50 Int. J. Mobile Communications, Vol. 7, No.

1, 2009

The empirical study of automotive telematics

acceptance in Taiwan: comparing three
Technology Acceptance Models

Huei-Huang Chen*
Department of Information Management,
Tatung University,
40 Chungshan North Road, Section 3,
Taipei 104, Taiwan, ROC
E-mail: hhchen@ttu.edu.tw
*Corresponding author

Shih-Chih Chen
Department of Computer Science
and Engineering,
Tatung University,
40 Chungshan North Road, Section 3,
Taipei 104, Taiwan, ROC
E-mail: onizuka.dr@gmail.com

Abstract: Automotive telematics comprises the applications of Global

Positioning System (GPS) navigation, multimedia entertainment, wireless
communications and automatic driving assistance systems. This study
examined users’ acceptance of automotive telematics. Following the theory
comparison approach, we evaluated including the Technology Acceptance
Model (TAM), the Theory of Planned Behaviour (TPB) and Combined
Technology Acceptance Model (TAM)-TPB model (C-TAM-TPB), that could
explain users’ automotive telematics acceptance decisions. The respective
models were evaluated using web-based survey data collected from 345 users
about their perceptions of automotive telematics. Overall, the results revealed
that the effect of Perceived Ease Of Use, Attitude and Perceived Behavioural
Control were very important but that usefulness and Subjective Norm did not
influence an individual’s Behavioural Intention. The implications of this study
were also discussed.

Keywords: automotive telematics; Combined Technology Acceptance Model;

CTAM; Combined TAM-TPB; C-TAM-TPB; mobile communications;
Technology Acceptance Model; TAM; telematics; Theory of Planned
Behaviour; TPB.

Reference to this paper should be made as follows: Chen, H-H. and Chen, S-C.
(2009) ‘The empirical study of automotive telematics acceptance in Taiwan:
comparing three Technology Acceptance Models’, Int. J. Mobile
Communications, Vol. 7, No. 1, pp.50–65.

Copyright © 2009 Inderscience Enterprises Ltd.

 The empirical study of automotive telematics acceptance in Taiwan 51

Biographical notes: Huei-Huang Chen received his PhD in Computer Science

from the University of Illinois at Urbana-Champaign in 1984. He is the Dean of
and Professor at the Tatung University’s Management and Design College,
Taiwan. He was with the information systems division of Tatung Co. for
14 years. His research interests include data warehousing and business
intelligence, information technology applications for e-businesses and
m-commerce.

Shih-Chih Chen is a Doctoral Candidate at the Tatung University in Taiwan.

His current research interests include quantitative analysis, mobile commerce
and managerial issues of emerging technologies.

1 Introduction

It was a luxurious facility only seen in few top cars to control the operation of all systems
in the vehicle through computing before 1980s. However, the relationship between the
automobile design, production and the technology has been closed. Since 1990s, an
automobile had been changed to an intelligent digital vehicle. Information services
provided to such a vehicle over wireless communication networks are referred to as
automotive telematics. There were two reasons causing such trend. First, due to the
automobile manufacturers began greatly relying on the computing for the mechanical
structure design of car body. The other was, due to the demand of car drivers and
passengers for in-car mobile communications and satellite navigation services were
steadily increasing.
Telematics, a union of telecommunications and informatics for the transport sector, is
expected to revolutionise automobile industries. With more than 40 million vehicles sold
worldwide each year and more than 935 million cellular customers by the end of 2001,
the automotive telematics market is poised for explosive growth (Zhao, 2002).
Automotive telematics and navigation – including both hardware and services – will
experience strong growth in several regions, generating in total revenues of 38.3 billion
dollars in 2011(ABI Research, 2006a,b). Strategy Analytics (2001) estimated that by
2007, approximately 55% of all new cars will have a telematics-capable terminal, as
compared to approximately 7.5% in 2000 (Telematics Research Group, 2002). For
modern people, their cars are almost the place where they take longest time to stay in
addition to their homes and offices. Therefore, it seems that one-way information
receiving cannot satisfy the users. What the user requests is total information control.
When the car becomes a black hole for information transmission, the longer the time is
for the user to stay in the car, the longer the time is for him/her being unable to access
the information. Telematics integrates such functions as internet, communication,
satellite positioning, safety monitoring and entertainment to provide the service of
in-car information transmission and exchange to satisfy the demand for in-car mobile
computing. One of the most popular cited examples is OnStar, a car navigation
system that bases its geographical information on a satellite-based infrastructure
(OnStar, 2006; Junglas, 2007).
52 H-H. Chen and S-C. Chen

Safety and convenience is and will always be the core for vehicle design, ubiquitous
context information affecting safety and convenience is hidden in three entities: vehicle,
people and environment (Zhang, Wang and Hackbarth, 2004). Most telematics research
has focused on the technology developments and functional applications essential to its
success (Melvin, 1999; Park et al., 2003; Tsou, 2004; Zhang, Wang and Hackbarth, 2004;
Tanizaki, Maruyama and Shimada, 2005). On the other hand, lot of recent research has
applied Technology Acceptance Model (TAM) to understand mobile commerce related
use (Chau and Hu, 2002; Yang, 2005; Han et al., 2006; Hung, Hwang and Hsieh, 2007;
Junglas, 2007; Lee and Jun, 2007; Lu, Wang and Yu, 2007; Chen, 2008), but few for
automotive telematics. This study examined and compared automotive telematics system
acceptance to the TAM (Davis, 1989; Davis, Bagozzi and Warshaw, 1989) the Theory of
Planned Behaviour (TPB; Ajzen, 1991) and the integrated model (Taylor and Todd,
1995a). Using the responses from the web-based survey that involved 345 users, this
research outcome evaluated and compared the extent to which the respective models
could explain individual user’s intention to use automotive telematics technology. The
causal paths specified by each model were also examined.

2 Telematics and its technology adoption

2.1 Telematics and automotive telematics

Information Technology (IT) has revolutionised automobiles in many ways. Today,
vehicles incorporate a high level of computer technology to meet the needs of consumers
for safety, environmental protection, communications and even entertainment needs.
In fact, electronic equipment now controls more than 86% of all systems in a typical
vehicle (Alliance of Automobile Manufacturers, 2006). The Alliance of Automobile
Manufacturers defines it as below:
“integrating the automobile electronics with the communication technology to
provide the driver driving guidance and information exchange services
to improve personal driving safety and security.”
The automotive industry quickly adopted the term to describe any system that provides
location-based services for a vehicle over the wireless telecommunications network.
In other words, generally, telematics refers to any automotive system that combines
wireless technology with location-based services. One of the most important information
services in automotive telematics is the car navigation service, which is a service that
calculates a route and guides the user to follow the route (Zhao, 2002; Park et al., 2003).
An automotive telematics platform consists of five key components (Aloi, Alsliety
and Akos, 2007):
1 positioning device
2 communication device
3 entertainment device
4 vehicle interface
5 user interface.
 The empirical study of automotive telematics acceptance in Taiwan 53

Besides, such as in-vehicle displays for drivers, digital maps with dynamic route planning
and navigation, are the popular applications for automotive telematics (Finn and Breen,
1996). It connects the On Board Unit (OBU), the carrier and the service content provider
with the purpose of providing in-vehicle communication and information services. The
method for service provision is to ensure the position of the car through the global
positioning system (GPS), to use OBU to collect driving information and accept the user
demand and to satisfy the demand of users for information transmission and exchange
through the link between the wireless communication network and the service content.
Lots of automotive telematics industries in development are focused on either increasing
driver safety or giving the driver more pleasurable options, instead of taking the driver’s
attention on the road.

2.2 Theoretical background

The development of TPB is originally based on the Theory of Reasoned Action (TRA)
(Fishbein & Ajzen, 1975; Ajzen & Fishbein, 1980). TPB hypothesised that intention to
perform a behaviour is based on: attitude, Subjective Norms (SN) and Perceived
Behavioural Control (PBC) (Ajzen, 1991). Attitude (ATT) towards performing the
behaviour is defined as a person’s general feeling that performing that behaviour is a
favourable or unfavourable action. SN are perception of whether people who are
important to the person think they should or should not perform an activity; this assumes
that the more an individual perceives that others think he or she should engage in a
behaviour, the more likely it is that the person will do so. PBC is assumed to reflect past
experience as well as anticipated obstacles. The more resources and opportunities that
individuals think they possess and the fewer obstacles they anticipate, the greater their
perceived control over the behaviour. Behavioural Intention (BI) is the individual’s
subjective probability that he or she will engage in that behaviour. The immediate
determinant of behaviour is the individual’s intention to perform or not perform that
behaviour. BI depends on three major factors: attitude towards performing the behaviour,
SN and PBC. The relative weights of these three components are expected to vary with
the kind of behaviour being predicted and with the conditions under which the behaviour
is to be performed.
TAM was initiated by Davis (1989) based on TRA, whose purpose was to generally
explain the decisive factor of users for the IT acceptance, to test and verify it by the
theory and to interpret most technology use behaviours. It is considered that belief will
influence attitude toward use, which will further influence BI to use. TAM theorises two
critical beliefs determining a user’s adoption intention and actual usage of IT. The first
belief is Perceived Usefulness (PU), reflecting a person’s salient belief in using the
technology, will be helpful in improving performance. The second belief is
Perceived-Ease-Of-Use (PEOU), explaining a person’s salient beliefs in using the
technology, will be free of any effort. And there is obviously positive relation between
the BI to use and the system usage. However, the difference from TRA is the exclusion of
SN in the model study. In light of the acceptance of users for the information systems,
TAM brings up two special beliefs, that is, PU and PEOU.
54 H-H. Chen and S-C. Chen

Figure 1 Technology Acceptance Model, Theory of Planned Behaviour

and Combined TAM-TPB model

Source: Davis (1989), Ajzen (1991), Taylor and Todd (1995a).

Taylor and Todd (1995b) believed that the ability of TAM to anticipate the user’s BI to
use new technology and the actual use had been supported by a great deal of empirical
researches but the model has not included the other two factors (social and control factor)
which have been proved by many studies to have the remarkable ability influencing the
user’s actual use in using new technology (as shown in Figure 1;Taylor and Todd, 1995a).
Combined TAM and TPB (C-TAM-TPB) is a hybrid model which combines the
constructs of TPB and TAM. Experience was incorporated into this model in between
experienced and inexperienced users. PU, ATT toward behaviour and PBC were all more
salient with increasing experience while SN became less salient with increasing
experience (Venkatesh et al., 2003).
Furthermore, the explanation for user behaviours to accept new technologies is
always a hot issue of the study on the information systems introduction, whose related
researches are one of the most mature research fields in modern information management
literatures. For this issue, many theoretical models have been developed from information
management to sociology and psychology since 1980, including TRA (Fishbein and
Ajzen, 1975; Ajzen and Fishbein, 1980), TPB (Ajzen, 1991) and TAM (Davis, 1989), etc.
In views that the theory of user acceptance behaviour for new technology equipments is
highly valued and continually developed, it is thus clear for the importance of the study
on technology acceptance behaviour to the successful introduction of information
systems. But the variables and the causal relations in various theoretical models are
different. With different theoretical bases and cases, each theoretical model has different
influence variables and causal relations. Therefore, for the application of technology
acceptance behaviour theory, it still needs more empirical analyses to test and verify the
application of theoretical models and their influence variables to different industries.
 The empirical study of automotive telematics acceptance in Taiwan 55

TAM and TPB have emerged as two dominant models and together provide a
theoretical foundation upon which C-TAM-TPB was developed and examined
(Taylor and Todd, 1995a; Chau and Hu, 2002). With inclusion of SN and PBC in TAM,
they proposed to combine the TPB and the TAM. Such two influence factors are also key
variables in TPB. Hence, the cognitive influences specified by TAM may serve as
important precedents of attitudinal beliefs in TPB, which reciprocally may enhance the
explanatory power of TAM via its potential for adding dimensions essential to individual
technology acceptance. Based on the empirical result of (Taylor and Todd, 1995a; Chau
and Hu, 2002) it is found that C-TAM-TPB obtained from the combination of TAM and
TPB has great Goodness-of-Fit for explaining the user’s behaviour in using new
technology.

2.3 Research framework and hypothesis development

Based on these theories and prior research, our basic hypotheses were:
For the result of related studies, it shows that the more positive the user’s attitude
toward the system, the higher the user’s intention to use such system (Davis, 1989; Davis,
Bagozzi and Warshaw, 1989; Taylor and Todd, 1995a). This research deduces the first
hypothesis. H1: Attitude positively influences BI.
Previous research provides evidence of the significant effect of PU on usage intention
(Davis, 1989; Davis, Bagozzi and Warshaw, 1989; Jackson, Chow and Leitch, 1997;
Venkatesh, 2000; Venkatesh, Morris and Phillip, 2000; Lee and Jun, 2007). For the result
of related studies, it shows that the more the user believes his/her work performance will
be enhanced after using new system, the more he/she will be willing to use such system.
This research deduces the second hypothesis. H2: PU positively influences BI.
Extensive research has consistently discussed that there is a positive relationship
between PU and PEOU with acceptance of IT (Agarwal and Prasad, 1999; Dishaw and
Strong, 1999; Venkatesh, Morris and Phillip, 2000). For the result of related studies, it
shows that the attitude toward adopting the system will be more positive when the user
perceives the system is easier to learn and it is more unnecessary to make efforts to use
such system (Davis, 1989; Davis, Bagozzi and Warshaw, 1989; Taylor and Todd, 1995a).
This research deduces the third hypothesis. H3: PEOU positively influences Attitude.
The result of related studies shows that the attitude toward adopting the system will
be more positive when the user believes it will be helpful to enhance the work
performance if adopting new IT (Davis, 1989; Taylor and Todd, 1995a; Wu and Chen,
2005). This research deduces the fourth hypothesis. H4: PU positively influences
Attitude.
The related study shows that the user will more believe it is helpful to enhance the
work performance after adopting the system when he/she perceives such system is easier
to learn. This research deduces the fifth hypothesis. H5: Perceived Ease Of Use positively
influences PU.
For the result of related studies, it shows that the user will be more willing to use new
system when he/she perceives that the important person concerned thinks it shall be more
to use such system (Ajzen, 1991; Taylor and Todd, 1995a; Wu and Chen, 2005). This
research deduces the sixth hypothesis. H6: Subject Norms positively influence BI.
56 H-H. Chen and S-C. Chen

The related study shows that the user’s BI to use the system will be stronger when
he/she believes he/she has the ability to use such system or his/her relevant resources are
more (Ajzen, 1991; Taylor and Todd, 1995a; Wu and Chen, 2005). This research deduces
the seventh hypothesis. H7: PBC positively influences BI.

3 Research methods

3.1 Measures
To ensure the questionnaire satisfies the content validity, the constructs in the study are
modified from relevant prior research, and three command steps are employed to choose
items for measurement. Firstly, the items from the existing literature were translated into
Chinese. Secondly, two university professors and two programmers of GPS navigation,
who are proficient in English and familiar with IT, were asked to provide assistance in
examining the appropriateness of the Chinese version of the scale, translated from the
original English measurement items. Any inappropriate items were eliminated. Finally, a
reexamination for the measurements was repeated throughout the pilot test. In addition, to
ensure desired balance and randomness in the questionnaire, some of the items were
worded with proper negation and all items were randomly sequenced on the questionnaire
in order to reduce the potential ceiling (or floor) effect that induces monotonous
responses to the items designed to measure a particular construct.

3.2 Subjects
A survey agency conducted a web-based survey to evaluate the research model for one
month. The participants for this study were 386 undergraduate students. After eliminating
insincere responses through data filtering, we selected 345 usable responses as the sample
finally. The variables and measurement scales are shown in Table 1. We used the value
of Cronbach’s D for identifying the reliability of the questionnaires, and conducted factor
analysis for convergent validity.
The result of our pilot test showed the high reliability of all the questionnaires. All the
questionnaires used in this survey have been validated in previous studies. Each item of
the questionnaire was measured on a seven-point Likert scale with the end points of
‘strongly disagree (1)’ and ‘strongly agree (7)’. Because of the homogenous nature of our
subjects, we did not check for non-response bias.
After data collection, a two-step procedure proposed by Anderson and Gerbing
(1988) is applied during the Structural Equation Model (SEM) test. The first step
involves developing an effective measurement model with confirmatory factor analysis,
while the second step analyses the structural model. Both SPSS 14 and AMOS 6.0 are
adopted as the tools for analysing the data. Model estimation was executed through use of
the generalised least squares technique.
 The empirical study of automotive telematics acceptance in Taiwan 57

Table 1 Question items used in the study

Construct Item Measure Source

Behavioural BI1 Whenever possible, I intend not to Davis (1989), Taylor and Todd
Intention (BI) use automotive telematics while (1995a), Venkatesh and Davis
driving (1996), Pavlou (2003) and Wu
and Chen (2005)
BI2 I intend to use automotive telematics
while driving as often as needed
BI3 To the extent possible, I would
use automotive telematics while
driving
Attitude (ATT) ATT1 Using automotive telematics is Ajzen and Fishbein (1980),
a wise idea Taylor and Todd, (1995a) and
Wu and Chen (2005)
ATT2 Using automotive telematics
would be pleasant while
driving
ATT3 I dislike the idea of using
automotive telematics
Perceived Ease PEOU1 My interaction with automotive Davis, (1989), Taylor and
Of Use (PEOU) telematics while driving is clear Todd, (1995a), Pavlou, (2003)
and understandable and Wu and Chen (2005)
PEOU2 It will be difficult to learn how
to use automotive telematics
PEOU3 It will be easy to operate automotive
telematics while driving
Perceived PU1 I would find automotive telematics Davis, (1989), Taylor and
Usefulness (PU) not useful while driving Todd, (1995a), Pavlou (2003)
and Wu and Chen (2005)
PU2 Using automotive telematics will
enhance my effectiveness while
driving
PU3 The advantages of using
automotive telematics will
outweigh the disadvantages
Subjective SN1 People who influence my behaviour Ajzen and Fishbein (1980),
Norms (SN) would think that I should use Taylor and Todd (1995a) and
automotive telematics Wu and Chen (2005)
SN2 People who are important to me
would think that I should use
automotive telematics
Perceived PBC1 I have the resources, knowledge, Ajzen and Fishbein (1980),
Behavioural and ability to use automotive Taylor and Todd (1995a) and
Control (PBC) telematics Wu and Chen (2005)
PBC2 Using automotive telematics is
entirely within my control
58 H-H. Chen and S-C. Chen

4 Data analysis results

4.1 Reliability and validity of research constructs

Before the structural model analysis, this study used Cronbach’s D to test the
measurement scale reliability of three models’ components. Reliability can reflect the
internal consistency of the indicators measuring a given factor. The item reliability for
each scale was examined using Cronbach’s D to confirm internal consistency of the
measures. Nunnally (1978) suggested that a scale can be considered to have high
reliability if Cronbach’s D is greater than 0.70 and should be removed if it was lower than
0.35. Results showed that every component of the three clusters’ models had strong
reliability with all Cronbach’s D greater than 0.73 (shown in Table 2).
Convergent validity is achieved if different indicators used to measure the same
construct obtain strongly correlated scores. In SEM, convergent validity can be assessed
by reviewing the t-tests for the factor loadings (Hatcher, 1994). For all three models in
this study, all factor loadings for indicators measuring the same construct are statistically
significant (as shown in Table 2), showing that all indicators effectively measure their
corresponding construct and support convergent validity (Anderson and Gerbing, 1988).
Hence, this supports convergent validity. In addition, using groups with telematics
experience, rather than those without telematics experience, helps facilitate improved
external validity of the current study.

Table 2 Summary of measurement scales

Variance Cronbach’s D
Construct Indicators Mean SD extracted (reliability)
Behavioural BI1 6.07 0.66 0.58 0.81
intention (BI) BI2 5.63 0.87 – –
BI3 5.72 0.87 – –
Attitude (ATT) ATT1 5.73 0.80 0.68 0.87
ATT2 5.76 0.81 – –
ATT3 5.75 0.81 – –
Perceived Ease PEOU1 5.56 0.86 0.52 0.77
Of Use (PEOU) PEOU2 5.72 0.72 – –
PEOU3 5.64 0.8 – –
Perceived PU1 6.08 0.71 0.62 0.83
Usefulness (PU) PU2 6.13 0.68 – –
PU3 6.06 0.7 – –
Subjective SN1 5.1 1.14 0.60 0.73
Norms (SN) SN2 4.95 1.24 – –
Perceived PBC1 5.76 0.98 0.64 0.78
Behavioural
PBC2 5.96 0.9 – –
Control (PBC)
 The empirical study of automotive telematics acceptance in Taiwan 59

Table 3 Chi-square difference tests for examining discriminant validity

TAM (unconstrained) TPB (unconstrained) C-TAM-TPB (unconstrained)

F2 = 73.44 (d.f. = 48) F2 = 64.39 (d.f. = 29) F2 = 145.41(d.f. = 90)
Constrained F2 Constrained F2 Constrained F2
Construct pair (d.f. = 49) F2 difference (d.f. = 30) F2 difference (d.f. = 91) F2 difference
(BI, ATT) 450.56* 377.12* 500.48* 436.09* 595.26* 449.85*
(BI, PEOU) 760.88* 687.44* – – 842.57* 697.16*
(BI, PU) 1087.53* 1014.09* – – 1116.72* 971.31*
(BI, SN) – – 459.23* 394.84* 679.06* 533.65*
(BI, PBC) – – 583.98* 519.59* 700.22* 554.81*
(ATT, PEOU) 423.64* 350.20* – – 516.16* 370.75*
(ATT, PU) 766.85* 693.41* – – 759.28* 613.87*
(ATT, SN) – – 319.34* 254.95* 600.53* 455.12*
(ATT, PBC) – – 428.15* 363.76* 471.44* 326.03*
(PEOU, PU) 763.87* 690.43* – – 807.61* 662.20*
(PEOU, SN) – – – – 551.64* 406.23*
(PEOU, PBC) – – – – 450.91* 305.50*
(PU, SN) – – – – 664.74* 519.33*
(PU, PBC) – – – – 750.43* 605.02*
(SN, PBC) – – 357.77* 293.38* 496.79* 351.38*
Note: BI, Behavioural Intention; ATT, Attitude; PEOU, Perceived Ease Of Use; PU,
Perceived Usefulness; SN, Subjective Norms; PBC, Perceived Behavioural
Control.
*Significant at the 0.001 overall significance level by using the Bonferroni method.
The chi-squared (F2) difference test can be used to assess the discriminant validity
(Hatcher, 1994). Discriminant validity is demonstrated if the F2 difference (with 1
degrees-of-freedom (d.f.)) is significant, meaning that the model in which the two
constructs are viewed as distinct (but correlated) factors is superior. By using the
Bonferroni method under the overall 0.01 level, the critical value of the F2 test is
(1,0.01/15) = 11.58. Since the F2 difference statistics for every two constructs all exceed
11.58 for each model (as shown Table 3), discriminant validity is successfully achieved.

4.2 Empirical results

Every construct in the final measurement models is measured using at least two indicator
variables. Seven common model-fit measures were used to assess the model’s overall
Goodness-of-Fit: chi-squared/ d.f.; Goodness-of-Fit Index (GFI); Adjusted Goodness-of-
Fit Index (AGFI); Normalised Fit Index (NFI); Comparative Fit Index (CFI); Root Mean
square Residual (RMR) and Root Mean Square Error of Approximation (RMSEA). As
shown in Table 4, Comparison of all fit indices with their corresponding recommended
values provided evidence of a good model fit (F2/ d.f. smaller than 3.0, GFI, AGFI, CFI,
NFI all greater than 0.9 and RMR, RMSEA smaller than 0.08), thus demonstrating that
the measurement model exhibited a fairly good fit with the data collected (Anderson and
Gerbing, 1988; Hinkin, 1995; Hu and Bentler, 1995). However, no statistical differences
were found in any included model Goodness-of-Fit measures, suggesting that all the
investigated models were of comparable fit.
60 H-H. Chen and S-C. Chen

Based on the entire sample of TAM, TPB and C-TAM-TPB, two paths are not
supported (Hypotheses H2 and H6 are not supported), while the remaining paths are all
significant at the 0.01 level (H1, H3, H4, H5 and H7 are supported). Properties of the
causal paths, including standardised path coefficients and explanation of variance for
each equation in the hypothesised model are presented in Figures 2–4. The summary of
hypothesis testing results is shown in Table 5.
The explanatory power of each model was also examined. Table 6 summarises the
relative strength of each path specified by the respective models. The strength of an
individual path was evaluated using the standardised path coefficient. The total effect of
one factor on another is obtained by summing up its direct and indirect effects via
relevant intervening factors. Bollen (1989) strongly advocates the importance not only of
the direct effects, but also of the indirect and the total effects when interpreting results in
a SEM.
Table 4 Overall fits of models

Recommended
Fit index value TPB TAM C-TAM-TPB
F2/ d.f. ” 3.0 2.07 1.59 1.86
GFI • 0.90 0.97 0.96 0.94
AGFI • 0.80 0.94 0.94 0.91
NFI • 0.90 0.96 0.96 0.93
CFI • 0.90 0.97 0.98 0.97
RMR ” 0.08 0.03 0.02 0.04
RMSEA ” 0.08 0.06 0.04 0.05

Figure 2 Results of model 1: Technology Acceptance Model

Figure 3 Results of model 2: Theory of Planned Behaviour

 The empirical study of automotive telematics acceptance in Taiwan 61

Figure 4 Results of model 3: Combined Technology Acceptance Model-Theory of Planned

Behaviour

*p < 0.05; **p < 0.01 and ***p < 0.001.

Table 5 Summary of hypothesis testing result

Hypothesis Results
H1: Attitude positively influences Behavioural Intention (ATTĺBI) Supported
H2: Perceived Usefulness positively influences Behavioural Intention (PUĺBI) Not supported
H3: Perceived Ease Of Use positively influences Attitude (PEOUĺATT) Supported
H4: Perceived Usefulness positively influences Attitude (PUĺATT) Supported
H5: Perceived Ease Of Use Positively influences Perceived Usefulness Supported
(PEOUĺPU)
H6: Subject Norms positively influence Behavioural Intention (SNĺBI) Not supported
H7: Perceived Behavioural Control positively influences Behavioural Intention Supported
(PBCĺBI)

Table 6 Strengths of individual factors

Effect on Behavioural Intention TPB TAM C-TAM-TPB

Direct effect
PU – 0.04 0.04
ATT 0.94 0.94 0.87
SN 0.10 – 0.025
PBC 0.22 – 0.13
Indirect effect
PU – 0.35 0.34
PEOU – 0.64 0.62
Total effect
PU – 0.39 0.38
PEOU – 0.64 0.62
ATT 0.94 0.94 0.87
SN 0.10 – 0.025
PBC 0.22 – 0.13
62 H-H. Chen and S-C. Chen

5 Discussions

Several implications can be obtained from the study results. Attitude may be the single
most important driver in individual automotive telematics system acceptance decision-
making. From the three models of this study, it can verify that the attitude toward use is
exactly one of important factors in the adoption of users. Therefore, as promoting
telematics service, it shall be provided with various product packing styles to satisfy the
fondness of users for in-car decorations and for the visual effect of product shapes. In
addition, the service content of Taiwan is mainly navigation, safety and security services.
In future, we can increase diversified entertainment services to enhance the fondness of
users or passengers for the system service.
In this study, PU appeared to have no significant effects on BI. However, based on the
path analysis result of TAM and C-TAM-TPB, we can find that the influence of the PU
on the attitude toward use is larger than the PU directly on the BI. This result is different
from other researches, but the same with Jackson, Chow and Leitch (1997). However, if
we expect users could continue using the system service, a useful service may be a key
point for the positive enhancement of attitude toward use.
PEOU positively influences attitude, rather than PU. We can suggest that the user
interface is simple to learn, yet incredibly powerful and efficient in use, is what users
need. With the automotive telematics system, drivers should establish a direct connection
to the dealership through a touch, even a simple voice command. In another word,
automotive telematics system should be intuitive and easy to the user’s needs. Enjoyable
and hassle-free interactive design is the key element to attracting customers.
SN appeared to have no significant effects on BI. Users are likely to develop
independent evaluations and consequently may place less weight on others’ opinions in
this study.
PBC appeared to have significant influence on BI but not to an extent comparable to
attitude. A plausible explanation for the significant but modest effect is that the
operations of telematics technology may not be particularly complicated, especially the
considerations of safety, system developers of have to take account that users cannot
distract attentions from driving.

6 Conclusions and future work

Creating applications likely to be accepted by target users is critical to harnessing a new

technology’s potential. This study investigated factors essential to acceptance of
telematics technology by individual users in Taiwan. We took the comparison of TAM
related approaches and used BI as a measure of technology acceptance. Findings of the
study suggest several potential areas where users of automotive telematics systems or
developers might fundamentally and interestingly differ in technology acceptance
decision-making, compared with the user populations commonly examined in prior
research.
Through the collection of relevant literatures, this research thinks the TAM applicable
to automotive telematics systems includes TAM, TPB and C-TAM-TPB under
consideration of the characteristics of each theoretical model and the system
characteristics in the field of technology acceptance researches. Furthermore, in order to
understand the applicability of related technology acceptance theories to the empirical
 The empirical study of automotive telematics acceptance in Taiwan 63

industry, this research processes the testing and verification and makes the fit analysis as
well as the model comparative analysis for those models and their variables. For the
analysis of competition indices, we can find that TAM is a better model to apply to the
automotive telematics industry. For the path analysis, TAM and TPB has one path that
does not only attain to the level of significance, but there are two paths in C-TAM-TPB
which is unattained to the level of significance. The explanations of these three models
for latent variables are beyond the level except for the explanation for use behaviours.
For future research of this study, we have several suggestions. Firstly, due to closely
relation between the numbers or frequency of system uses and the driving frequency of
users, we use the frequency of system uses to explain that the actual use may be
interfered with the number of driving. And the explanation of actual uses by the time for
system uses may be examined with the driving distance of users. Secondly, future
researches can use the ratio of the frequency of system uses to the frequency of actual
driving to measure in order to reduce the interference by the frequency of driving.
Besides, they also can make use of the ratio of the time for system uses to the driving
time to measure in order to reduce the interference by the driving distance. Thirdly, this
research is designed to study the comparison between models to seek a better model.
In future, with the model of this research as the basis, we can increase other theories
related to the consumer behaviour model to study the factor influencing the willingness
of adoption. Finally, the scope of this research is limited to Taiwan telematics users. We
can expand the group of samples to Asian areas and further study the prevalence of the
model in the telematics industry in future.
This study suffers from some limitations. The first limitation is the possibility of a
common method bias by using a single questionnaire to measure all constructs. Another
limitation was that this study used Taiwan areas as its scopes and the automobile forum
members as its convenience samples. Additional research across different countries or
cultures will be required in order to generalise the findings. The third limitation is that the
study uses several different levels of significance (p < 0.001; p < 0.01; p < 0.05).
However, the significance level of 0.10 should be used with greater caution during the
interpretation of automotive telematics’ BIs.

References
ABI Research (2006a) ‘Automotive navigation forecasts’. Available at: http:// www.
abiresearch.com/products/market_data/Automotive_Navigation_Forecasts. Accessed on
August 2007.
ABI Research (2006b) ‘Automotive telematics: global consumer telematics markets and
technologies’. Available at: http://www.abiresearch.com/ products/market_research/
Automotive_Telematics. Accessed on August 2007.
Agarwal, R. and Prasad, J. (1999) ‘Are individual differences germane to the acceptance of new
information technologies?’, Decision Sciences, Vol. 30, pp.361–391.
Ajzen, I. (1991) ‘The theory ofplanned behavior’, Organizational Behavior and Human Decision
Process, Vol. 50, pp.179–211.
Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behavior.
New Jersy, NJ: Prentice-Hall.
Alliance of Automobile manufacturers (2006) ‘Today’s automobile: a computer on wheels’.
Available at: http://www.autoalliance.org/innovation/. Accessed on August 2007.
64 H-H. Chen and S-C. Chen

Aloi, D.N., Alsliety, M. and Akos, D.M. (2007) ‘A methodology for the evaluation of a GPS
receiver performance in telematics applications’, IEEE Transactions on Instrumentation and
Measurement, Vol. 6, pp.11–24.
Anderson, J.C. and Gerbing, D.W. (1988) ‘Structural equation modeling in practice: a review and
recommended two-step approach’, Psychological Bulletin, Vol. 103, pp.411–423.
Bollen, K.A. (1989) Structural Equations with Latent Variables. New York, NY: Wiley.
Chau, P.Y.K. and Hu, P.J-H. (2002) ‘Investigating healthcare professionals’ decisions to accept
telemedicine technology: an empirical test of competing theories’, Information and
Management, Vol. 39, pp.297–311.
Chen, L.D. (2008) ‘A model of consumer acceptance of mobile payment’, Int. J. Mobile
Communications, Vol. 6, pp.32–52.
Davis, F.D. (1989) ‘Perceived usefulness, perceived ease of use, and user acceptance of
information technology’, MIS Quarterly, Vol. 13, pp.319–340.
Davis, F.D., Bagozzi, P.R. and Warshaw, P.R. (1989) ‘User acceptance of computer technology: a
comparison of two theoretical models’, Management Science, Vol. 35, pp.982–1003.
Dishaw, M.T. and Strong, D.M. (1999) ‘Extending the technology acceptance model with
task-technology fit constructs’, Information and Management, Vol. 36, pp.9–21.
Finn, B. and Breen, P. (1996) ‘The use of transport telematics in inter-urban and rural bus service’,
IEEE Transactions on Public Transport Electronic Systems, pp.5–10.
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behavior: An Introduction to
Theory and Research Reading. MA: Addison-Wesley.
Han, S., Mustonen, P., Seppanen, M. and Kallio, M. (2006) ‘Physicians’ acceptance of mobile
communication technology: an exploratory study’, Int. J. Mobile Communications, Vol. 4,
pp.210–230.
Hatcher, L. (1994) Step-by-step Approach to using the SAS System for Factor Analysis and
Structural Equation Modeling. Cary, NC: SAS Institute Inc.
Hinkin, T.R. (1995) ‘A review of scale development practices in the study of organizations’,
Journal of Management, Vol. 21, pp.967–989.
Hu, L. and Bentler, P.M. (1995) ‘Evaluating model fit’, in R.H. Hoyle (Ed.), Structural Equation
Modeling: Concepts, Issues, and Applications. Thousand Oaks, CA: Sagex.
Hung, M.C., Hwang, H.G. and Hsieh, T.C. (2007) ‘An exploratory study on the continuance of
mobile commerce: an extended expectation-confirmation model of information system use’,
Int. J. Mobile Communications, Vol. 5, pp.409–422.
Jackson, C.M., Chow, S. and Leitch, R.A. (1997) ‘Toward an understanding of the behavioral
intention to use an information system’, Decision Sciences, Vol. 28, pp.357–389.
Junglas, I. (2007) ‘On the usefulness and ease of use of location-based services: insights into the
information system innovator's dilemma’, Int. J. Mobile Communications, Vol. 5, pp.389–408.
Lee, T.M. and Jun, J.K. (2007) ‘The role of contextual marketing offer in mobile commerce
acceptance: comparison between mobile commerce users and nonusers’, Int. J. Mobile
Communications, Vol. 5, pp.339–356.
Lu, J., Wang, L-Z. and Yu, C-S. (2007) ‘Is TAM for wireless mobile data services applicable in
China? A survey report from Zhejiang, China’, Int. J. Mobile Communications, Vol. 5,
pp.11–31.
Melvin, D. (1999) ‘Integrating GPS receivers into consumer mobile electronics’, IEEE MultiMedia,
Vol. 6, pp.88–90.
Nunnally, J. (1978) Psychometric Theory (2nd ed.). New York, NY: McGraw-Hill.
OnStar (2006) ‘GM to offer simplest, smartest navigation system in North America’. Available at:
http://209.235.195.221/releases_detail.php?ItemID=343. Accessed on February 2007.
Park, C-W., Joe, M-J., Byun, S-C., Choi, H-W., Jung, M-H. and Kim, Y-I. (2003) ‘A turn-by-turn
navigation system for automotive telematics terminals’, IEEE Proceedings, Intelligent
Vehicles Symposium, pp.21–24.
 The empirical study of automotive telematics acceptance in Taiwan 65

Paul, J., Hu, P.Y.K.C., Olivia, R. and Liu Sheng, K.Y.T. (1999) ‘Examining the technology
acceptance model using physician acceptance of telemedicine technology’, Journal of
Management Information Systems, Vol. 16, pp.91–112.
Pavlou, P.A. (2003) ‘Consumer acceptance of electronic commerce-integrating trust and risk with
the technology acceptance model’, Int. J. Electronic Commerce, Vol. 7, pp.69–103.
Strategy Analytics (2001) ‘In-car telematics terminals market 2000–2007’, In-Car Telematics
Terminals Forecast 2001–2007 Report, January 2002.
Tanizaki, M., Maruyama, K. and Shimada, S. (2005) ‘Acceleration technique of snake-shaped
regions retrieval method for telematics navigation service system’, International Conference
on Data Engineering, pp.949–957.
Taylor, S. and Todd, P. (1995a) ‘Assessing IT usage: the role of prior experience’, MIS Quarterly,
Vol. 19, pp.561–570.
Taylor, S. and Todd, P. (1995b) ‘Understanding Information technology usage: a test of competing
models’, Information Systems Research, Vol. 6, pp.144–176.
Telematics Research Group (2002) Auto Manufacturers Seek Younger Buyers for Telematics
Systems and Services. Telematics Research Group Inc.
Tsou, M-H. (2004) ‘Integrated mobile GIS and wireless internet map servers for environmental
monitoring and management’, Cartography and Geographic Information Science, Vol. 31,
pp.153–165.
Venkatesh, V. (2000) ‘Determinants of perceived ease of use: integrating control, intrinsic
motivation, and emotion into the technology acceptance model’, Information Systems
Research, Vol. 11, pp.265–342.
Venkatesh, V. and Davis, F.D. (1996) ‘A model of the antecedents of perceived ease of use:
development and test’, Decision Sciences, Vol. 27, pp.451–481.
Venkatesh, V., Morris, G.M. and Phillip, L.A. (2000) ‘A longitudinal field investigation of gender
difference in individual technology adoption decision-making process’, Organizational
Behavior and Human Decision Process, Vol. 83, pp.33–60.
Venkatesh, V., Morris, M.G., Davis, G.B. and Davis, F.D. (2003) ‘User acceptance of information
technology: toward a unified view’, MIS Quarterly, Vol. 27, pp.425–478.
Wu, I-L. and Chen, J-L. (2005) ‘An extension of trust and TAM model with TPB in the initial
adoption of online tax: an empirical study’, Int. J. Human Computer studies, Vol. 62,
pp.784–808.
Yang, K. (2005) ‘Exploring factors affecting the adoption of mobile commerce in Singapore’,
Telematics and Informatics, Vol. 22, pp.257–277.
Zhang, D., Wang, X-H. and Hackbarth, K. (2004) ‘OSGi based service infrastructure for context
aware automotive telematics’, IEEE Transactions on Vehicular Technology, pp.2957–2961.
Zhao, Y. (2002) ‘Telematics: safe and fun driving’, IEEE Intelligent Systems, Vol. 17, pp.10–14.