Technological Forecasting & Social Change 126 (2018) 284–291

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

journal homepage: www.elsevier.com/locate/techfore

Analysis of public acceptance of electric vehicles: An empirical study in T

Shanghai
Ning Wang, Linhao Tang⁎, Huizhong Pan
School of Automotive Studies,Tongji University, 4800 Caoan Road, Jiading District,Shanghai 201804, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Developing Electric Vehicles (EVs) has a significant meaning in protecting the environment. However, the low
Electric vehicles public acceptance of EVs hinders the promotion of EVs in China. Factor analysis method and structural equation
Public acceptance model are used in this essay to explore the potential factors that affect consumers' acceptance of EVs in Shanghai
Influencing factors based on the questionnaire survey. The results show that only 18.1% of the respondents are willing to purchase
Factor analysis method
EVs to replace conventional vehicles. Technical level, marketing, perceived risks and environmental awareness
Structural equation model
Promotion solutions
have significant impacts on EV acceptance. Finally, the promotion solutions are put forward accordingly.

1. Introduction deeply study on key factors influencing consumers' acceptance of EVs.

In this paper, we analyze the key influencing factors of EV public
In recent years, China's consumption of oil in the transport sector acceptance based on a questionnaire survey of EV potential consumers
is > 50% of the total oil consumption, which is expected to reach over in Shanghai which has the largest ownership of EVs in China. This
60% in 2020. Besides, with the increase of car ownership, China's paper is divided into 5 sections, the first being this introduction. The
carbon emissions in 2014 have exceeded those of the United States and second section presents the factors affecting consumers' acceptance of
the European Union, indicating that the environmental problem is be- EVs through literature review and expert interviews. The third section
coming increasingly prominent (Anon, 2015.). Under the pressures of provides the research model and methodology. The fourth section
energy security and environmental pollution, China should change the provides the research results based on the questionnaire survey data
existing energy consumption structure to reduce fuel consumption and and data analysis methods. The fifth section presents the research
emissions. The adoption of EVs can reduce China's dependence on conclusions.
foreign oil and effectively reduce the environmental pollution pro-
blems. 2. Literature review
Compared with the traditional gas-powered vehicle, EVs have much
higher purchase cost, lower charging infrastructure availability and According to recent researches, consumers' acceptance of EVs is
longer charging time, which make consumers unwilling to purchase EVs influenced by many factors, which can be divided into four categories:
(Carley et al., 2013a). In order to promote EV's adoption, Chinese EV technical performance, external environmental factors, consumers`
central government and local governments have introduced plenty of demographics and personalities, and consumers ` perceived values to
fiscal and non-fiscal policies. The fiscal policies include purchase sub- EVs.
sidies, purchase tax exemption, charging infrastructure construction
subsidies and electricity price subsidies. The non-fiscal policies mainly (1). The technical performance of EVs is the most important factor.
include road tolls exemptions and free public charging in designated Based on the research of hybrid vehicle owners, Ozaki and
charging places (Li et al., 2016). Stimulated by such policies, the Sevastyanova (2011) concluded that the factors related to the ve-
ownership of EVs approached 500,000 by the end of 2015. However, hicle performance such as high ride quality, low noise, automatic
the market share of EVs is still relatively small. Furthermore, EV sold transmission and easy operation will affect consumers' acceptance
badly in the private domain where Chinese government prefers to of EVs. (Jensen et al., 2013) made a survey of 369 Danish battery
seeing a great success of EV. In the private domain, which is the most electric vehicle owners and found that the EV models, driving
important is consumers` preference to EV. Therefore, it is necessary to range and security will affect consumers' willingness to buy EVs.

⁎
Corresponding author.
E-mail addresses: wangning@tongji.edu.cn (N. Wang), tanglinhao228@outlook.com (L. Tang), huizhong_pan@hotmail.com (H. Pan).

http://dx.doi.org/10.1016/j.techfore.2017.09.011
Received 7 June 2017; Received in revised form 29 August 2017; Accepted 10 September 2017
Available online 27 October 2017
0040-1625/ © 2017 Elsevier Inc. All rights reserved.
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Egbue and Long (2012) found that the immature battery tech- showed that the factors affecting consumers' acceptance of EVs are not
nology and the high battery price are the main obstacles to the only the product performance, but also consumers' perceptions of EV
large-scale promotion of EVs. driving pleasure and symbolic meaning.
(2). Besides product technical attributes, consumers' preferences to al- This research gap may be related to the methods used. Consumers'
ternative vehicles also hinge on their demographics and personal- willingness to buy EVs is commonly regarded as a discrete choice
ities. Ziegler, 2012 investigated 598 potential car consumers in problem using the multinomial logit model, nested logit model and
Germany and the results showed that young consumers with en- mixed logit model. Bolduc et al., 2008 and Sikes et al., 2010 applied the
vironmental awareness are more willing to buy EVs. Carley et al. multinomial logit model to study the influencing factors of consumers'
(2013a, 2013b) investigated the American drivers of their will- acceptance of EVs, while Potoglou and Kanaroglou, 2007 applied the
ingness to purchase plug-in hybrid electric vehicles, and found that nested logit model to study the EV public acceptance. However, in the
consumers who are interested in EVs are well educated, and hybrid discrete choice problem, there exists rational inattention where some
electric vehicle owners are eco-conscious and concerned about respondents give up processing information about the unknown values
dependence on imported oil. The research of Hidrue et al. (2011) of the available options to maximize the expected value of the chosen
showed that consumers' educational level, income and environ- option due to its high cost (Mckay and Matejka, 2011).
mental protection awareness are positively correlated with the In order to found a comprehensive research model for studying the
purchase intention of EVs. Tian (2012) concluded that consumers EV public acceptance, this paper applies the structural equation models
who have the intention to purchase EVs are tech-conscious and which allows for the simultaneous analysis of whole system of regres-
environment-friendly. Plötz et al. (2014) made a survey of German sion equations. Furthermore, Structural Equation Modeling is able to
EV consumers and found that the most likely group of EV buyers deal with latent (non-observable) variables which are not directly but of
are middle-aged men with technical professions living in rural or utmost importance for the research (The Structural Equation Model,
suburban multi-person households. 2007). Besides, this model can make it easy to study the relationship
(3). In the early stage of EV's industrialization and commercialization, between each latent variable and the manifest variables (Grewal et al.,
government plays a crucial role to make EV's diffusion self-sus- 2004).
taining. Lane and Potter (2007) found that the environmental
regulations, oil price policy, purchase subsidies and the charging
infrastructure construction will affect the market penetration of 3. Research model and research method
cleaner vehicles. The studies of Eppstein et al. (2011), Beresteanu
and Li, 2011, and Gallagher and Muehlegger, 2011 all indicated Based on the classification from above literature review, the re-
that the government financial subsidies play an important role in search model of EV public acceptance is shown in Fig. 1. Here we as-
the large-scale EV promotion. sume that these four kinds of factors all have significant impacts on
consumers' acceptance of EVs. These assumptions are verified in the
In addition, charging infrastructure determines EV's convenience following sections.
and affects consumers' willingness to purchase EVs in result. Sierzchula The questionnaire was designed based on the research model, which
et al. (2014) studied the EV adoption in 30 countries in 2012 and consisted of four parts. Part I covers 43 questions about the factors
concluded that the charging infrastructure construction is a key factor affecting the consumers' acceptance of EVs, except those factors related
affecting the EV adoption. to consumers. All the factors are measured by multiple items on a 5-
The relative advantage of EV's usage cost depends on fuel price and point “Likert” scale that ranges from 1 = Strongly Disagree to
electricity price which encourages consumers to buy EVs. (Plötz et al., 5 = Strongly Agree. Part II gathers some basic information about travel
2012) analyzed the impacts of the purchase cost and usage cost on the behavior including private car ownership, average daily travel distance
EV adoption based on the total life cycle cost. The research of Choi and and the expectations to EVs such as driving range, maximum speed and
Inha (2010) concluded that the main factors affecting consumers' ac- price. Part III measures consumer's acceptance of EVs. Part IV is related
ceptance of EVs are the purchase price, the usage cost and the main- to the demographic variables, including sex, age and educational level,
tenance cost. The studies of Zubaryeva et al. (2012) and Dijk et al. etc.
(2013) also confirmed the influence of oil price and electricity price on The survey was done in the 4S stores and EV test drive centers of
the EV purchase intention. Shanghai whose respondents all have a good understanding about EVs
and are the potential EV consumers. Finally, 458 valid questionnaires
(4). Finally, the car is not only a means of transportation, but also a were collected. To analyze consumers' acceptance of EVs with the col-
symbol of identity. Kurani (Heffner et al., 2007) found that the lected data, Spss21.0 and Amos21.0 software were applied as the sta-
symbolic meanings of hybrid electric vehicles will affect the pur- tistical data analysis tools to establish a structural equation model. The
chase willingness of the 25 households in the United States. The main statistical analysis methods were as follows:
respondents hold the opinion that driving a hybrid electric vehicle
will express their personalities such as maturity, intelligence and
distinction. Raham-Rowe et al. (2012) (surveyed 40 UK households
of their driving experience of EVs and found that some drivers felt
good because of the EV environmental benefits, while some drivers
felt embarrassed because of the poor performance and appearance
of EVs.

Although there have been many researches about this subject, most
of them only incorporate one or several influencing factors without
building up a comprehensive model. And the results seem unilateral
and distorted according to other researches. For example, through the
investigation of the Swedish car owners, Jansson, 2011 found that the
consumers' willingness to purchase EVs was affected by consumers'
subjective norms and attitudes, the innovative personality and per-
Fig. 1. Research model of EV public acceptance.
ceived EV performance. The research of Schuitema et al., 2013 also

285
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

(1). Descriptive statistical analysis: Frequency analysis is used to con- are not willing to purchase EVs is the highest. The percentage of
duct various statistical analysis for the demographic variables and respondents within the age of 26 to 30 who are willing to purchase
the research variables. EVs as a second car is the highest. The percentage of respondents
(2). Factor analysis: a statistical analysis method that uses a few factors within the age of 41 to 50 who are willing to purchase EVs to re-
to reflects the majority of the information of the original elements. place the traditional fuel vehicles is relatively higher.
To determine whether the data is suitable for factor analysis, the 3). Education level: respondents whose education level is junior middle
Bartlett spherical test and KMO (Kaiser-Meyer-Olkin) test should school or lower have the highest percentage to choose not to buy
be performed before the factor analysis. EVs. Respondents with bachelor degrees or above have the highest
(3). Structural equation model: a popular method in social science re- percentage to choose to buy EVs as a second car.
search, which integrates the methods of factor analysis, path ana- 4). Annual household income: respondents with annual income
lysis and multiple linear regression analysis. In this paper, the SEM of < 50,000 yuan or between 200,000 and 300,000 yuan have the
model include the measurement model and structural model and is highest percentage to choose not to buy EVs. Respondents with
applied to analyze the relationship among the observing variables, annual income of > 500,000 yuan have the highest percentage to
the latent variables and consumers' acceptance of EVs. choose to buy EVs as a second car. Respondents with annual income
between 300,000 and 500,000 yuan have the highest percentage to
4. Research result choose to buy EVs to replace traditional fuel vehicles.

4.1. Sample descriptive statistics 4.1.3. Consumers' travel features and preferences
The study has also investigated respondents` vehicle ownership,
4.1.1. Demographics travel behaviors and expectations to EVs. In terms of private car own-
Appendix 1 illustrates the demographic attributes of the re- ership, 67.9% of the respondents own one private car at least. Daily
spondents in this study. 52.9% of the respondents are male and the travel distance of most respondents (81.6%) is below 20 km, which is
majority of the respondents are within the age group of 18–40 (81.1%). much lower than the expected driving range of EVs. 43.3% of the re-
This group could be the potential consumers for electric vehicles. In spondents expect that the driving range of EVs is above 200 km. The
terms of education level, most respondents have acquired high aca- expected maximum speed is above 80 Km/h or above 100 Km/h.
demic degrees. The proportion of respondents who have bachelor de- Consumers are very sensitive to price. Half of the respondents could not
grees is 67.9%. When it comes to household income, 70.0% of the re- accept that EVs costs more than traditional fuel vehicles. 35.0% of the
spondents are within the income group of 50,000–200,000 yuan. In respondents could accept that the price of EVs is 10% higher than that
2015, the average annual income per urban resident in China is 31,195 of traditional fuel vehicles. Only 4.2% could accept that the EV price is
yuan, which is lower than the income of the majority of respondents. In 30% or above higher than traditional fuel vehicles.
terms of occupation, > 20% of the respondents are the ordinary staff
and technical staff. The party and government cadres / teachers / po- 4.1.4. Descriptive statistics of the scale variables
licemen, company managers and students are in the proportion be- The descriptive statistics results of 43 scale variables in Part I are
tween 10% and 20%. shown in Appendix 2, including the minimum value, maximum value,
mean value, standard deviation, skewness and kurtosis. It can be seen
4.1.2. Consumers' willingness to purchase EVs from the values of skewness and kurtosis that the variables do not obey
The investigation shows that > 60% of the respondents express the normal distribution. Therefore, they are standardized for further
their willingness to purchase electric vehicles. However, the re- analysis and the results are showed as follows:
spondents are more willing to buy EVs as the second car. The propor-
tion of respondents who are willing to buy electric vehicles as a second 1). EV technology performance: it is believed by respondents that EVs
car (46.1%) is much higher than that of respondents willing to buy have advantages in low noise, easiness to handle without gearbox,
electric vehicles to replace traditional fuel vehicles (18.1%). The results excellent digital entertainment and navigation functions and fash-
are shown in Table 1. It is concluded that the public acceptance for EVs ionable interior and appearance.
is very low and there are few consumers willing to buy EVs to replace 2). EV adoption context: The main advantage of EVs caused by the
their traditional fuel vehicles. price difference between electricity and fuel is recognized by re-
In this paper, cross contingency tables are used to analyze the cor- spondents. In terms of EV policies, the respondents think that the
relations between the demographic variables and consumers` will- subsidies for purchasing EVs are relatively generous while the
ingness to purchase EVs. The results of the cross analysis are shown as electricity subsidies and non-fiscal subsidies are not enough. As for
below: EV charging services, the respondents think that the EV public
charging time is long and current charging availability are in-
1). Gender: the proportion of men who are not willing to buy EVs is adequate. In addition, though the majority of respondents know
higher than that of women. Women are more willing to buy EVs as a about some well-known EV manufacturers and models, automotive
second car. companies need to put more efforts on advertising and promotion.
2). Age: the percentage of respondents within the age of 18 to 25 who 3). EV perceived social values: in that EVs make great contributions to
energy saving and environmental protection, driving EVs is re-
Table 1
garded by respondents as an identity sign of environmentalist.
Consumers' acceptance of EVs. Nonetheless, the respondents also approve that EVs take risks in
being charged in time, battery life short and the technological level.
Item Choice N Percentage

Y: Are you willing to Unwilling 164 35.8%

4.2. Reliability and validity analysis
purchase an EV? Willing Willing to purchase an EV 211 46.1%
as a second car Reliability Analysis is a method to measure whether a comprehen-
Willing to purchase an EV 83 18.1% sive evaluation system has certain stability and reliability. In this paper,
to replace traditional fuel
the reliability is assessed by measuring the Cronbach's α coefficient to
vehicles
Sample 458 100.0% check the internal consistency among the items. The acceptable value of
Cronbach's α in reliability analysis is above 0.70 (Thompson, 2004)

286
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Table 2 the real situation and recognized by experts in the industry.

Results of the reliability analysis.
4.4. Structural equation model of EV public acceptance
Element Cronbach's α Item

EV technical performance 0.757 9 Based on the above results of the factor analysis, a structural
EV adoption context 0.800 14 equations model is established using AMOS 21.0 software. The max-
EV value perception and communication 0.780 9
imum likelihood estimation method is utilized for the confirmatory
Consumers' characteristics 0.729 11
Entire sample 0.898 43 factor analysis. The dependent variable Y has three choices, which are:
not willing to purchase EVs (Y = 0), willing to purchase EVs to replace
traditional fuel vehicles (Y = 1) and willing to purchase EVs as a
Table 3 second car (Y = 2). The latent variables are the above 11 factors ob-
Results of validity analysis. tained from the factor analysis. The structural equations causal model is
shown in Fig.2. In this study, the following are used as the criteria: χ2/
Element KMO Bartlett's test p-Value
df < 3; RMSEA < = 0.05; TLI ~ 1; CFI > =0.95; SRMR < = 0.10
EV technical performance 0.775 835.142 0.000 (Dion, 2008). The indicators of the model fitting are as follows: χ2/
EV adoption context 0.783 1803.613 0.000 df = 2.85; RMSEA = 0.048; TLI = 0.938; CFI = 0.963;
EV value perception and communication 0.727 1265.777 0.000 SRMR = 0.056, meeting the above criteria.
Consumers' characteristics 0.743 953.955 0.000
According to the parameter estimation result shown in Appendix 3,
the 4 factors —technical level, marketing, perceived risks and en-
(Grewal et al., 2004). The reliability analysis result is shown in Table 2. vironmental awareness— have significant impact on the EV public ac-
The Cronbach's α for the overall scale of each factor is within ceptance. EV technical level and environmental are positively corre-
0.729–0.898, suggesting a very strong consistency among the items for lated with EV public acceptance, while marketing and perceived risks
each kind of factors. are negatively correlated with EV public acceptance.
In this study, the Kaiser-Meyer-Olkin (KMO) statistics that can test
the partial correlations between variables and Bartlett's spherical hy- (1). Technical level: battery life and driving range are the 2 factors that
pothesis test are used for the validity analysis. Generally, if the value of consumers are concerned about and the impact of driving range
KMO is closer to 1, it would be more adequate for factor analysis. variable on the technical level factor is greater than the battery life
Bartlett's spherical test is used to test whether each variable is in- variable. Low cruise range of EVs causes anxiety when driving.
dependent. The results are shown in Table 3. The values of KMO sta- Short battery life could increase expense in the total lifetime of
tistics are > 0.7, indicating that the data is suitable for factor analysis. EVs. Therefore, in order to improve the technical level of EVs,
The F value of the Bartlett's spherical test is equal to 0.000, which automotive manufacturers should invest in the R & D of battery
means that the data has good construct validity and is suitable for the technology.
following factor analysis. (2). EV marketing factor: the after-sales service, advertising, pur-
chasing channels and brand effect will influence the effectiveness
of marketing whose sequence is according to the influence degree.
4.3. Factor analysis
Therefore, in order to enhance the marketing effect, the after-sales
service quality should be improved and the attractive advertising
The SPSS 21.0 software is used for the factor analysis. One of the
campaign should be carried out. In addition, the purchase channels
purposes is to integrate the original variables into fewer factors and
should be broadened to make the purchase process more con-
greatly reduce the analysis workload. The other purpose is to decrease
venient and multiple brands and more types of EVs should be
the multicollinearity among the original four variables and to lay the
provided for consumers to choose.
foundation for the following regression model. The factorial loads are
(3). Perceived risks factor: according to the order of relevance, im-
determined using the principal component analysis. The Varimax ro-
proving EV technical level and quality reliability is the most im-
tation is utilized for the interaction process and the criteria is that the
portant for reducing consumers' perceived risks of EVs. Besides, the
eigenvalue should be > 1 and the factorial loads must present values
charging infrastructure should be well equipped to meet the needs
above 0.5 (Hair et al., 2009). The factor analysis results are shown in
of EV consumers.
Table 4. Some variables are gradually removed if the values of factorial
(4). Environmental awareness factor: with the development of our so-
loads are < 0.5. The 31 remaining variables are grouped into 11 factors
ciety, consumers are more and more concerned about the moral
explaining 63.96% of the data variance. The matrix reveals a con-
responsibility for environment protection. The companies should
vergence after 8 iterations. Then the 11 factors are named according to
increase publicity about EV's environmental benefits and make
more consumers recognize it.
Table 4
Results of factor analysis.
5. Conclusions
Initial elements Factor Items
In this paper, we analyzed the public acceptance of EVs and its key
EV technical performance Technical level Q5, Q6
Inner performance Q2, Q3, Q4 influencing factors to provide decision support for the government and
External performance Q8,Q9 car manufacturers to develop relevant policies and marketing strate-
EV adoption context Cost of ownership Q11, Q13, Q14, gies. Based on the literature review and summary analysis, the research
Q15 model of EV public acceptance is determined. The questionnaire was
Charging services Q18, Q19
Marketing Q20, Q21, Q22,
obtained from 458 potential consumers of EVs in Shanghai. The re-
Q23 search results were obtained with the data analysis methods of factor
EV value perception and Perceived values Q25, Q26 analysis and structural equation model. Finally, the feasible suggestions
communication Perceived risks Q27, Q28, Q29 were put forward according to the research findings.
Consumers' characteristics Innovative personality Q33, Q34, Q37
The findings of this research indicate that the respondents are more
Conformist mentality Q38, Q39, Q40
Environmental awareness Q41, Q42, Q43 willing to purchase EVs as a second car and the proportion of re-
spondents who are willing to purchase EVs to replace the regular gas-

287
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Fig. 2. The structural equation model of EV public acceptance.

powered vehicles is relatively low. In the current situation, the public government and car manufacturers should put more efforts on R & D,
acceptance of EVs is low. In the aspect of demographic variables, marketing and charging infrastructure to improve EV's technical level
consumers who have the characteristics of men, age of 18 to 25 years and decreasing EV's risks.
old, educational level of junior middle school or lower, household in- Although this research has come up with some findings of the
come below 50,000 yuan and between 200,000 and 300,000 yuan are consumers' purchase willingness, there are still some limitations. For
more unwilling to purchase EVs. Consumers who have the character- example, the research object is limited to the consumers of private cars
istics of female, age of 26 to 30 years old, educational level of bachelor with no consideration of the commercial fleets. In addition, when
degree or above, annual household income of > 500,000 yuan are analyzing the public acceptance of private electric vehicles, we didn't
more willing to buy EVs as the second car. Consumers who have the differentiate the battery electric vehicles and plug-in hybrid electric
characteristics of 41 to 50 years old, household income between vehicles. These should be further studied in the future work.
300,000 and 500,000 yuan are more likely to buy EVs to replace the
conventional vehicles. At the early stage of EV industry development, Acknowledgements
the government and car manufacturers had better regard people who
have the same characteristics as the target group to purchase EVs and This study was supported by grants from the Ministry of Science and
develop the special strategies for EV promotion. Technology of the People’s Republic of China (Grant no.
Based on the result of the structural equation model, the 4 factors of 2015BAG11B00), the Shanghai Soft Science Research Project (Grant no.
technical level, marketing, perceived risks and environmental aware- 16692103700), and the Fundamental Research Funds for the Central
ness have significant impacts on the EV public acceptance. The Universities (Grant no. 1700219155).

Appendix 1. Demographics of the respondents

Sample characteristics Percentage

Sample size 458

Gender Male 52.9%

Female 47.1%
Age 18–25 21.3%
26–30 29.7%
31–40 30.1%
41–50 10.5%
51–60 7.5%
> 61 0.9%
Marital Status Single 41.0%
Married 59.0%

288
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Education Level Junior middle school or lower 0.4%

Senior middle school or equivalent 6.6%
Associate 12.1%
Bachelor 67.9%
Master 11.9%
Ph.D. 1.1%
Occupation Party and government cadres/Teacher/Policeman 11.9%
Ordinary staff 27.7%
Business owners/Shareholders 1.3%
Technical staff 21.9%
Worker/Service personnel 7.5%
Company managers 11.1%
Freelancers 6.2%
Retirees 1.8%
Students 10.6%
Annual Income (yuan) < 50,000 8.5%
50,000–80,000 17.6%
80,000–120,000 22.5%
120,000–150,000 15.4%
150,000–200,000 14.5%
200,000–300,000 6.9%
300,000–500,000 4.0%
> 500,000 4.0%
No stable income 6.5%

Appendix 2. Descriptive statistics of scale variables (N = 458)

Variable Minimum Maximum Mean Std. Skewness Kurtosis

deviation

Q1 The acceleration time of EVs is short and the power is strong 1 5 2.838 1.085 − 0.233 − 0.611
Q2 EVs are easy to handle without gearbox 1 5 3.336 1.154 − 0.537 − 0.329
Q3 EVs are safe 1 5 3.007 1.133 − 0.212 − 0.646
Q4 The noise of EVs is small 1 5 3.744 1.021 − 0.701 − 0.087
Q5 The battery life of EVs is long 1 5 2.969 1.059 − 0.271 − 0.500
Q6 The driving range of EVs is long 1 5 2.810 1.085 − 0.019 − 0.513
Q7 There are enough EV models for consumers to choose 1 5 2.612 1.205 − 0.092 − 1.008
Q8 The interior and appearance of EVs are fashionable and cool 1 5 3.215 1.053 − 0.359 − 0.227
Q9 EVs have excellent digital entertainment and navigation functions 1 5 3.286 0.985 − 0.556 − 0.317
Q10 The EV purchase price of 100,000 to 300,000 yuan is reasonable 1 5 2.963 1.190 − 0.234 − 0.834
Q11 The electricity cost of EVs is lower than fuel cost 1 5 3.673 1.106 − 0.689 − 0.100
Q12 The maintenance cost of EVs is low 1 5 3.373 1.139 − 0.463 − 0.430
Q13 The government subsidies for EV purchase are large 1 5 3.369 1.074 − 0.477 − 0.178
Q14 The government subsidies for electricity cost are provided 1 5 3.306 1.018 − 0.440 0.009
Q15 The government develops policies such as license plate exemption and 1 5 3.247 1.210 − 0.416 − 0.646
parking free.
Q16 The EV charging time at home is not long (6 h) 1 5 3.073 1.081 − 0.398 − 0.377
Q17 The EV charging time in public places is rather long (0.5-2 h) 1 5 3.244 1.030 − 0.429 0.004
Q18 The EV charging infrastructure has been well equipped 1 5 2.274 1.130 0.297 − 1.014
Q19 The EV charging is very convenient 1 5 2.331 1.187 0.374 − 0.924
Q20 Some well-known car manufacturers have launched EVs 1 5 3.195 1.129 − 0.344 − 0.415
Q21 The EV purchase process is easy and fast 1 5 2.982 1.066 − 0.350 − 0.384
Q22 The EV after-sales service is guaranteed 1 5 2.996 0.989 − 0.279 0.000
Q23 The EV advertising and promotion are attractive 1 5 2.854 1.071 − 0.033 − 0.492
Q24 EVs have made great contributions to energy-saving and environmental 1 5 3.952 1.113 − 1.026 0.476
protection
Q25 Driving EVs is the sign of status as an environmentalist 1 5 3.523 1.147 − 0.630 − 0.174
Q26 EVs make me more fashionable and can reflect my personality 1 5 3.189 1.133 − 0.312 − 0.463
Q27 I worry about the short battery life of EVs 1 5 3.930 0.994 − 0.775 0.221
Q28 I worry about the low technological level and unreliable quality of EVs 1 5 3.736 1.066 − 0.667 0.006
Q29 I′m worried that the EVs could not be charged in time 1 5 3.998 1.018 − 0.930 0.405
Q30 EVs have values for commercialization 1 5 3.689 0.985 − 0.603 0.285
Q31 The EVs will be popular in the near future 1 5 3.466 0.997 − 0.381 − 0.076
Q32 I will recommend EVs to my family, friends or colleagues 1 5 3.175 1.123 − 0.294 − 0.481

289
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Q33 I like to pursuit things that are fashionable and new 1 5 3.312 1.076 − 0.295 − 0.307
Q34 I am more likely to accept new technologies than others 1 5 3.462 0.986 − 0.423 0.036
Q35 If something is broken, I will replace it rather than repair it 1 5 2.796 1.199 − 0.070 − 0.929
Q36 Compared to the price, I am more concerned about the quality 1 5 3.763 1.047 − 0.751 0.138
Q37 I like to own attractive things 1 5 3.119 1.066 − 0.381 − 0.173
Q38 When my choices are different from others, I would hesitate 1 5 3.195 1.038 − 0.444 − 0.183
Q39 I do not like mavericks 1 5 3.066 1.082 − 0.268 − 0.323
Q40 When going shopping, I often listen to and take the advice of my friends 1 5 3.475 0.998 − 0.529 0.076
Q41 I always keep an environment- friendly lifestyle in my daily life 1 5 3.545 0.982 − 0.496 0.170
Q42 Using energy saving products will be respectful 1 5 3.658 1.064 − 0.627 − 0.017
Q43 I am willing to pay extra money for environmental protection 1 5 3.298 1.064 − 0.475 − 0.065

Appendix 3. Parameter estimates from the structural equation model

Explained variable Predictor variable Estimate Std. err Z-value P (> |z|)

EV public acceptance Technical level 0.216 0.087 2.477 0.013

Inner performance 0.024 0.101 0.238 0.812
External performance −0.009 0.080 − 0.109 0.913
Cost of ownership 0.119 0.105 1.136 0.256
Charging services −0.025 0.066 − 0.380 0.704
Marketing −0.287 0.136 − 2.111 0.035
Perceived values 0.073 0.079 0.923 0.356
Perceived risks −0.156 0.061 − 2.540 0.011
Innovative personality 0.102 0.070 1.445 0.148
Conformist mentality −0.004 0.118 − 0.037 0.971
Environmental awareness 0.174 0.081 2.142 0.032
Technical level Q5 1.000
Q6 1.030 0.414 2.489 0.013
Inner performance Q4 1.000
Q3 0.953 0.145 6.555 0.000
Q2 1.150 0.185 6.217 0.000
External performance Q9 1.000
Q8 0.054 0.483 0.112 0.911
Cost of ownership Q11 1.000
Q13 1.895 0.201 9.421 0.000
Q14 1.696 0.180 9.438 0.000
Q15 1.212 0.148 8.174 0.000
Charging services Q18 1.000
Q19 1.401 3.797 0.369 0.712
Marketing Q20 1.000
Q21 1.595 0.235 6.784 0.000
Q22 1.854 0.267 6.948 0.000
Q23 1.659 0.242 6.852 0.000
Perceived values Q26 1.000
Q25 0.312 0.293 1.064 0.287
Perceived risks Q27 1.000
Q28 0.887 0.072 12.232 0.000
Q29 0.801 0.068 11.781 0.000
Innovative personality Q33 1.000
Q34 0.925 0.126 7.321 0.000
Q37 0.599 0.086 6.990 0.000
Conformist mentality Q40 1.000
Q39 1.153 0.204 5.645 0.000
Q38 1.171 0.209 5.603 0.000
Environmental awareness Q41 1.000
Q42 1.115 0.109 10.265 0.000
Q43 1.116 0.109 10.264 0.000

References Beresteanu, Arie, Li, Shanjun, 2011. Gasoline price, government support and the demand
for hybrid vehicles in the United States. Int. Econ. Rev. 52 (1), 161–182.
Bolduc, Denis, Boucher, Nathalie, Alvarez-Daziano, Ricardo, 2008. Hybrid choice mod-
Anon, 2015. Anon “Global Carbon Project” of the International Environmentalist Group. eling of new technologies for car choice in Canada. Trans. Res. Rec. J. Trans. Res.

290
N. Wang et al. Technological Forecasting & Social Change 126 (2018) 284–291

Board 2082 (1), 63–71. Germany. Transp. Res. A Policy Pract. 67, 96–109.
Carley, S., Krause, R.M., Lane, B.W., et al., 2013a. Intent to purchase a plug-in electric Potoglou, Dimitris, Kanaroglou, Pavlos S., 2007. Household demand and willingness to
vehicle: a survey of early impressions in large US cites[J]. Transp. Res. Part D: pay for clean vehicles. Transp. Res. Part D: Transp. Environ. 12 (4), 264–274.
Transp. Environ. 18, 39–45. Raham-Rowe, Ella, et al., 2012. Mainstream consumers driving plug-in battery-electric
Carley, Sanya, et al., 2013b. Intent to purchase a plug-in electric vehicle: a survey of early and plug-in hybrid electric cars: a qualitative analysis of responses and evaluations.
impressions in large US cites. Transp. Res. Part D: Transp. Environ. 18, 39–45. Transp. Res. A Policy Pract. 46 (1), 140–153.
Choi, Hyundo, Inha, Oh., 2010. Analysis of product efficiency of hybrid vehicles and Schuitema, G., Anable, J., Skippon, S., et al., 2013. The role of instrumental, hedonic and
promotion policies. Energ Policy 38 (5), 2262–2271. symbolic attributes in the intention to adopt electric vehicles[J]. Transp. Res. A
Dijk, Marc, Orsato, Renato J., Kemp, Rene, 2013. The emergence of an electric mobility Policy Pract. 48, 39–49.
trajectory. Energ Policy 52, 135–145. Sierzchula, W., Bakker, S., Maat, K., et al., 2014. The influence of financial incentives and
Dion, Paul A., 2008. Interpreting structural equation modeling results: a reply to Martin other socio-economic factors on electric vehicle adoption[J]. Energ Policy 68,
and Cullen. J. Bus. Ethics 83, 365–368. 183–194.
Egbue, Ona, Long, Suzanna, 2012. Barriers to widespread adoption of electric vehicles: An Sikes, Karen, et al., 2010. Plug-in Hybrid Electric Vehicle Market Introduction Study:
analysis of consumer attitudes and perceptions. Energ Policy 48, 717–729. Final Report. No. ORNL/TM-2009/019. Oak Ridge National Laboratory (ORNL).
Eppstein, Margaret J., et al., 2011. An agent-based model to study market penetration of The Structural Equation Model, 2007. The successful use of information in multinational
plug-in hybrid electric vehicles. Energ Policy 39 (6), 3789–3802. companies. DUV 132–135.
Gallagher, Kelly Sims, Muehlegger, Erich, 2011. Giving green to get green? Incentives and Thompson, Bruce, 2004. Exploratory and Confirmatory Factor Analysis: Understanding
consumer adoption of hybrid vehicle technology. J. Environ. Econ. Manag. 61 (1), Concepts and Applications. American Psychological Association.
1–15. Tian, Xu., 2012. The Identification of Potential Consumers on Domestic Electric Vehicles
Grewal, R., Cote, J.A., Baumgartner, H., 2004. Multicollinearity and measurement error and the Research of Consumer Behavior [D]. Wuhan University of Technology.
in structural equation models: implications for theory testing[J]. Mark. Sci. 23 (4), Ziegler, Andreas, 2012. Individual characteristics and stated preferences for alternative
519–529. energy sources and propulsion technologies in vehicles: a discrete choice analysis for
Hair, J., Anderson, R., Black, W., 2009. Multivariate Data Analysis, 7th ed. Prentice Hall, Germany. Transp. Res. A Policy Pract. 46 (8), 1372–1385.
Upper Saddle River, USA. Zubaryeva, Alyona, et al., 2012. Assessing factors for the identification of potential lead
Heffner, Reid R., Kurani, Kenneth S., Turrentine, Thomas S., 2007. Symbolism in markets for electrified vehicles in Europe: expert opinion elicitation. Technol.
California's early market for hybrid electric vehicles. Transp. Res. Part D: Transp. Forecast. Soc. Chang. 79 (9), 1622–1637.
Environ. 12 (6), 396–413.
Hidrue, Michael K., et al., 2011. Willingness to pay for electric vehicles and their attri- Wang Ning was born in 1977 in Shang Dong Province and is a Masters and PhD su-
butes. Resour. Energy Econ. 33 (3), 686–705. pervisor in the School of Automotive Studies, Tongji University. Wang Ning is Director of
Jansson, J., 2011. Consumer eco-innovation adoption: assessing attitudinal factors and the Automotive Market and Behavior Data Institute. He gained a PhD degree from the
perceived product characteristics. Bus. Strateg. Environ. 20 (3), 192–210. Department of Economics and Management, Tongji University in 2006 and a Master's
Jensen, Fjendbo, Anders, Cherchi, Elisabetta, Mabit, Stefan Lindhard, 2013. On the sta- degree in the Department of Transportation in 2003. He is mainly engaged in the study of
bility of preferences and attitudes before and after experiencing an electric vehicle. automotive marketing, consumer behavior, business models of electric vehicles and
Transp. Res. Part D: Transp. Environ. 25, 24–32. charging facilities, industrial policy, the internet of vehicles, car sharing, automotive E-
Lane, Ben, Potter, Stephen, 2007. The adoption of cleaner vehicles in the UK: exploring commerce, and big data analysis and applications, among others.
the consumer attitude–action gap. J. Clean. Prod. 15 (11), 1085–1092.
Li, W., Long, R., Chen, H., 2016. Consumers' evaluation of national new energy vehicle
policy in china: an analysis based on a four paradigm model. Energ Policy 99, 33–41. Tang Linhao was born in 1995 in Henan Province. He gained a Bachelor's degree in 2017
Mckay, A., Matejka, F., 2011. Rational Inattention to Discrete Choices: A New Foundation and he is a postgraduate in the School of Automotive Studies, Tongji University. His
for the Multinomial Logit Model. Meeting Papers. Society for Economic Dynamics. research interests are in the following fields: assessments of policy incentives for pro-
Ozaki, Ritsuko, Sevastyanova, Katerina, 2011. Going hybrid: an analysis of consumer moting electric vehicles, business models of electric vehicles, among others.
purchase motivations. Energ Policy 39 (5), 2217–2227.
Plötz, Patrick, Gnann, Till, Wietschel, Martin, 2012. Total ownership cost projection for Pan Huizhong was born in 1994 in Anhui Province. She gained a Bachelor's degree in
the German electric vehicle market with implications for its future power and elec- Vehicle Engineering at Beijing Institute of Technology in 2015 and is a postgraduate in
tricity demand. In: 7th Conference on Energy Economics and Technology the School of Automotive Studies, Tongji University. Her research interests are in the
Infrastructure for the Energy Transformation. 27. following fields: the internet of vehicles, business models of electric vehicles, and the
Plötz, Patrick, et al., 2014. Who will buy electric vehicles? Identifying early adopters in development mode of solar energy charging stations, among others.

291