
tical implications are discussed in this paper.
1.Introduction
In recent years,automated vehicles (AVs)have become a main-
stream topic that attracts massive attention from the World’s Tech Gi-
ants,car manufacturers as well as mobility service providers (Duan et
al., 2023). AVs utilize the technology to partially or entirely displace
human-driven vehicles while preventing road hazards and providing
productive time to the driver.Although researchers,industry leaders
and university professors have predicted different timelines for market
penetration and full adoption from the public,most people anticipate
that fully AVs (level 5 automation)technology will mature around
2030 (Underwood,2014). Therefore,it is important to gauge the public
perception towards using fully AVs before they are on the market.For
example,a better understanding of public perception can help the gov-
ernment make investment decisions and future planning,help automo-
bile manufacturers design products that fit the customer and indicate
the future research direction for the industry.
Although many studies have analysed public perception towards
AVs,few studies have undertaken an in-depth investigation of public
perception of fully AVs,especially in the Australian context.For exam-
ple, (Cunningham et al., 2019)investigated public perception and opin-
ions of AVs by employing a national survey in the Australian context
and received 5089 respondents.The study found that although most
Australian people realised the potential benefits of AVs,they still have
considerable concerns about AVs.In a different study by Soltani et al.
(2021)at the University of South Australia,Adelaide,152 students took
part in a survey.The study used structural equation modeling to exam-
ine students'views on AVs.It was found that students in applied sci-
ences and those who drive to campus,especially younger males,are
more willing to accept AVs.Nonetheless,many are worried about cy-
bersecurity and whether AVs can be trusted to work safely.The insights
from the study are useful in promoting self-driving cars in specific areas
like university campuses.However,the study only looked at campus
setting,so public adoption of AVs in a broader context needs further in-
⁎Corresponding author at:School of Engineering,RMIT University,City Campus,124 La Trobe Street,Melbourne 3000,Victoria,Australia.
E-mail address:s3598361@student.rmit.edu.au (Y.Chen).
https://doi.org/10.1016/j.cstp.2023.101072
Received 27 February 2023;Received in revised form 23 June 2023;Accepted 28 August 2023
2213-624/© 20XX
Note: Low-resolution images were used to create this PDF. The original images will be used in the final composition.
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
vestigation.A study by Acheampong et al. (2023)focused AVs adoption
in Greater Manchester in the UK and Melbourne in Australia.The study
discussed various potential uses of AVs,including public transport,per-
sonal and shared-use,and urban freight through participatory work-
shops.It also recognized the challenges for AVs adoption,including the
need for strong governance,data privacy,and cybersecurity.
Another study Kaye et al. (2021)focused on individual acceptance
for AVs from automation levels 3 to 5 by using meta-analytic structural
equation modelling.By analysing perceived ease of use (PEU), per-
ceived behavioural control,subjective norms,attitude and perceived
usefulness (PU), the study found that these factors significantly posi-
tively affect intentions to use AVs.However,other factors have not
been considered in the study by Kaye et al. (2021),such as trust and
data privacy.The study Acharya and Mekker,2022)revealed that trust
directly impacts perceived data privacy and security on Connected Ve-
hicles (CV)by developing a user acceptance model in the US context.
Additionally,a study from (Korkmaz et al., 2022)implemented the uni-
fied theory of acceptance and use of technology 2 (UTAUT2)model to
gauge autonomous public transport systems acceptance and concluded
performance expectancy,social influence,habit,and safety,as well as
trust,have significant positive impacts toward intention to use.How-
ever,the data sample is limited to young people,and the perception of
AV technology is limited to self-driving cars,given that different AV
types could be developed for public transport (PT).
Hoff and Bashir (2015)systematically reviewed the factors that can
affect trust in automation by investigating three layers of trust:disposi-
tional trust,learned trust and situational trust,which could help de-
sign procedures to encourage trust.Meyer-Waarden and Cloarec
(2022)developed a conceptual model using the rationale of technol-
ogy acceptance by incorporating performance expectancy,social
recognition,technology security and personal concerns.The survey
used in the model was conducted in France and the results from struc-
tural equation modelling demonstrated that there are positive relation-
ships between integration to use AI-powered AVs and well-being,so-
cial recognition and technology trust.Similarly,Chen (2019)utilised
the technology acceptance model (TAM)to investigate the factors af-
fecting people's acceptance of autonomous shuttle services by inviting
participants to have test rides.The study concluded that the perceived
usefulness is associated to the attitude in a positive way but not to the
intention to use.
In summary,there has been an increasing trend in using user accep-
tance models to explore the public acceptance of AVs with different lev-
els of automation in different countries.Comparatively,the study of
public acceptance of fully AVs has been less investigated.Consideration
of other theoretical constructs (e.g., data privacy and trust)in tradi-
tional user acceptance models to investigate public acceptance towards
AVs is an emerging area.This study aims to evaluate the behavioural in-
tention of people to use fully AVs (level 5)in Australia by extending the
TAM model.The main contributions from this study are as follows:
a)We investigate public acceptance of fully AVs in Australia by us-
ing a modified Technology Acceptance Model (TAM)by including trust
as well as data privacy constructs.According to our knowledge,it is the
first study in Australia to use the modified TAM model by including
trust and data privacy constructs to explore the public acceptance of
fully automated vehicles.Due to demographic and psychological con-
texts,different countries have different perceptions towards AVs (Chen
et al., 2022). Although the comprehensive study by Cunningham et al.
(2019)demonstrated that most Australian people tend to accept AVs,
the author stated that perceived value as well as trust in this technology
would be a concern with the majority of Australian people not wanting
to pay more for the fully AVs than the conventional cars.Therefore,we
conducted a questionnaire survey in Australia to obtain more insights
into these perceived values via modified TAM,and analysis of causal re-
lationships that could affect the decision-making process.The survey
questionnaire developed in this study can be used by other researchers
investigating Australian public acceptance of fully AVs in future.The
findings from this study will help decision-makers better understand
users'acceptance to formulate suitable strategies to increase the adop-
tion and deployment of fully AVs.
b)Although many studies have utilised the different user acceptance
models in assessing AVs acceptance,limited studies incorporated the
perceived trust and data privacy in assessing fully AVs by extending the
TAM model.By considering these two important factors,the study will
provide a more comprehensive investigation of AVs acceptance in Aus-
tralia,given that Australians tend to be concerned with data privacy is-
sues (Cunningham et al., 2019).
The paper is structured into the following sections.Section 2 illus-
trates the research background by conducting the literature review for
user acceptance models.Then section 3 describes the proposed model
using the rationale of TAM and the relevant hypothesis development,
while section 4 illustrates survey design,data collection and methodol-
ogy of data analysis.After that,Sections 5 and 6 exhibit the results and
discuss their theoretical and practical implications,respectively.Fi-
nally,section 7 will summarise this study's key findings and limitations.
Table 1 presents the acronyms used in this study.
2.Literature review
2.1.User acceptance models
The current trend in transport mode choice is that people in Aus-
tralia prefer using private cars because of the long distance to public
transport services and less population density (Boulange et al., 2017).
Therefore,the introduction of new mobility services such as AVs will
change the norm;however,it is not clear how the trend will evolve.
Numerous studies have evaluated the perceptions of AVs from the
public in different countries based on various factors.The key factors
that could impact people's acceptance of AVs are safety,perceived ease
of use as well as perceived usefulness (Xu et al., 2018). Other demo-
graphic factors that play important roles are educational level,age,
gender and income level.
In order to investigate and identify the factors impacting people's
motivation to utilize new technology (AVs), researchers and industry
widely used the TAM (Davis,1989). Davis (1989)illustrated that the
idea of adopting a new technology counts on two critical factors:per-
ceived ease of use (PEU)and perceived usefulness (PU)of the product
or technology.PEU is defined as the extent to which people believe us-
Table 1
Acronyms.
Abbreviation Explanation
AVs Automated Vehicles
TAM Technology Acceptance Model
PT Perceived Trust
PEU Perceived Ease of Use
PU Perceived Usefulness
BI Behavioural Intention
PDP Perceived Data Privacy
UTAUT Unified Theory of Acceptance and Use of
Technology
PT Public Transport
TPM Theory of Reasoned Action
TPB Theory of Planned Behaviour
BEV Battery Electric Vehicles
ABS Australian Bureau of Statistics
CFA Confirmatory factor analysis
SRMR Standardised Root Mean Square Residual
RMSEA Root Mean Square Error Approximation
SAVE Square Root of Average Variance Extracted
CR Composite Reliability
SEM Structural equation modelling
ATU Attitude
2
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
ing a system will take fewer efforts,while PU is related to the extent to
which people believe using the technology can enhance job perfor-
mance.Acharya and Mekker (2022)examined the public acceptance to-
wards connected vehicles (CVs)by developing a new acceptance model
which extends the TAM.Perceived data privacy as well as security for
the new technology,were found to affect the acceptance of CVs on top
of the original TAM constructs:perceived ease of use as well as useful-
ness.
A study by Hegner et al. (2019)used TAM as a foundation to exam-
ine autonomous vehicle adoption.The study integrated elements like
trust in technology and concerns about relinquishing control,along
with traditional TAM factors such as perceived ease of use and per-
ceived usefulness.Additionally,it incorporated “driving enjoyment”as
a barrier and “personal innovativeness”as a facilitator to adoption in
the proposed model.Through an online survey,the paper concluded
that trust,concerns about control,and perceived usefulness are signifi-
cant factors for the adoption of autonomous vehicles.While driving en-
joyment hindered the adoption innovativeness promoted the adoption
of AVs.Zhang et al. (2020)also employed the TAM to understand pub-
lic acceptance of AVs and modified it by incorporating social and per-
sonal factors such as initial trust,social influence,and personality traits
(e.g., neuroticism,conscientiousness,agreeableness and openness).
Through a questionnaire survey of 647 drivers in China,the study
found that social influence and initial trust are pivotal in the early
adoption of AVs.Additionally,individuals open to new experiences and
sensation seekers were found to be more inclined to trust and using
AVs.
Based on TAM,Venkatesh et al. (2003)incorporated eight previ-
ously established models and developed a unified model titled the Uni-
fied Theory of Acceptance and Use of Technology (UTAUT). The unified
model includes three major determinants for intention towards using
(performance expectancy,effort expectancy as well as a social influ-
ence)and two major factors for practice behaviour (facilitating condi-
tions and intention). Similarly,Adnan et al. (2018)conducted a com-
prehensive review of user acceptance and ethical implications of au-
tonomous vehicles.The paper extended UTAUT by integrating ethical
implications and cost as factors influencing user acceptance.The impor-
tance of trust was emphasized in the study.Further,the study advo-
cated for incorporating ethical considerations in future studies to ad-
dress consumer acceptance by predicting that autonomous vehicles will
mature by 2030 and dominate by 2050.
Zhang et al. (2019)tested an extended TAM model to understand
people's attitudes and intentions to use AVs and concluded that trust is
the most critical factor affecting people's motivation to use AVs.How-
ever,Zhang et al. (2021)found perceived ease of use makes the least
contribution,except in Europe,Asia,partially AVs and young sub-
groups.Another study investigated the factors affecting the adoption
of the autonomous public transport system by expanding the UTAUT
model.The proposed model found that social influence,trust,safety as
well as performance expectancy have significant positive effects on the
adoption intention (Korkmaz et al., 2022).
By extending the Theory of Reasoned Action Model (TRA) (Fishbein
and Ajzen,1977), the theory of Planned Behaviour (TPB)contains three
components:attitudes (TAM), perceived behavioural control as well as
subjective norms (Ajzen,1991). The study (Karuppiah and Ramayah,
2023)utilised TPB as a base theory to identify factors significantly im-
pacting consumers'intention to buy hybrid cars in Malaysia.However,
the study did not consider another important factor:brand trust.An-
other study Haustein and Jensen (2018)based on TPB incorporated
personal norms,perceived mobility necessities and battery electric ve-
hicles (BEV)experience to investigate the critical factors impacting the
adoption of BEV in Denmark and Sweden.Although the results indi-
cated that charging infrastructure is the most essential factor in encour-
aging people to use BEV,the study expected the results to be only trans-
ferable to other European countries,but this needs to be confirmed by
future research.
A study by Cugurullo and Acheampong (2023)examined public in-
tention to use Artificial Intelligence (AI)technology,particularly in au-
tonomous cars,by integrating the TBP,TAM,and Technology Diffusion
Theory into a proposed framework named “SCALE”,which includes
constructs on fear,perceived usefulness,perceived ease of use,image,
status and instrumentality.The framework considered the advantages
of AI alongside fears by categorizing benefits into individual,urban,
and global aspects.The study concluded that while there is substantial
fear among the public,the perceived benefits drive AI adoption.Ad-
dtiontally,Zhang et al. (2023)reviewed the literature on autonomous
vehicle acceptance,focusing on factors (e.g., trust,cost,and safety)and
theoretical modelling methodology,including TAM,UTAUT,TPB and
IDT (Innovation Diffusion Theory). The study emphasized the need for
more in-depth experimental studies in diverse contexts as technology
matures.Further,it was suggested to improve human-vehicle collabo-
ration through better human–machine interfaces and understanding
context-specific influencing factors to enhance acceptance.
2.2.Data privacy
Keszey (2020)provided a systematic review of factors influencing
the adoption of AVs and extended the framework by testing new
variables.The study defined data privacy as concerns about the man-
agement and protection of personal information by AV's intercon-
nected systems.According to the study,data privacy played a signifi-
cant role in accepting AVs.It is argued that if people are apprehen-
sive about how AVs handle and secure their personal information,
they are likely to have a reduced intention to use the technology.
The study concluded that for AVs adoption,individuals must trust
that AVs offer sufficient data privacy and security.
Similarly,Tan and Taeihagh (2021)discussed that data privacy in
AVs involves managing the personal information collected by AVs,such
as travel habits.The study emphasized the need for stringent gover-
nance,as companies operating AVs may exploit or sell this data,poten-
tially leading to privacy breaches or systemic discrimination.The paper
suggested that clear data privacy regulations and guidelines specific to
AVs are essential for public acceptance and successful deployment,as
individuals are likely to be concerned about how their data is being
used and protected.Furthermore,Panagiotopoulos and
Dimitrakopoulos (2018)discussed data privacy in relation to AVs as
part of the perceived trust factor.Data privacy in that study referred to
the protection of personal and sensitive information that AVs may col-
lect and process.The study revealed that for consumers to accept and
adopt AVs,they need to trust the technology,which includes having
confidence in how AVs handle data privacy.Gaining trust from end-
users concerning data privacy is highlighted as critical for the wide-
spread acceptance and deployment of AVs in the future.
Lastly,Khan et al. (2023)discussed the crucial role of data privacy
as one of the six cyber impediments impacting the deployment of AVs.
It highlighted that data privacy concerns are threefold:consumer-level
information,including location and personal details,proximity espi-
onage through AV sensors,and intellectual property of vendors.Khan
et al. (2023)also revealed that data privacy is perceived differently
based on gender,age,education level,understanding of AVs,and cyber-
security knowledge.Specifically,women and older individuals are
more concerned about data privacy.As AV understanding and cyberse-
curity knowledge increased,data privacy became a more significant
barrier to AV adoption.
2.3.AI in AVs
Brauner et al. (2023)indicated that AI’s pervasiveness,including in
AVs,is evaluated diversely in terms of likelihood and acceptability.
3
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
The balance between technological advancements and societal value
could affect public perception of AI in AVs.The study revealed differ-
ences in individual perceptions,which may impact the adoption of AI-
based technologies like AVs.The paper emphasized the necessity for
public discourse and regulatory frameworks regarding AI applications
and highlighted the importance of education in AI to foster informed
discussions and decision-making.
AI is also presented as a central component in the development of
AVs,mainly driving their functionality and safety (Giannopoulos and
Munro,2019). AI's role in enabling vehicle-to-vehicle communication
and learning from the dynamic environment is crucial for AVs'safety,
which can significantly impact public perception and adoption.The in-
vestments and innovations in AI within the AV sector are primarily
from the private sector,and strategic partnerships are playing a signifi-
cant role.Giannopoulos and Munro (2019)suggested that as AI contin-
ues to improve the safety and efficiency of AVs,it is likely to positively
influence people's perception and adoption.
Carpenter (2020)discussed how AI in AVs affects public perception
and adoption.It highlighted the shift in people’s roles from drivers to
passengers and the importance of understanding their emotional re-
sponses in an AI-controlled environment.By framing the AVs as an ex-
tension of personal spaces,the paper underlined the emotional attach-
ment and territoriality associated with it (Carpenter,2020). Addition-
ally,the study brought attention to the idea of a “kill switch”allowing
humans to override AI control,and how it could be crucial for user ac-
ceptance by providing a sense of safety and control over the AI system.
3.Extended technology acceptance model and hypothesis
development
This study is conducted by using the extension of the original TAM
(Davis,1989)incorporating two important factors:perceived trust and
data privacy as shown in Fig.1.It has been shown that trust (Kaur and
Rampersad,2018)and data privacy (Rahimi et al., 2020)may influence
the adoption of automated vehicles.
Kaur and Rampersad (2018)utilised TAM theory and established a
new model by containing security,performance expectancy,reliability,
privacy,adoption as well as trust.The data was collected from staff and
university students in Australia to investigate the critical factors im-
pacting the adoption of AVs to connect to the nearby station.The study
revealed that the significant concerns included data privacy and secu-
rity.Similarly,another study concluded that trip privacy and data pri-
vacy concerns presented barriers in all model users for the adoption of
AVs by using latent class clustering analysis and structural equation
models (Rahimi et al., 2020). In addition,the study Molnar et al. (2018)
concluded that one of the most important factors that emerged as a
strong influence on the adoption of automated vehicles is trust.
Fig.1.Proposed TAM.
The actual usage of new technology is a critical factor in measuring
technology acceptance;however,it could not be assessed before the ac-
tual technology comes into reality.Therefore,the behavioural intention
could be used as a key factor of adopting technology acceptance be-
cause several studies have verified the relationship between the inten-
tion to use new technology and behavioural intention (Venkatesh et al.,
2012). In the following sub-sections,the extended TAM variables are
described along with the relevant hypothesis.
3.1.Attitude
Attitude represents the people's intention to adopt the new technol-
ogy.Attitude towards adopting the new technology is defined as an ex-
tent to which people have positive or negative sentiments about the
new technology:AVs.
3.2.Perceived usefulness
In the original TAM model, (Davis,1989)proposed the perceived
usefulness as a subjective likelihood of using a specific system to in-
crease quality of work within an organisational context.Therefore,the
perceived usefulness is an important factor in describing how helpful
and useful this technology fulfils their expectations.The benefits from
AVs in terms of mobility and safety could increase the motivation to use
AVs for an individual.This study assumed that:
H1:Perceived usefulness positively influences attitude to use AVs
3.3.Perceived ease of use
Perceived ease of use which is defined by Davis (1989)is the extent
to which the users expect the target system will take fewer efforts.
Therefore,H2 and H3 are hypothesised in the original TAM model.
H2:Perceived ease of use positively influences attitude to use AVs
H3:Perceived ease of use positively influences perceived usefulness of
AVs
3.4.Perceived trust
Perceived trust is defined as the extent to which users find AVs tech-
nology trustworthy,reliable and dependable.The association of trust
with perceived usefulness as well as perceived ease of use is ambiguous
because several studies found that trust is the most critical factor affect-
ing the adoption of new technology (Susanto and Aljoza,2015;Kenesei
et al., 2022).
Therefore,H4 is hypothesised that trust towards AVs technology in-
fluences people's attitude towards using AVs,which will affect their in-
tention to use AVs.
H4:Perceived trust positively influences attitude to use AVs
Also,trust demonstrated a substantial direct impact on behavioural
intention as well as perceived usefulness (Choi and Ji,2015)and this
indicates the trust is a major factor in understanding the adoption of au-
tonomous vehicles.Zmud et al. (2016)found that lack of trust is a top
reason for people being unlikely to use AVs.However,there are some
studies assuming that trust is not related to PEOU and PU because the
three factors affect BI independently (Buckley et al., 2018). Therefore,
H5 is hypothesised that the perceived trust influences perceived ease of
use.
H5:Perceived trust positively influences perceived ease of use
3.5.Perceived data privacy
Perceived data privacy is the extent to which people are worried
about the data collection of any personal trip.Therefore,the extent to
which an individual care about the unauthorised collection and access
of data represents the perceived data privacy.Kim et al. (2008)con-
4
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
cluded that the impact of uncertainties as well as risks on the accep-
tance of new technology is affected by trust;therefore,hypothesis H6 is
established to identify the effects of data privacy on the acceptance of
AVs.
H6:Perceived data privacy positively influences trust
All factors,including behavioural intention,perceived usefulness,
perceived ease of use as well as attitude,are affected by public opinions
of online media (Lin and Kim,2016). Wang and Zhao (2019)concluded
that the perceived data privacy negatively influences the risk on the be-
havioural intention for using ride-sharing services.Therefore,hypothe-
ses H7 to H9 are proposed:
H7:Perceived data privacy positively influences perceived usefulness of
AVs
H8:Perceived data privacy positively influences perceived ease of use of
AVs
H9:Perceived data privacy positively influences attitude towards of us-
ing AVs
3.6.Behavioural intention
When the new technology is rolled out,the perceived usefulness
positively influences on behavioural intentions for the new technology.
Additionally,TAM indicated that attitude positively influences on be-
havioural intentions as well.Therefore,hypotheses H10 and H11 are
proposed:
H10:Perceived usefulness positively influences on the behavioural inten-
tions of AVs
H11:Attitude positively influences on the behavioural intentions of AVs
4.Methodology
4.1.Data collection
Qualtrics,a private research company which specialised in building
and distributing surveys via online panels of respondents,was engaged
to collect the online survey data.809 Australian participants data were
collected,and they were of at least of 18 years of age and illustrative of
the Australian general population structure regarding age,gender and
number of people in different regions.The sample data was collected by
aligning with the census data from the Australian Bureau of Statistics
(ABS,2021).
There are five levels of automation of AVs.Our questions in the sur-
vey are related to level 5 (full automation), meaning the vehicle does
not require any human control to drive the vehicle.In the question-
naire,we briefly described the automated cars and different levels of
automation to the participants as follows:“It is anticipated that driver-
less cars will be on the road in the next 10 to 15 years.Automated cars
use advanced technology and software to automate driving tasks such
as steering,acceleration,lane changing,parking,and braking.For ex-
ample,an automated car can transport you from home to your work-
place or vice versa without you driving the car.While inside the auto-
mated car,you can relax and do your other work on your mobile,laptop
or enjoy the environment outside while the automated car drives you
towards your destination.“
The survey questionnaire was designed to understand the percep-
tion towards AVs and transport mode choice in different situations after
consulting with three academics and two transportation industry peo-
ple.Then the survey was distributed to them for a pilot test,and the sur-
vey was modified to incorporate their feedback for the final survey.The
questionnaire has four sections,starting with socio-demographics ques-
tions,followed by transport mode choice preference questions in differ-
ent hypothetical situations,automated car perceptions and motivation-
related questions for AVs.
Firstly,“soft launch data”was collected to identify any concerns for
data quality.Then the low-quality data was removed by comparing
with median response time and crosschecking some answers (e.g.peo-
ple who chose “no driving experience”but they chose “private vehicles”
as primary mode,are contradictory to each other). The “full launch”
was conducted between mid-October 2022 to early November 2022.
The order of survey items was randomised to mitigate side-on effects.
All the participants'information was completely anonymous,and the
survey items were checked and approved by University Human Re-
search Ethics Committee (Review Reference:2022-25201-18632). On
average,participants took 16 min to complete the survey.
A sample of 809 people from different states of Australia filled in the
survey.Of those,most people are from New South Wales (33.3%) and
Victoria (26.1%), followed by Queensland (19.8%). There is an equal
gender ratio between males (49.0%) and females (50.4%). Moreover,
age group 25 to 34 (21.4%) represented the most people in the survey,
followed by the age group 45 to 54 (18.4%) and 35 to 44 (17.1%). The
distribution of states,gender and age group closely align with ABS 2021
data.
By looking at details of the socio-demographic information,most
people have “Certificates/Diploma/Vocational,or Equivalent”degree
(46.5%) followed by “Undergraduate:Bachelor's degree,Certificates/
Diploma,or Equivalent”degree.Of those 809 respondents,around the
majority of people (68.4%) claimed that “driving their own private ve-
hicles”as a primary mode for commuting purposes and around 31.6%
(the most)and 36% (the most)of these people stated that they spent 10
to 20mins for travel time and 5 to 7 times a week respectively by using
their private vehicles.Therefore,the data samples are a good represen-
tative of the Australian population and some insights for their travel
habits can also be obtained.
The summary of socio-demographics for respondents is shown in
Table 2:
4.2.Measurements of the latent variables
Table 3 illustrates the variables and relevant measurement items.
Of the five variables,one was the outcome variable which measures
the intention to use the AVs.For each variable,there are several mea-
surement items and these items were rated by respondents on a Likert
scale of 1 to 5.One represents “strongly disagree”while five repre-
sents “strongly agree”.
4.3.Method
Confirmatory factor analysis (CFA)was implemented to investigate
and review its suitability of structural equation modelling.As Kline
(2011)recommended,the Comparative Fit Index (CFI), Chi-square
value,Root Mean Square Error Approximation (RMSEA)and Standard-
ised Root Mean Square Residual (SRMR)to the degree of freedom were
treated as goodness-of-fit indices.Convergent validity could assess
whether the survey items within a group are associated with each other
or not.In order to guarantee convergent validity,factor loadings for
each the measured items on the factor should be significant and higher
than 0.7.Moreover,the average variance extracted (AVE)for the factor,
as a way to assess the convergent validity,should be higher than 0.5.
Discriminant validity represents the extent to whether the factors(con-
structs)in the model differ from one another (Hamid et al., 2017). If the
square root of average variance extracted (SAVE)for each factor is
higher than bivariate correlations in the proposed model,the factor is
deemed to have discriminant validity (Fornell and Larcker,1981). In
addition to that,the internal consistency of each measured items in the
factor was checked using composite reliability (CR)as well as Cron-
bach's alpha.Cronbach's alpha as well as composite reliability can be
regarded as good internal consistency if their value are higher than 0.7
(Fornell and Larcker,1981). Structural Equation Modeling (SEM)tech-
nique of Partial Least Squares (PLS)was applied to assess structural
models.SEM was run using SmartPLS3 software.A bootstrap procedure
5
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
Table 2
Descriptive statistics of the survey.
Demographic
variable Value Frequency Proportion
(%)
Gender Male 396 49.0%
Female 408 50.4%
Other 1 0.1%
Prefer not to say 4 0.5%
Age group 18–24 116 14.3%
25–34 173 21.4%
35–44 138 17.1%
45–54 149 18.4%
55–64 115 14.2%
65–74 81 10.0%
75–84 32 4.0%
>=85 5 0.6%
State South Australia 58 7.2%
New South Wales 269 33.3%
Queensland 160 19.8%
Tasmania 18 2.2%
Western Australia 76 9.4%
Northern
Territory 8 1.0%
Victoria 211 26.1%
Australian Capital
Territory 9 1.1%
Educational
background Did not complete
high school or
none
50 6.2%
High School,
Certificates/
Diploma/
Vocational,or
Equivalent
376 46.5%
Undergraduate:
Bachelor's degree,
Certificates/
Diploma,or
Equivalent
297 36.6%
Master's degree 77 9.5%
Doctoral degree 9 1.1%
Income level
(AUD per
month after
tax)
$0 16 2.0%
$1–$3000 282 34.9%
$3001–$5000 281 34.7%
$5001–$7000 135 16.7%
$>7000 95 11.7%
Primary mode
for commuting
(pre-COVID19)
Rail 77 9.5%
Bus 72 8.9%
Drive your own
private car 553 68.4%
Car-pooling 7 0.9%
Tram 10 1.2%
Walk/On foot 31 3.8%
Cycling 5 0.6%
As a passenger in
car 34 4.2%
Riding services
(e.g.Uber)6 0.7%
Motorcycle 2 0.2%
Other 12 1.5%
What is your
travel time for
your trip made
by your
primary
transport
mode?
<10 mins 119 14.7%
10 to 20 mins 256 31.6%
21 to 30 mins 209 25.8%
31 to 40 mins 113 14.0%
41 to 50 mins 47 5.8%
Above 50 mins 65 8.0%
Table 2 (continued)
Demographic
variable Value Frequency Proportion
(%)
How often do
you drive a
private car?
<2 times a week 33 6%
2–4 times a week 126 22.8%
5–7 times a week 199 36%
Above 7 times a
week 193 34.9%
None 2 0.4%
On average,
how long do
you drive the
private car per
weekday?
<20 mins 113 20.4%
20 mins–40 mins 206 37.2%
40 mins–60 mins 108 19.5%
60 mins–80 mins 60 10.8%
>80 mins 66 11.9%
On average,
what is your
commuting
distance
travelled using
a private car
per weekday?
<10 km 152 27.4%
11 km–20 km 191 34.5%
21 km–30 km 105 18.9%
>30 km 105 18.9%
Driving license Yes 761 94.1%
No 48 5.9%
Driving
experience No experience 37 4.6%
<1 year 29 3.6%
1–5 years 107 13.2%
5–10 years 119 14.7%
10–20 years 132 16.3%
>20 years 385 47.6%
Number of cars
owned per
household/
family
0 43 5.3%
1 337 41.7%
2 306 37.8%
>2 123 15.2%
including 1000 sub-samples was implemented to evaluate the signifi-
cance level of the models’theoretical constructs.We used Partial Least
Squares Multigroup Analysis (PLS-MGA), which is a non-parametric
significance test for finding the difference between group-specific re-
sults based on PLS-SEM bootstrap results (Hair Jr.et al., 2021).
5.Results
5.1.Assessment of measurement model
Before implementing the structural equation modelling,we assessed
the adequacy of the measurement model by undertaking the PLS algo-
rithm.As demonstrated in Table 4,all the factor loadings were signifi-
cant and higher than 0.7 except PEU1(0.682)and PEU2(0.672), which
are close to 0.7.Additionally,all averaged variances extracted (AVEs)
were above 0.5,while composite reliability (CR)and Cronbach's Alpha
(Cr.α)of each latent construct exceeded 0.7.Table 4 shows the square
root of AVE was larger than the associations with other constructs,
which demonstrates a strong discriminant validity (Fornell and Larcker,
1981). Table 5 indicated that no indicator loadings are higher than op-
posing constructs (Hair et al., 2011). Table 6 revealed that every item in
the model has higher loadings on its own parent construct compared to
other constructs,proving a strong discriminant validity.Therefore,the
parameters of the measurement model provide strong evidence for the
reliability and validity of construct measures.
5.2.Structural equation modelling
After demonstrating our measurement model's results and ade-
quacy,we tested the hypothesis.The hypotheses in terms of the paths of
the relationship between all variables were examined using SEM,and
6
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
Table 3
Variables and measurement items.
Variables Measurement items Sources
Perceived
trust (PT)1.Automated cars are
dependable (PT1)
2.Automated cars are
reliable (PT2)
3.I can trust automated
cars (PT3)
4.I feel confident that I
could use an automated
car (PT4)
(Venkatesh et
al., 2003)
Perceived
ease of use
(PEU)
1.It is easier to learn how
to operate an
automated car than a
conventional one.
(PEU1)
2.I would find the
technical assistance for
automated cars to be
available from specific
individuals or groups
(PEU2)
3.Learning to use a fully
automated car will be
easy for me (PEU3)
4.I will find it easy to get
a fully automated car
to do what I want it to
do (PEU4)
5.It will be easy for me to
become skilful at using
an automated car
(PEU5)
6.I will find automated
car easy to use (PEU6)
(Nees,2016;
Choi and Ji,
2015;
Venkatesh and
Davis,2000)
Perceived
usefulness
(PU)
1.A fully automated car
will let me do other
tasks,such as eating on
my trip (PU1)
2.Using an automated car
will decrease my
accident risk (PU2)
3.I find an automated car
to be useful when I'm
impaired (e.g.drowsy)
(PU3)
4.I believe that the use of
automated cars can
improve traffic quality.
(PU4)
5.I believe that the use of
automated cars can
improve environmental
quality (PU5)
6.I believe that the use of
automated cars can
improve traffic safety
(PU6)
(Venkatesh et
al., 2003)
1.Table 3 (continued)
Variables Measurement items Sources
Perceived
data
privacy
(PDP)
1.Automated cars would
not collect too much
information about your
data (PDP1)
2.Automated cars would
not share your
information with other
parties without your
permission (PDP2)
3.Automated cars would
have strong security
measures to protect
your personal
information (PDP3)
(Glancy,2012;
Schoettle and
Sivak,2014)
Attitude
(ATU)1.Using automated cars is
a good idea (ATU1)
2.Using automated cars is
a wise idea (ATU2)
3.Using automated cars is
pleasant (ATU3)
(Rejali et al.,
2023)
Behaviour
intention
(BI)
1.I intend to use the
automated car (BI1)
2.When I see automated
cars on the roads,I
would intend to adopt
them (BI2)
3.I plan to use automated
cars in the future for
my workplace trip
(BI3)
4.I intend to use
automated cars in
future if they are
available in the market
(BI4)
5.I would travel in
automated cars with
other passengers as
well (BI5)
(Choi and Ji,
2015;Nees,
2016;
Venkatesh,
2000)
then the results are provided in Fig.2and Table 7.All the paths within
the proposed model were found to be supported.
Eleven out of 11 hypothesised relationships were significant.Hy-
pothesis 1,which relates to the positive effect of perceived usefulness
on attitude toward using AVs was supported (β=0.374,t=9.626,
p<0.01). As predicted,perceived ease of use had a significant positive
effect on perceived usefulness (Hypothesis 3;β=0.618,t=22.066,
p<0.01). Therefore,perceived ease of use had a significant positive
effect on attitude toward using (Hypothesis 2;β=0.146,t=3.968,
p<0.01). Additionally,the perceived trust had a significant positive
effect on perceived ease of use (Hypothesis 5;β=0.681,t=26.153,
p<0.01), which indirectly proves that perceived trust had a signifi-
cant positive effect on attitude toward using (Hypothesis 4;β=0.380,
t=9.290,p<0.01).
Perceived data privacy (concern)had significant positive impacts on
perceived trust,perceived ease of use and perceived usefulness,respec-
tively (Hypothesis 6;β=0.614,t=23.160,p<0.01,Hypothesis 8;
β=0.155,t=5.248,p<0.01,Hypothesis 7;β=0.270,t=8.887,
p<0.01). Therefore,perceived data privacy also had a positive impact
on attitude toward using (Hypothesis 9;β=0.054,t=2.130,
p<0.05). As predicted,perceived usefulness and attitude toward us-
ing both had a positive impact on behavioural intentions (Hypothesis
10;β=0.279,t=7.715,p<0.01,Hypothesis 11;β=0.619,
t=18.053,p<0.01).
7
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
Table 4
Results of the confirmatory factor analysis (Internal consis-
tency,reliability and convergent validity of the measurement
model).
Variable Items Convergent validity Cr.α
Factor
loading CR AVE
Perceived trust PT1
PT2
PT3
PT4
0.899
0.911
0.905
0.734
0.922 0.748 0.885
Perceived ease of
use PEU1
PEU2
PEU3
PEU4
PEU5
PEU6
0.682
0.672
0.816
0.852
0.860
0.873
0.912 0.635 0.882
Perceived
usefulness PU1
PU2
PU3
PU4
PU5
PU6
0.749
0.870
0.778
0.867
0.802
0.884
0.928 0.683 0.906
Perceived data
privacy PDP1
PDP2
PDP3
0.851
0.891
0.875
0.905 0.761 0.844
Attitude ATU1
ATU2
ATU3
0.936
0.938
0.907
0.948 0.859 0.918
Behavioural
Intentions BI1
BI2
BI3
BI4
BI5
0.851
0.875
0.883
0.906
0.828
0.939 0.755 0.919
Note:Cr.α:Cronbach's Alpha;CR:Composite Reliability;AVE:Average Vari-
ance Extracted.
SRMR =0.069 (estimated model), Chi-square =2463.138 (estimated model).
Table 5
Discriminate validity of the measurement model.
Variables ATU BI PDP PEU PT PU
ATU 0.927
BI 0.846 0.869
PDP 0.604 0.641 0.872
PEU 0.761 0.753 0.573 0.797
PT 0.814 0.808 0.614 0.776 0.865
PU 0.813 0.783 0.624 0.772 0.77 0.827
Note:the data on the diagonal is the square roots for AVE,and the data below
are the correlation coefficients.
5.3.Mediation analysis
In order to test if there is a mediation between PT,PEU and PU,PDP,
ATU and BI,we analysed direct and indirect effects by using a bootstrap
method (Alfons et al., 2022). This method allows us to separate each in-
direct effect to identify the mediating effects of perceived usefulness,
perceived ease of use and perceived trust separately instead of in paral-
lel.In this bootstrap study,we used a bootstrap sample size of 1000
with a 95%confidence level.
Table 8 indicates that PT has a significant impact on BI with the to-
tal coefficient of 0.512 and PEU,attitudes as well as PU have full medi-
ation between PT and BI as there is no direct link.Similarly,Table 9
demonstrates that PDP also has a significant impact on BI with the total
coefficient of 0.548 with full mediation from attitude,PEU,PT and PU.
Table 6
Factor loadings (in bold)and cross-loadings.
ATU BI PDP PEU PT PU
ATU1 0.936 0.798 0.55 0.696 0.762 0.765
ATU2 0.938 0.78 0.577 0.707 0.759 0.756
ATU3 0.907 0.775 0.553 0.712 0.742 0.738
BI1 0.672 0.851 0.566 0.642 0.641 0.613
BI2 0.73 0.875 0.599 0.672 0.683 0.682
BI3 0.752 0.883 0.545 0.636 0.748 0.681
BI4 0.794 0.906 0.548 0.671 0.762 0.728
BI5 0.722 0.828 0.528 0.652 0.669 0.69
PDP1 0.492 0.548 0.851 0.462 0.513 0.512
PDP2 0.496 0.52 0.891 0.448 0.491 0.495
PDP3 0.582 0.6 0.875 0.574 0.59 0.612
PEU1 0.553 0.595 0.528 0.682 0.544 0.556
PEU2 0.52 0.553 0.471 0.672 0.526 0.554
PEU3 0.589 0.547 0.386 0.816 0.59 0.592
PEU4 0.704 0.671 0.501 0.852 0.712 0.706
PEU5 0.605 0.591 0.42 0.86 0.637 0.623
PEU6 0.641 0.633 0.439 0.873 0.675 0.643
PT1 0.711 0.69 0.544 0.665 0.899 0.696
PT2 0.734 0.719 0.54 0.662 0.911 0.696
PT3 0.768 0.791 0.59 0.683 0.905 0.722
PT4 0.592 0.582 0.437 0.682 0.734 0.535
PU1 0.586 0.552 0.413 0.646 0.594 0.749
PU2 0.744 0.697 0.564 0.684 0.722 0.87
PU3 0.585 0.553 0.415 0.637 0.555 0.778
PU4 0.713 0.721 0.573 0.633 0.643 0.867
PU5 0.66 0.633 0.5 0.575 0.605 0.802
PU6 0.724 0.703 0.603 0.662 0.686 0.884
As Table 10 shows,PU has a significant impact on BI with the total co-
efficient of 0.511 with partial mediation from attitudes as there is a di-
rect effect between PU and BI (β=0.279).
5.4.Multigroup analysis
The multigroup analysis enables the assessment of whether esti-
mates of group-specific parameters,such as path coefficients,change
significantly across pre-defined data groups.
In terms of gender,the effect of PU->BI was statistically significant
and weaker among women than among men (B=−0.141,P=0.041).
Similarly,the effect of PU on ATU was also statistically significant and
weaker among women than among men (B=−0.202,P=0.014).
However,the effect of PEU on ATU was statistically significant and
stronger among women than among men (B=0.161,P=0.024).
Moreover,the effect of PDU->PU was statistically significant and
weaker among women than among men (B=−0.167,P=0.007).
The effect of ATU->BI was statistically significant and stronger in
the 18–24 age group than in the 75–85 age group (B=0.433,
P=0.013). However,the effects of PU->ATU and PU->BI were sig-
nificant but weaker in comparisons between the same groups,having
values of (B=−0.456,P=0.028)and (B=−0.521,P=0.002), re-
spectively.Furthermore,the effects of PDP->PT and PDP->PU were
significant but weaker when comparing 25–24 to 65–74,with values of
(B=−0.228,P=0.049)and (B=−0.349,P=0.001), respectively.
However,among the comparisons of the same groups,the effect of
PEU->ATU was significant and stronger (B=0.224,P=0.027). In
terms of income level,the impact of the ATU->BI was statistically sig-
nificant and weaker among the groups of income levels of $1-$3000 vs.
$5001-$7000 and $3001-$5000 vs. $5001-$7000,having values of
(B=−0.232,P=0.009)and (B=−0.1852,P=0.035), respec-
tively.
8
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
Fig.2.Results of the structural model indicate relationships, **: p<0.01; *p<0.05.
Table 7
Results of the structural model.
Structural
relations Hypothesis B P-value T-
value Result
PU ->
ATU H1 0.374 <0.01 9.626 H1 is
accepted
PEU ->
ATU H2 0.146 <0.01 3.968 H2 is
accepted
PEU ->
PU H3 0.618 <0.01 22.066 H3 is
accepted
PT ->
ATU H4 0.380 <0.01 9.290 H4 is
accepted
PT ->
PEU H5 0.681 <0.01 26.153 H5 is
accepted
PDP ->
PT H6 0.614 <0.01 23.160 H6 is
accepted
PDP ->
PU H7 0.270 <0.01 8.887 H7 is
accepted
PDP ->
PEU H8 0.155 <0.01 5.248 H8 is
accepted
PDP ->
ATU H9 0.054 <0.05 2.130 H9 is
accepted
PU -> BI H10 0.279 <0.01 7.715 H10 is
accepted
ATU ->
BI H11 0.619 <0.01 18.053 H11 is
accepted
Note:Significance level was p =0.05.
6.Discussion
6.1.Theoretical implications
There is little research investigating public opinion about fully AVs
within the Australian context.In order to achieve this,an online survey
has been distributed to Australian people.This study undertook a com-
prehensive and detailed examination of how public opinions on fully
AVs are formed based on demographic differences and psychological
perception.
This study proposed an extended TAM by including perceived trust
as well as perceived data privacy in the original TAM.The two variables
in the original TAM model –PU and PEU,affect attitudes significantly in
this study,with total effects of 0.374 and 0.377,respectively.This
shows that the proposed model in this study has the ability to model
AVs acceptance behaviour.Although previous study (Acharya and
Mekker,2022)combines BI and attitude into one variable,the results
are quite similar in terms of the total effects of PU and PEU.According
to the mediation analysis,trust mediated the impact of perceived data
privacy on the AVs acceptance in this study.The result also aligns with
the findings from (Acharya and Mekker,2022).
It turns out that attitudes significantly mediate the effect of PU on BI
in the present study,which demonstrates a similar result from (Rejali et
al., 2023)although this result is based on the data from Iran.Similarly,
PU also mediates the effects of PDP and PEU on attitudes;therefore,
Australian people tend to have a good attitude toward AVs because
most Australian people agree with the many potential benefits of the
new technology (Cunningham et al., 2019).
The present study also shows that the predictors (PU and attitude)of
BI significantly impact intentions to use AVs.The result aligns with an-
other research (Xiao and Goulias,2022)which demonstrates perceived
usefulness strongly correlated with the intention to adopt AVs.In this
study,attitude is significantly affected by perceived trust,followed by
PDP,PEU and PU.This finding is in line with (Zmud and Sener,2017),
arguing that trust may be the potential reason instead of age and vehi-
cle ownership.This finding also aligned with (Kenesei et al., 2022)that
user should have enough trust to reduce perceived risk,thus by increas-
ing the future usage.Additionally, (Adnan et al., 2018)believed that
one of the great challenges is to build the trust towards AVs technology.
The importance of trust is also backed by Choi and Ji (2015)that trust is
one of the most important determinants of intention to use AVs.Previ-
ous research (Zhang et al., 2019)revealed that performance trust had
the strongest effect on the intention to use,which indirectly supports
this study's finding:perceived trust significantly affected attitude,and
attitude significantly affected behavioural intention.Therefore,similar
Table 8
Direct and indirect effects of PT on BI.
Total effect (PT->BI)Direct effect (PT->BI)Indirect effects of PT on BI
Coefficient p-value Coefficient p-value Coefficient T value p-value SD
0.512 0.000 N/A N/A H:PT -> PEU -> Attitudes -> BI 0.061 3.817 0.000 0.016
H:PT -> Attitudes -> BI 0.235 7.855 0.000 0.03
H:PT -> PEU -> PU -> Attitudes -> BI 0.098 8.107 0.000 0.012
H:PT -> PEU -> PU -> BI 0.117 6.969 0.000 0.017
9
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
Table 9
Direct and indirect effects of PDP on BI.
Total effect (PDP->
BI)Direct effect (PDP->
BI)Indirect effects of PDP on BI
Coefficient p-value Coefficient p-value Coefficient T
value p-
value SD
0.548 0.000 N/A N/A H:PDP -> Attitudes -> BI 0.033 2.085 0.037 0.016
H:PDP -> PEU -> Attitudes -> BI 0.014 3.036 0.002 0.005
H:PDP -> PT -> Attitudes -> BI 0.144 7.334 0.000 0.02
H:PDP -> PT -> PEU -> PU -> BI 0.072 6.858 0.000 0.011
H:PDP -> PEU -> PU -> BI 0.027 3.873 0.000 0.007
H:PDP -> PU -> Attitudes -> BI 0.063 6.377 0.000 0.01
H:PDP -> PT -> PEU -> PU ->
Attitudes -> BI 0.06 7.958 0.000 0.008
H:PDP -> PU -> BI 0.075 5.684 0.000 0.013
H:PDP -> PT -> PEU -> Attitudes -> BI 0.038 3.796 0.000 0.01
H:PDP -> PEU -> PU -> Attitudes ->
BI 0.022 4.49 0.000 0.005
Table 10
Direct and indirect effects of PU on BI.
Total effect (PU->BI)Direct effect (PU->BI)Indirect effects of PU on BI
Coefficient p-value Coefficient p-value Coefficient T value p-value SD
0.511 0.000 0.279 0.000 H:PU -> Attitudes -> BI 0.232 9.318 0.000 0.025
to the study from (Zhang et al., 2021), trust plays the most important
factor in determining AV acceptance.
Although the present study indicated that PEU and PU were signifi-
cant constructs of attitude (Davis,1989), PDP was more significant than
PEU and PU.Past studies were concerned with informational privacy
perception because it mainly drives the acceptance of AVs (Walter and
Abendroth,2020). This study confirms that perceived data privacy has
significantly influenced attitude and BI,respectively.Another study by
Dirsehan and Can (2020)extending TAM model indicated that there is a
strong significant association between sustainability concerns and trust,
which means people take care about data privacy,trust and then adopt
AVs.However,the current findings is contrary to previous research by
Dimitrakopoulos et al. (2021)that stated data privacy concerns are im-
portant but somehow redundant.Likewise,the research from Slovenia
(Ljubi and Groznik,2023)revealed that privacy is another important
factor but less important than safety concerns.Therefore,our study
shows the importance of understanding the public perceptions of data
privacy and trust and how it relates to their decisions regarding the
adoption of AVs.
Similarly,the empirical result of the multigroup study contributes to
the AV literature.The effect of the PU on BI was also statistically signifi-
cant and greater among men than women,and the effect of PU on ATU
was statistically significant and larger among men than women.This
demonstrates the difference in perception between the different gen-
ders.This highlights the disparity in perception between the genders.
The authors (Rice and Winter,2019)also found that women were less
likely than men to travel in AVs owing to fear.Similarly,the effect of
ATU on BI was statistically significant and higher in those aged 18–24
than in those aged 75–85.This reinforces the notion that older people
may have a different understanding of technology than younger genera-
tions (Lee and Coughlin,2014;Chakraborty et al., 2016).
6.2.Practical implications
This study also underlines the importance of perceived data privacy
regarding preventative design and policy development.By acknowledg-
ing the importance of data privacy,the government should establish ef-
fective legislation by integrating it with current road traffic laws.How-
ever,there are still some discussions to be made,such as what types of
data should be stored and what purpose this data will be used (Khan et
al., 2020). Similar to the study by Zmud et al. (2016),both studies re-
veal that data privacy is an important factor in AVs adoption,with peo-
ple who are more concerned with data privacy will be less likely to use
AVs.Although both studies show the intention to use is higher for some
population segments,age is not a significant factor overall.A study by
Soltani et al. (2021),showed that Australian students are worried about
data privacy.Also,the study found differences in perceptions of AVs be-
tween male and female students,which aligns with the finding from the
current study that there are differences between male and female per-
ceptions regarding AVs.
From the manufacturer's point of view,integrating data security fea-
tures when conducting vehicle design should be prioritised.Some tech-
niques,including transfer layer solutions (e.g.authentication code)and
physical layer security solutions,shall be considered by the manufac-
turers (Rathore et al., 2022). Like what (Lee et al., 2022)had suggested,
in order to gain public acceptance of a new technological system,it is
important to be deployed under a clear condition that societal values
and potential concerns are addressed,especially privacy.Government
should provide policy guidance,such as penalties for privacy violations
and mandatory data reporting of accidents and collision.
Another important factor towards the acceptance of AVs is per-
ceived trust.This study concurred with research conducted by Zhang et
al. (2020),which emphasized the significance of establishing initial
trust.These findings highlight the importance of peer opinions and
first-hand knowledge in augmenting user inclination towards AVs.
Though the intention to use AVs exhibited variation among different
demographic groups,devising different strategies that cater to these di-
verse groups could effectively foster trust in AVs.Although the per-
ceived trust varies in AV with different driving styles (Ekman et al.,
2021), vehicle manufacturers should focus on designing control strate-
gies and human–machine interfaces to achieve desirable levels of per-
ceived trust.This also includes customized design solutions for each in-
dividual user to counter the lack of trust,which increases the adoption
of AVs (Hamburger et al., 2022). Government should concentrate on
building social trust to help the public overcome psychological barriers,
such as promoting benefits realisation.The good news is that most of
our participants hold neutral and somewhat positive opinions towards
AVs.Therefore,combining efforts from the government and manufac-
turers will be the key to facilitating the acceptance of AVs.
10
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
The multigroup analysis (shown in section 5.4)indicated the statis-
tical disparity between gender and age groups,which is of the utmost
importance to the decision-maker.To improve women's perceptions of
AV adoption,for example,a more proactive and normative stance is re-
quired.Therefore,consumer information management and education
regarding the utility of AVs would be extremely advantageous.In addi-
tion,an elderly-tailored opportunity must be provided in order to im-
prove the weaker perceptions of older age groups regarding the adop-
tion of AVs (Chakraborty et al., 2016). Nevertheless,women and people
in older age groups will benefit the most from transportation based on
AVs.
6.3.Study limitations
This study still has its limitations that could be addressed by further
research.Although level 5 automation will not be a reality in a few
years,big technology companies (e.g.Tesla)plan to roll out level 5 au-
tomation in the next few years (Altman,2022). Given that level 5 auto-
mated vehicles have not existed in the market,some people who filled
out the questionnaire may not have a complete understanding of AVs,
although the definition was provided in the questionnaire.This may af-
fect participants’perception of AVs ease of use and usefulness.How-
ever,based on media news on AVs and other various sources,people
may have some idea of AVs and their benefits.Likewise,the study has
investigated the impact of perceived data privacy and trust on the over-
all acceptance of AVs.However,the impact may vary between different
situations,such as owning AV (purchased), shared AV(rented)and the
situation people recommend AV to their peers.Moreover,although
around 90%of our participants have heard of AVs before,future re-
search study may investigate whether people who have prior long-term
experiences in AVs could affect the acceptance of AVs.In future,partici-
pants could be provided with an opportunity for a test ride of AV and
follow it up with a questionnaire survey on their perceptions of AVs.
This may increase the reliability of the model’s constructs.
7.Conclusion
In this study,by extending the original TAM,we have incorporated
the two important factors:data privacy as well as perceived trust in the
TAM to investigate the public perceptions of fully AVs.The model was
assessed using 809 respondents'data from Australia.The research re-
vealed that perceived trust and perceived data privacy is the first and
second most important variable affecting the attitude,followed by per-
ceived ease of use as well as usefulness.The results of structural equa-
tion modelling suggested that attitudes and perceived usefulness di-
rectly impact behavioural intention,and attitudes significantly mediate
perceived usefulness on behavioural intention.This study highlighted
the importance of perceived trust and data privacy,among other psy-
chosocial variables.The findings of this study not only provide in-depth
theoretical constructs but also provide valuable findings to the policy-
makers to develop strategies to promote the adoption and deployment
of fully AVs,especially in Australia.Due to the important role of per-
ceived trust and data privacy,policymakers may consider promoting
AVs through awareness campaigns and media on what measures have
in place to ensure AVs safety and security,including cybersecurity mea-
sures.The implementation of safety and security measures may be
achieved by adopting suitable policies when implementing regulations
for fully AVs.Also,decision-makers may consider promoting the advan-
tages and use of AVs as perceived usefulness as well as ease of use posi-
tively influenced the intention to use AVs.During promotional cam-
paigns for AVs,there is an opportunity to proactively address any pub-
lic concerns regarding the use of AVs and protection against data pri-
vacy or security breach.
Uncited references
CRediT authorship contribution statement
Yilun Chen :Conceptualization,Methodology,Soft-
ware,Formal analysis,Investigation,Data curation,Writing
–original draft,Writing –review &editing.Shah Khalid
Khan :Formal analysis,Writing -original draft.Nirajan Shi-
wakoti :Conceptualization,Writing –review &editing,Su-
pervision,Funding acquisition.Peter Stasinopoulos :Super-
vision,Funding acquisition.Kayvan Aghabayk :Investiga-
tion,Writing –review &editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
References
Acharya,S., Mekker,M., 2022.Public acceptance of connected vehicles:An extension of
the technology acceptance model.Transp.Res.Part F:Psychol.Behav.88,54–68.
Acheampong,R.A., Legacy,C., Kingston,R., Stone,J., 2023.Imagining urban mobility
futures in the era of autonomous vehicles—insights from participatory visioning and
multi-criteria appraisal in the UK and Australia.Transp.Policy 136,193–208.
Adnan,N., Nordin,S.M., Bahruddin,M.A.B., Ali,M., 2018.How trust can drive forward
the user acceptance to the technology? In-vehicle technology for autonomous vehicle.
Transp.Res.A Policy Pract.118,819–836.
Ajzen,I., 1991.The theory of planned behavior.Organ.Behav.Hum.Decis.Process.50
(2), 179–211.
Alfons,A., Ates,N.Y., Groenen,P.J.F., 2022.A ROBUST BOOTSTRAP TEST FOR
MEDIATION ANALYSis.Organ.Res.Methods 25 (3), 591–617.
Altman,A., 2022.Automotive World.[Online]Available at:https://
www.automotiveworld.com/articles/the-race-to-level-5-will-be-won-through-
collaboration/[Accessed 10 December 2022].
Boulange,C., Gunn,L., Giles-Corti,B., Mavoa,S., Pettit,C., Badland,H., 2017.Examining
associations between urban design attributes and transport mode choice for walking,
cycling,public transport and private motor vehicle trips.J.Transp.Health 6,
155–166.
Brauner,P., Hick,A., Philipsen,R., Ziefle,M.2023.What does the public think about
Artificial Intelligence?—A criticality map to understand bias in the public perception
of AI.Front.Comput.Sci., 5,19.
Buckley,L., Kaye,S.-A., Pradhan,A.K., 2018.Psychosocial factors associated with
intended use of automated vehicles:A simulated driving study.Accid.Anal.Prev.
115,202–208.
Carpenter,J., 2020.Kill switch:The evolution of road rage in an increasingly AI car
culture.In:Living With Robots.Academic Press,pp.75–90.
Chakraborty,R., Lee,J., Bagchi-Sen,S., Upadhyaya,S., Raghav Rao,H., 2016.Online
shopping intention in the context of data breach in online retail stores:An
examination of older and younger adults.Decis.Support Syst.83,47–56.
Chen,C.-F., 2019.Factors affecting the decision to use autonomous shuttle services:
Evidence from a scooter-dominant urban context.Transp.Res.F 67,195–204.
Chen,Y., Shiwakoti,N., Stasinopoulos,P., Khalid Khan,S., 2022.State-of-the-art of
factors affecting the adoption of automated vehicles.Sustainability 14 (11), 1–29.
Choi,J.K., Ji,Y.G., 2015.Investigating the importance of trust on adopting an
autonomous vehicle.Int.J.Hum.Comput.Interact.31 (10), 692–702.
Cugurullo,F., Acheampong,R.A., 2023.Fear of AI:an inquiry into the adoption of
autonomous cars in spite of fear,and a theoretical framework for the study of
artificial intelligence technology acceptance.AI Soc.1–16.
Cunningham,M.L., Regan,M.A., Horberry,T., Weeratunga,K., Dixit,V., 2019.Public
opinion about automated vehicles in Australia:Results from a large-scale national
survey.Transp.Res.A 129,1–18.
Davis,F.D., 1989.Perceived usefulness,perceived ease of use,and user acceptance of
information technology.MIS Q.13,319–340.
Dimitrakopoulos,G., Tsakanikas,A., Panagiotopoulos,E., 2021.Autonomous Vehicles
Technologies,Regulations,and Societal Impacts.
Dirsehan,T., Can,C., 2020.Examination of trust and sustainability concerns in
autonomous vehicle adoption.Technol.Soc.63,101361.
Duan,L., Wei,Y., Zhang,J., Xia,Y., 2023.Addressing the urban-scale vehicle assignment
and rebalancing problems in shared autonomous vehicle system while simultaneously
considering immediate,reservation,shareable,and unshareable requests.Comput.
Ind.Eng.177,109025.
Ekman,F., Johansson,M., Karlsson,MariAnne,Strömberg,H., Bligård,L.-O., 2021.Trust
in what? Exploring the interdependency between an automated vehicle’s driving style
and traffic situations.Transport.Res.F:Traffic Psychol.Behav.76,59–71.
Fishbein,M., Ajzen,I., 1977.Belief,Attitude,Intention,and Behavior:An Introduction to
Theory and Research.Volume 10(2).
Fornell,C., Larcker,D.F., 1981.Evaluating structural equation models with unobservable
11
CORRECTED PROOF
Y.Chen et al.Case Studies on Transport Policy xxx (xxxx)101072
variables and measurement error.J.Mark.Res.18 (1), 39.
Giannopoulos,G.A., Munro,J.F., 2019.Case study X -US,EU,China,and Japan—The role
of public and private sector investment in EVs,AVs,and AI innovation.In:The
Accelerating Transport Innovation Revolution.Elsevier Inc Web., pp.333–339.
Glancy,D.J., 2012.Privacy in autonomous vehicles.Santa Clara Law Rev.52,1171.
Hair Jr., J.F.et al., 2021.Partial Least Squares Structural Equation Modeling (PLS-SEM)
Using R.
Hair,J.F., Ringle,C.M., Sarstedt,M., 2011.PLS-SEM:Indeed a silver bullet.J.Mark.
Theory Pract.19 (2), 139–152.
Hamburger,Y.A., Sela,Y., Kaufman,S., Wellingstein,T., Stein,N., Sivan,J., 2022.
Personality and the autonomous vehicle:Overcoming psychological barriers to the
driverless car.Technol.Soc.69,101971.
Hamid,M.R.A., Sami,W., Mohmad Sidek,M.H., 2017.Discriminant validity assessment:
Use of Fornell &Larcker criterion versus HTMT Criterion,p.012163.
Haustein,S., Jensen,A.F., 2018.Factors of electric vehicle adoption:A comparison of
conventional and electric car users based on an extended theory of planned behavior.
Sustain.Transp.12 (7), 484–496.
Hegner,S.M., Beldad,A.D., Brunswick,G.J., 2019.In automatic we trust:investigating the
impact of trust,control,personality characteristics,and extrinsic and intrinsic
motivations on the acceptance of autonomous vehicles.Int.J.Hum.Comput.Interact.
35 (19), 1769–1780.
Hoff,K.A., Bashir,M., 2015.Trust in automation:Integrating empirical evidence on
factors that influence trust.Hum.Factors 57 (3), 407–434.
Jeng,D.-J.-F., Tzeng,G.-H., 2012.Social influence on the use of clinical decision support
systems:revisiting the unified theory of acceptance and use of technology by the
fuzzy DEMATEL technique.Comput.Ind.Eng.62 (3), 819–828.
Karuppiah,V., Ramayah,T., 2023.Modeling hybrid cars adoption using an extended
version of the theory of planned behavior.Transp.Lett.15 (7), 780–792.
Kaur,K., Rampersad,G., 2018.Trust in driverless cars:Investigating key factors
influencing the adoption of driverless cars.Eng.Technol.Manage.48,87–96.
Kaye,S.-A., Somoray,K., Rodwell,D., Lewis,I., 2021.Users’acceptance of private
automated vehicles:A systematic review and meta-analysis.J.Saf.Res.79,352–367.
Kenesei,Z., Ásványi,K., Kökény,L., Jászberényi,M., Miskolczi,M., Gyulavári,T.,
Syahrivar,J., 2022.Trust and perceived risk:How different manifestations affect the
adoption of autonomous vehicles.Transp.Res.A 164,379–393.
Keszey,T., 2020.Behavioural intention to use autonomous vehicles:Systematic review
and empirical extension.Transp.Res.Part C:Emerg.Technol.119,102732.
Khan,S.K., Shiwakoti,N., Stasinopoulos,P., Chen,Y., 2020.Cyber-attacks in the next-
generation cars,mitigation techniques,anticipated readiness and future directions.
Accid.Anal.Prev.148,105837.
Khan,S.K., Shiwakoti,N., Stasinopoulos,P., Warren,M., 2023.A multinational empirical
study of perceived cyber barriers to automated vehicles deployment.Sci.Rep.13 (1),
1842.
Kim,D.J., Ferrin,D.L., Rao,H.R., 2008.A trust-based consumer decision-making model in
electronic commerce:The role of trust,perceived risk,and their antecedents.Decis.
Supp.Syst.44 (2), 544–564.
Kline,R.B., 2011.Principles and Practice of Structural Equation Modeling,Fourth ed.The
Gulford Press,New York.
Korkmaz,H., Fidanoglu,A., Ozcelik,S., Okumus,A., 2022.User acceptance of
autonomous public transport systems:Extended UTAUT2 model.J.Public Transp.24,
100013.
Lee,C., Coughlin,J.F., 2014.PERSPECTIVE:Older adults’adoption of technology:an
integrated approach to identifying determinants and barriers.J.Prod.Innov.Manag.
32 (5), 747–759.
Lee,D., Hess,D.J., Heldeweg,M.A., 2022.Safety and privacy regulations for unmanned
aerial vehicles:A multiple comparative analysis.Technol.Soc.71,102079.
Lin,C.A., Kim,T., 2016.Predicting user response to sponsored advertising on social media
via the technology acceptance model.Comput.Hum.Behav.64,710–718.
Ljubi,K., Groznik,A., 2023.Role played by social factors and privacy concerns in
autonomous vehicle adoption.Transp.Policy 132,1–15.
Meyer-Waarden,L., Cloarec,J., 2022.“Baby,you can drive my car”:Psychological
antecedents that drive consumers’adoption of AI-powered autonomous vehicles.
Technovation 109,102348.
Molnar,L.J., Ryan,L.H., Pradhan,A.K., Eby,D.W., St.Louis,R.M., Zakrajsek,J.S., 2018.
Understanding trust and acceptance of automated vehicles:An exploratory simulator
study of transfer of control between automated and manual driving.Transp.Res.F 58,
319–328.
Nees,M.A., 2016.Acceptance of self-driving cars:an examination of idealized versus
realistic portrayals with a self-driving car acceptance scale.Proceedings of the
Human Factors and Ergonomics Society Annual Meeting 60 (1), 1449–1453.
Panagiotopoulos,I., Dimitrakopoulos,G., 2018.An empirical investigation on consumers’
intentions towards autonomous driving.Transp.Res.Part C:Emerg.Technol.95,
773–784.
Rahimi,A., Azimi,G., Asgari,H., Jin,X., 2020.Adoption and willingness to pay for
autonomous vehicles:Attitudes and latent classes.Transp.Res.Part D:Transp.
Environ.89,102611.
Rathore,R.S., Hewage,C., Kaiwartya,O., Lloret,J., 2022.In-vehicle communication
cyber security:challenges and solutions.Sensor 17,22.
Rejali,S., Aghabayk,K., Esmaeli,S., Shiwakoti,N., 2023.Comparison of technology
acceptance model,theory of planned behavior,and unified theory of acceptance and
use of technology to assess a priori acceptance of fully automated vehicles.Transp.
Res.A 168,103565.
Rice,S., Winter,S.R., 2019.Do gender and age affect willingness to ride in driverless
vehicles:If so,then why? Technol.Soc.58,101145.
Schoettle,B., Sivak,M., 2014.A Survey of Public Opinion About Autonomous and Self-
Driving Vehicles in the US,the UK,and Australia.The University of Michigan
Transportation Research Institute,Ann Arbor,Michigan.
Soltani,A., Ananda,D., Rith,M., 2021.University students’perspectives on autonomous
vehicle adoption:Adelaide case study.Case Stud.Transp.Policy 9 (4), 1956–1964.
Susanto,T.D., Aljoza,M., 2015.Individual acceptance of e-government services in a
developing country:dimensions of perceived usefulness and perceived ease of use and
the importance of trust and social influence.Procedia Comput.Sci.72,622–629.
Synopsys,n.d. [Online]Available at:https://www.synopsys.com/automotive/
autonomous-driving-levels.html#[Accessed Accessed on January 19,2023].
Tan,S.Y., Taeihagh,A., 2021.Adaptive governance of autonomous vehicles:Accelerating
the adoption of disruptive technologies in Singapore.Gov.Inf.Q.38 (2), 101546.
Underwood,S., 2014.Automated,Connected,and Electric Vehicle Systems.Accessed on
January 19th,2023 ed.Institute for Advanced Vehicle Systems,Michigan.
Venkatesh,V., 2000.Determinants of perceived ease of use:integrating control,intrinsic
motivation,and emotion into the technology acceptance model.Inf.Syst.Res.11 (4),
342–365.
Venkatesh,V., Davis,F.D., 2000.A theoretical extension of the technology acceptance
model:four longitudinal field studies.Int.J.Hum.Comput.Interact.31 (10),
186–204.
Venkatesh,V., Morris,M.G., Davis,G.B., Davis,F.D., 2003.User acceptance of
information technology:toward a unified view.MIS Q.27,425–478.
Venkatesh,V., Thong,J.Y., Xu,X., 2012.Consumer acceptance and use of information:
existing the unified theory of acceptance and use of technology.MIS Q.36,157–178.
Walter,J., Abendroth,B., 2020.On the role of informational privacy in connected
vehicles:A privacy-aware acceptance modelling approach for connected vehicular
services.Telematics Inform.49,101361.
Wang,S., Zhao,J., 2019.Risk preference and adoption of autonomous vehicles.Transp.
Res.Part A:Policy Pract.126,215–229.
Xiao,J., Goulias,K.G., 2022.Perceived usefulness and intentions to adopt autonomous
vehicles.Transp.Res.A Policy Pract.161,170–185.
Xu,Z., Zhang,K., Min,H., Wang,Z., Zhao,X., Liu,P., 2018.What drives people to accept
automated vehicles? Findings from a field experiment.Transp.Res.Part C:Emerg.
Technol.95,320–334.
Zhang,T., Tao,D.a., Qu,X., Zhang,X., Lin,R., Zhang,W., 2019.The roles of initial trust
and perceived risk in public’s acceptance of automated vehicles.Transp.Res.C 98,
207–220.
Zhang,T., Tao,D., Qu,X., Zhang,X., Zeng,J., Zhu,H., Zhu,H., 2020.Automated vehicle
acceptance in China:Social influence and initial trust are key determinants.Transp.
Res.Part C:Emerg.Technol.112,220–233.
Zhang,T., Zeng,W., Zhang,Y., Tao,D.a., Li,G., Qu,X., 2021.What drives people to use
automated vehicles? A meta-analytic review.Accid.Anal.Prev.159,106270.
Zhang,Q., Zhang,T., Ma,L., 2023.Human acceptance of autonomous vehicles:Research
status and prospects.Int.J.Ind.Ergon.95,103458.
Zmud,J., Sener,I.N., Wagner,J., 2016.Consumer Acceptance and Travel Behviour Impacts
of Automated Vehicles,Texas A&M Transportation Institute.
Zmud,J.P., Sener,I.N., 2017.Towards an understanding of the
travel behavior impact of autonomous vehicles.Transp.Res.
Procedia 25,2500–2519.
Australian Bureau of Statistics (ABS), 2021 Population.
<https://www.abs.gov.au/statistics/people/population>,
accessed 12 March 2023.
12
