analysis
Content
Transportation Research Part C 15 (2007) 300–311
www.elsevier.com/locate/trc
Predicting electronic toll collection service adoption:
An integration of the technology acceptance model
and the theory of planned behavior
Chun-Der Chen *, Yi-Wen Fan, Cheng-Kiang Farn
Department of Information Management, National Central University, ChungLi, Taiwan, ROC
Received 11 August 2006; received in revised form 12 April 2007; accepted 12 April 2007
Abstract
In order to reduce the number of vehicles stuck in congestion, especially for stop-and-go traffic at toll plazas, the establishment of electronic toll collection (ETC) systems has been a hot issue and dominant trend in many countries. Taiwan has
joined the crowd, adding an ETC system to its toll roads in early 2006. However, despite the potential benefits for motorists, the utilization rate has been lower than expected during the introductory stage. The objective of this study is to
advance our understanding on the critical antecedents of motorists’ intention of ETC service adoption by integrating both
technology acceptance model (TAM) and theory of planned behavior (TPB) perspectives. Through empirical data collection and analysis from highway motorists who had not installed on-board units (OBU) for ETC service in Taiwan, we
found that system attributes, perceived usefulness and perceived ease of use, indeed, positively engender motorists’ attitudes towards ETC service adoption. Moreover, results also reveal that attitude, subjective norm and perceived behavioral
control positively influence the intention of ETC system adoption. Implications for practitioners and researchers, and suggestions for future research are also addressed in this study.
2007 Elsevier Ltd. All rights reserved.
Keywords: Electronic toll collection; Intelligent transportation systems; Technology acceptance model; Theory of planned behavior
1. Introduction
Faced with annually increasing demand for travel and transport of goods, transportation systems are
reaching the limits of their existing capacity. Heavy highway congestion has become one of most serious urban
problems for many countries worldwide. As congestion and service disruptions in highways continue to escalate, especially for stop-and-go traffic at toll plazas, government officials are beginning to realize that traditional transportation strategies are no longer effective. Faster, more convenient, and information
*
Corresponding author. Tel.: +886 3 422 7151x66548; fax: +886 3 425 4604.
E-mail addresses: [email protected] (C.-D. Chen), [email protected] (Y.-W. Fan), [email protected] (C.-K.
Farn).
0968-090X/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.trc.2007.04.004
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
301
technology-related solutions must be implemented in order to meet the demand for expansion of current highway systems. Consequently, the establishment of electronic toll collection (ETC) systems has been a hot issue
and dominant trend in many countries.
ETC is part of an intelligent transportation system and it is also a fairly mature technology that allows for
electronic payment of highway tolls by using vehicle-to-roadside communication technologies (e.g., microwave, infrared communication, or GPS technology). Motorists purchase on-board units (OBUs) that electronically identify vehicles as they pass through a toll plaza without stopping or even slowing down, with tolls
automatically charged to or debited from ‘‘smartcards’’ inserted in their OBUs. The removable credit-cardsized electronic purse contains stored value which can be periodically replenished when the balance is low
(Decorla-Souza and Kane, 1992). ETC not only eliminates the traffic queue at tollbooths and improves safety
for the motoring public, but is also coupled with potential impacts on personal travel behavior, commercial
vehicle operations, and greater electronic commerce opportunities in particular (Golob and Regan, 2001).
In Taiwan, Far Eastern Electronic Toll Collection Co. (FETC), the build-operate-transfer project contractor, was commissioned by the Taiwan Area National Freeway Bureau to install the nation’s first ETC system
in 22 toll plazas along two North-South highways, which carry 5–6 million vehicles per year. After being tested
in a variety of conditions, the construction of ETC was completed by end 2005 and launched on February 10,
2006. Furthermore, government officials announced that a satellite-based vehicle positioning system (VPS)
would be implemented in 2008, with the entire system being operational by July 2010. Taiwan authorities envision that all manual tolls will eventually be replaced.
However, during the initial stage of the first three months – and despite potential benefits for motorists,
large-scale TV commercials, and diverse promotion activities for the ETC system – the utilization rate of large
vehicles (e.g., tourist coaches, trucks, container vehicles) was only around 20%. Furthermore, the utilization
rate of ETC lanes among small vehicles (e.g., private cars) was only about 6.5% by May 6, 2006, showing that
the individual use of ETC service has been lower than expected (http://www.fetc.net.tw, 2006).
The objective of our study is to uncover the important factors affecting the intention of ETC service adoption. Although the application of ETC service is still in an early stage and why motorists choose (or do not
choose) ETC is far from being completely understood, comprehending the determinants of ETC service adoption would help our government officials better deploy and manage their information technology resources
and enhance overall effectiveness. Moreover, the initial adoption of an e-service is also one of the crucial driving forces to further influence continued use of the service (Kwon and Zmud, 1987). In order to provide a solid
theoretical basis for examining the important antecedents for ETC service adoption, this paper integrates two
important streams of literature under the nomological structure of the theory of reasoned action (TRA): (1)
the technology acceptance model (TAM) (Davis, 1989; Davis et al., 1989), and the theory of planned behavior
(TPB) (Ajzen, 1991).
In TAM, behavioral intention is determined by attitude towards usage as well as by the direct and indirect
effects of two system features: perceived usefulness and perceived ease of use (Davis, 1989, 1993). The value of
TAM in technology-driven contexts has been consistently important and widely accepted (Bernadette, 1996;
Venkatesh and Davis, 2000). However, as Rogers (1995) argued, diffusion of innovative technology is highly
related to communication channels, individuals, organizational members, and social system in addition to the
technology itself. As with most information systems, ETC service adoption could only be partially explained
by TAM since both human and social factors should also be incorporated and considered simultaneously.
Likewise, together with TAM, TPB was selected to provide a necessary theoretical premise for the research
model examined in this study.
TPB is a well-researched model which is widely used in predicting and explaining human behavior across a
variety of settings while also considering the roles of individual and social systems in the process (Ajzen, 1991).
TPB identifies three attitudinal antecedents of behavioral intention. Two reflect the perceived desirability of
performing the behavior: attitude toward outcomes of the behavior and subjective norm. The third, perceived
behavioral control, reflects perceptions that the behavior is personally controllable (Ajzen, 1987, 1991). As the
focus of this study is on the ETC service adoption setting, which is considered as an instance of the acceptance
of innovative technology intertwined with social systems and personal characteristics, the integration of TAM
and TPB for our research framework should be in a more comprehensive manner to examine the intention and
acceptance of ETC service.
302
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
This paper proceeds as follows: Section 2 reviews the theoretical foundation from previous literature and
then advances a research model and hypotheses. Section 3 details the methodology and research design, and
Section 4 presents the data analysis and hypotheses testing results. Section 5 discusses our research findings,
and finally, Section 6 concludes with limitations, implications, and potential topics for future research.
2. Theoretical framework and hypotheses
Fig. 1 identifies the key constructs and main relationships examined in the study. As shown, the dependent
variable – intention of ETC service adoption – is posited as the primary construct to determine motorist acceptance of ETC service. In addition, TAM variables (perceived usefulness and perceived ease of use) are posited
as key drivers of motorists’ attitudes. Lastly, three key building blocks of TPB, attitude, social norms, and
perceived behavioral control, are hypothesized to affect the intention of ETC service adoption. The following
section elaborates on these relationships and explains the theoretical underpinning of these hypotheses.
2.1. The theory of planned behavior
The theory of planned behavior (TPB) is an extension of the theory of reasoned action (TRA) (Fishbein
and Ajzen, 1975), which has as its main goals the prediction and understanding of human behavior. It postulates that individuals’ behavioral intentions are determinants of their actual behavior. Behavioral intention in
TPB is a function of three determinants: attitude, subjective norm, and perceived behavioral control.
Attitude refers to an individual’s positive or negative feelings about performing the target behavior. TPB
predicts that the more favorable an individual evaluates a particular behavior, the more likely he or she will
intend to perform that behavior (Ajzen, 1987). Subjective norm reflects a person’s perception that most people
who are important to him think he should or should not perform the behavior in question. The more an individual perceives that significant others think he or she should engage in the behavior, the greater an individual’s level of motivation to comply with those others (Ajzen and Fishbein, 1980). Perceived behavioral control
reflects perceptions of internal and external constraints on behavior (Taylor and Todd, 1995). This perception
of volitional control or the perceived difficulty towards completion of the act will affect an individual’s intent
as well as the successful performance of that behavior (Ajzen and Madden, 1986). TPB has shown strong pre-
Fig. 1. The proposed conceptual model and research hypotheses.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
303
dictive utility for a wide range of behavioral intentions and actual behaviors, and a recent meta-analysis conducted by Armitage and Conner (2001) also provides support for the efficacy of the TPB.
2.2. The technology acceptance model
The technology acceptance model (TAM) (Davis, 1989, 1993; Davis et al., 1989) is grounded in both TRA
and TPB. TAM was specifically tailored for modeling user acceptance of an information system with the aim
of explaining the behavioral intention to use the system. In this model, perceived usefulness and perceived ease
of use were considered as two predecessors affecting attitude toward a technology, which affects behavioral
intention to use that technology (which, in turn, leads to its actual use).
Perceived usefulness refers to the degree to which a person believes that using a particular system would
enhance his or her job performance, and perceived ease of use is defined as the degree to which a person
believes that using a particular system would be free of effort (Davis, 1989). The effects of external variables
(e.g., system design characteristics) on behavioral intention are mediated by these beliefs. Moreover, perceived
ease of use has a direct effect on perceived usefulness and perceived usefulness could affect behavioral intention
directly (Davis et al., 1989; Venkatesh and Davis, 2000). Based on TAM and applied to the context of ETC
service adoption, an ETC system’s OBU that is perceived as facilitating the transaction process and being easy
to operate is likely to be accepted by consumers. Research following TAM consistently showed that there is a
positive relationship between PU and PEOU with acceptance of information technology (e.g., Dishaw and
Strong, 1999; Venkatesh and Davis, 2000).
TAM has received substantial attention in information systems literature because it focuses on system use,
has reliable instruments with excellent measurement properties, is parsimonious, and is empirically sound
(Pavlou, 2003). For evaluating technologies or making comparisons between user groups or applications
within and across organizations, TAM is useful and robust. However, TAM is criticized for ignoring the social
influence on technology acceptance so it has limitations in being applied beyond the workplace (Fu et al.,
2006). Several empirical studies have suggested that TAM could be integrated with other acceptance and diffusion theories, thereby including variables related to both human and social factors and improving its predictive and explanatory power (e.g., Taylor and Todd, 1995; Venkatesh and Davis, 2000).
2.3. Hypotheses development
Neither TAM nor TPB have been found to provide consistently superior explanations or predictions of
behavior (Taylor and Todd, 1995; Venkatesh et al., 2003) since factors influencing consumers’ adoption of
technology can vary, depending on the technology, target consumers, and context. As Chau and Hu (2002)
stated, theory testing follows replication logic and hence makes theory comparison an attractive approach,
i.e. generating support for a theory (or some theories) and, at the same time, singling out the potential limitation of others (p. 298). Dishaw and Strong (1999) also suggested that an integrated model may provide more
explanatory power than can either model alone. Therefore, we proposed a research framework integrating elements derived from both TAM and TPB since they are leading theoretical models and have accumulated fairly
strong empirical support involving various end-users and business managers.
In the context of ETC service adoption, TPB suggests that a motorist is more willing to adopt ETC service if he
or she has positive attitude towards using ETC, wants to comply with other important people’s opinions on the
use of ETC, and has the requisite resources, skills, or opportunities. In addition, by using TAM as a basic structure
for our research context, favorable system attributes of the ETC system, perceived usefulness and perceived ease
of use, will positively influence motorists’ attitudes toward ETC system service. Moreover, perceived usefulness
has a direct effect on the intention of ETC service adoption, and perceived ease of use has a direct effect on perceived usefulness of an ETC system. Accordingly, the hypotheses are presented below and as indicated in Fig. 1.
Hypothesis 1 (H1): Attitude towards ETC system service positively increases the intention of ETC service
adoption.
Hypothesis 2 (H2): Subjective norms towards ETC system service positively increases the intention of ETC
service adoption.
304
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
Hypothesis 3 (H3): Motorist’s perceived behavioral control towards ETC system service positively increases
the intention of ETC service adoption.
Hypothesis 4 (H4): Perceived usefulness of ETC system service positively increases the attitude towards
ETC system service.
Hypothesis 5 (H5): Perceived usefulness of ETC system service positively increases the intention of ETC
service adoption.
Hypothesis 6 (H6): Perceived ease of use of ETC system service positively increases the attitude towards
ETC system service.
Hypothesis 7 (H7): Perceived ease of use of ETC system service positively increases the perceived usefulness
of ETC system service.
3. Methodology and research design
3.1. Sample and data collection
A self-administered questionnaire was used for our research purpose. Since this paper aimed to examine the
effects of TAM and TPB variables on the intention of ETC service adoption, we surveyed individual motorists
of private vehicles who had not yet installed OBU. We employed trained doctoral students as interviewers to
conduct the study and mall-intercept personal interviews (Bush and Hair, 1985), were also administered in several major rest areas along two North–South highways in Taiwan.
Participants were first asked whether they had installed an OBU for using ETC service. If so, after briefly
elucidating our research purpose, they were invited to participate and complete the survey questionnaire. A
total of 300 questionnaires were distributed and a total of 264 completed questionnaires were returned. Since
9 questionnaires were invalid, 255 responses were obtained and valid (85.00% response rate). Specific demographic information is shown in Table 1.
Table 1
Demographic profile of the respondents (N = 255)
Demographic variables
Frequency
Percentage
Gender
Male
Female
197
58
77.25
22.75
Age
18–29
30–39
40–49
50–59
More than 59
20
113
93
27
2
7.84
44.31
36.47
10.59
0.78
Education
Elementary school or less
Junior high school
High school
University
Postgraduate degree
1
5
28
130
91
0.39
1.96
10.98
50.98
35.69
1
35
70
71
65
13
0.39
13.73
27.45
27.84
25.49
5.10
Frequencies of passing toll plazas
Ten times or above per day
One time or above per day
Several times per week
Several times per month
One time for several months
Few/barely
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
305
As shown in Table 1, 77.25% are males and the majority of respondents (80.78%) are in the age group of
30–39 and 40–49 years. In addition, most respondents are highly educated with 86.67% of them having
attained university diplomas or postgraduate degrees. Moreover, the frequency of passing toll plazas show
that 41.57% of our respondents are frequent highway motorists. Despite passing through toll plazas quite
often (10 times per day, one time or above per day, and several times per week), they have still not installed
OBUs.
3.2. Measurement development
The operationalization, sources, and standardized loadings of measurement items are shown in Appendix
A. To test the framework, we paid particular attention to issues of operationalization and measurement in this
study, following Venkatraman and Grant (1986). We operationalized the variables in two ways: (1) for those
variables that have been previously employed in a research setting, we adopted the measures with acceptable
measurement quality; and (2) for those variables that were unique to our conceptual model, we conducted
belief elicitation interviews with motorists and developed the operational measures. In addition, respecting
Nunnally’s (1978) recommended procedures, we developed multiple items for each construct when possible.
The preliminary instrument was pilot tested and reviewed by motorists for clarity again, thereby confirming
the content validity of our instrument. All items were five-point, Likert-type scales anchored at ‘‘strongly disagree’’ (1), ‘‘strongly agree’’ (5), and ‘‘neither agree nor disagree’’ (3). The scales of the intention of ETC service adoption are borrowed from Yu et al. (2005). For measuring attitude, subjective norm and perceived
behavioral control, the items are adopted from Taylor and Todd (1995). The perceived usefulness scales
are adopted from Moore and Benbasat (1991). Finally, as for perceived ease of use, we adopted this from
Bhattacherjee (2000).
4. Data analysis and results
4.1. Convergent validity and discriminant validity
In analyzing the collected data, this study followed the two-step procedure suggested by Anderson and Gerbing (1988). We estimated and respecified the measurement model prior to incorporating the structural restrictions. Convergent and discriminant validity of the remaining items and scales were tested with confirmatory
factory analysis (CFA) using the LISREL 8.50 program (Bentler and Chou, 1987). The result of the CFA
indicated that the measurement model provided a very good fit to the data: x2(131) = 328.37, Bentler
Bonett Normed Fit Index (NFI) = 0.95, Non-Normed Fit Index (NNFI) = 0.96, Comparative Fit Index
(CFI) = 0.97, Goodness-of-Fit Index (GFI) = 0.88, and Root Mean Square Error of Approximation
(RMSEA) = 0.077.
Convergent validity was assessed based on the criteria that the indicator’s estimated pattern coefficient was
significant on its posited underlying construct factor. We evaluated for the measurement scales using the three
criteria suggested by Fornell and Larcker (1981):
(1) All indicator factor loadings (k) should be significant and exceed 0.5.
(2) Construct reliabilities should exceed 0.8.
(3) Average variance extracted (AVE) by each construct should exceed the variance due to measurement
error for that construct (e.g., AVE should exceed 0.5).
All k values in the CFA model exceeded 0.5 and were significant at p = 0.001 except two indicators for perceived usefulness which were dropped due to their cross-loadings. We dropped these items from further analysis. Composite reliabilities of constructs ranged from 0.87 to 0.97 (see Table 2). AVE, ranging from 0.63 to
0.95, was greater than the variance due to measurement error. Therefore, all three conditions for convergent
validity were met.
Before examining the discriminant validity, however, we found that some constructs – attitude, perceived
behavioral control and perceived usefulness – may cause a multicollinearity problem because of their high
306
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
Table 2
Reliability, correlation coefficients and AVE results
Variable
Mean
SD
Cronbach’s alpha
Composite reliability
AVE
(1)
(2)
(3)
(4)
(5)
(6)
ATT
SN
PBC
INTENT
PU
PEOU
12.81
4.94
12.58
5.05
15.95
13.39
4.37
2.41
4.04
2.48
5.37
4.34
0.94
0.97
0.89
0.98
0.93
0.96
0.93
0.96
0.87
0.97
0.91
0.95
0.77
0.93
0.63
0.95
0.77
0.83
0.88
0.69
0.74
0.69
0.72
0.62
0.97
0.62
0.67
0.64
0.49
0.79
0.68
0.61
0.76
0.98
0.63
0.57
0.88
0.55
0.91
Notes:
1. The main diagonal shows the square root of the AVE (average variance extracted).
2. ATT for attitude, SN for subjective norm, PBC for perceived behavioral control, INTENT for intention of ETC service adoption, PU
for perceived usefulness, and PEOU for perceived ease of use.
3. Significant at p < .01 level is shown in bold.
correlation. Multicollinearity may lead to artifactual outcomes because of the effects of two or more predictors
being substantially confounded (Miles and Shevlin, 2001). Multicollinearity is indexed by the variance inflation factor (VIF) statistic; there has been little agreement in the literature, however, as to a suitable VIF cutoff
point at which the problem should give cause for concern. Recommended VIF criteria for problematical levels
of multicollinearity encompass values of 2.50 (Allison, 1999), 4.00 (Miles and Shevlin, 2001), and 10.00 (Everitt, 1996). We followed the procedure suggested by Miles and Shevlin (2001) and the VIF values of these constructs are all less than 3.4, showing there is no serious multicollinearity problem.
Finally, discriminant validity was shown when the square root of each construct’s AVE is larger than its
correlations with other constructs (Chin, 1998). As illustrated in Table 2, the square root of the AVE is larger
than all other cross-correlations. Hence the latter test of discriminant validity was also met. In addition to reliability coefficients and AVE values, Table 2 also reports the correlation matrix, means, and standard deviations of the study’s principal constructs.
4.2. Hypothesis testing
The research model presented earlier was tested using the structural equation modeling (SEM) approach.
Overall, the goodness-of-fit of the structural model was comparable to that of the previous CFA model and
provided evidence of adequate fit. As shown in Fig. 2, attitude (b = 0.22, p < 0.05) is positively associated with
the intention of ETC service adoption, rendering support for H1. In addition, both subjective norm (b = 0.23,
p < 0.001) and perceived behavioral control (b = 0.36, p < 0.001) were also significant predictors of the intention of ETC service adoption, thereby supporting H2 and H3. Furthermore, among these three building
blocks of TPB, perceived behavioral control has the strongest effect on the intention of ETC service adoption.
The impact of perceived usefulness (b = 0.58) and perceived ease of use (b = 0.38) on attitude are significant
at p = 0.001. Therefore, H4 and H6 can be accepted. This shows that both system attributes – perceived usefulness and perceived ease of use – are important antecedents in forming the motorist’s attitude towards ETC
service adoption. In other words, an ETC system service that is not advantageous for motorists and does not
help them perform their toll collection is not likely to be received favorably nor will a positive attitude be
formed towards it in spite of careful implementation efforts.
Meanwhile, it shows that perceived usefulness has no significant impact on the intention of ETC service
adoption, and thus H5 cannot be accepted. For the result in perceived usefulness, previous empirical studies
on TAM have shown inconsistency with either significant influence (Moore and Benbasat, 1991; Chau, 1996)
or with insignificant influence on behavioral intention to use (Chen et al., 2002). It might imply an indirect
influence of perceived usefulness on the intention of ETC service adoption via the mediator (e.g., attitude)
towards ETC service adoption in our model.
Moreover, perceived ease of use has a strong effect on perceived usefulness (b = 0.61, p < 0.001), validating
H7 and allowing the inference that perceived ease of use fosters a motorist’s perceived usefulness towards ETC
system service. We will discuss these findings in detail in the next section.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
307
Fig. 2. Data analysis results.
5. Discussion
This study aims to shed light on the critical antecedents of motorists’ intentions of ETC service adoption by
integrating both TAM and TPB perspectives. The overall explanatory power of our research model was relatively high; a R-square of 61% for the intention of ETC service adoption, a R-square of 74% for attitude
towards ETC service adoption, and a R-square of 38% for perceived usefulness were obtained, suggesting that
both theories are capable of explaining a relatively high proportion of the variation in the intention to adopt
ETC service.
Several insightful results could be summarized from our research framework as follows. First, motorist’s
perceived behavioral control was the most important predictor of the intention of ETC service adoption. This
finding is consistent with Mathieson’s (1991) identification of perceived behavioral control as an important
determinant to behavioral intention. Our result also reinforces that the inclusion of perceived behavioral control significantly improved the prediction of intentions (Ajzen, 1991). Following TPB, therefore, it is expected
that motorist’s perceived constraints of carrying out the transaction of ETC affects whether or not that behavior will be carried out. For example, higher price and the installation fee of the OBU might increase the degree
of perceived external constraints of a motorist, thereby decreasing the willingness to adopt ETC service.
Second, subjective norm suggests that behavior is instigated by one’s desire to act as how others act or think
one should act, and it appeared to be the second most important determinant of a motorist’s intention to
adopt ETC service. Our finding is similar to those reported by Bhattacherjee (2000) and Karahanna et al.
(1999) which indicated that subjective norm influences behavioral intention towards system use and subjective
norm could influence intention to use as strongly as attitude does. Although some prior studies (e.g., Venkatesh and Davis, 2000; Mathieson, 1991; Davis et al., 1989) found that subjective norm could significantly
determine intention to use in a mandatory-usage context, its impact would be less significant where users were
in a voluntary-usage context as with the case of ETC service adoption in this study. However, subjective norm
has been found to be more important in the early stages of system development and was a more important
predictor of intention for people without prior experience (Taylor and Todd, 1995). In particular, as ETC service adoption in this study is at the introduction stage, there is a lack of references from prior adopters such as
friends, peers and superiors. Furthermore, after the formal launch of the ETC system in Taiwan, disputes,
public complaints, and allegations of impropriety have been rife (e.g., perceived social pressure from TV
308
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
news). Accordingly, it is reasonable to expect that the effect of subjective norm on the intention of ETC service
adoption should indicate significance and have as strong an effect as attitude does.
Third, our finding supports TPB and TAM, which both predict that attitude towards behavior is a significant determinant of behavioral intention. Attitude has long been shown to influence behavioral intentions
(Ajzen and Fishbein, 1980). This relationship has received substantial empirical support. For our research
context, the motorists’ attitudes could strongly determine their willingness to use OBU for ETC service.
Fourth, we also confirmed that both perceived usefulness and perceived ease of use determine attitude
towards ETC service adoption. The results reveal that motorists’ perceptions of the ease and effectiveness
of ETC service may indirectly lead to their intention to adopt ETC service via their attitude. In addition,
the effect of perceived usefulness on attitude is greater than that of perceived ease of use (0.58 > 0.38). As
Keil et al. (1995) concluded, in determining whether or not to use a technology, usefulness is more important
than ease of use. In our context, an essential factor for a motorist’s acceptance of ETC service is whether the
technology fills his or her service needs. In other words, ETC is acceptable is only when demonstrating proven or desired utility in its practices. Moreover, the significant positive effect of perceived ease of use on perceived usefulness is also confirmed, and this is consistent with the results from many prior studies that
examined TAM.
Lastly, perceived usefulness appeared to have no significant effect on the intention of ETC service adoption.
A plausible explanation is that perceived usefulness might often indicate its influence on the intention through
the mediator of attitude. Moreover, the context of this study is focused on the stage of initial adoption and
voluntary use of ETC service. Motorists’ positive perceived usefulness in using ETC service may not immediately lead to a behavioral intention to use, but rather to firstly form a favorable attitude toward using ETC
service. In other words, potential users such as motorists would need to take a period of time to carefully
change their psychological state in order to adopt ETC service.
6. Conclusions
Given the turbulence of many industries, understanding what facilitates the delivery of products and services to satisfy customers’ needs offers scholars and practitioners continuously and increasingly important
challenges. By integrating two important intention-based theoretical perspectives, TAM and TPB, we expect
to provide meaningful insights in a more comprehensive manner that jointly predicts user acceptance of ETC
service. This could be especially valuable for government policymakers and BOT vendors for their policy
development and marketing practices.
However, we also acknowledge that a number of research limitations exist in our research which should be
overcome in the future. First, the conclusions drawn from our study are based on cross-sectional data. As suggested by Ajzen (1991), the relative importance of attitude, subjective norm, and perceived behavioral control
in predicting behavioral intention may vary across behavior and situation. With our cross-sectional data, we
only took a snapshot of this model. A stricter test of our argument, however, would be to use a longitudinal
study to evaluate this aspect more critically since the implementation of ETC service is only just beginning in
Taiwan. By using a longitudinal study in the future, we could investigate our research model in different time
periods and make comparisons, thus providing more insights into ETC adoption phenomenon and more contributions for both TAM and TPB perspectives.
Second, although our model provides some insights to explain the intention of ETC service adoption, some
possible moderating effects between TPB variables (e.g., attitude, subjective norm, and perceived behavioral
control) and the intention of ETC service adoption are not well understood. Future studies may benefit from
articulating the possible moderating factor (e.g., government’s procedural justice) that enhance or impede such
ETC adoption intention that is most compatible with such purposes. In sum, these questions open up fertile
grounds for future research opportunities.
These findings might have policy implications for the implementation of ETC and e-government related
services. Our research results revealed that attitude, subjective norm, and perceived behavioral control could
positively influence motorist’s intention of ETC service adoption. Moreover, the impact of perceived behavioral control has the strongest effect on adoption. Accordingly, this highlights the need for several strategic
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
309
plans to be carried out synchronously for gaining motorist’s trust and engendering the intention of ETC service adoption as soon as possible.
First, as discussed earlier, perceived behavioral control refers to perceptions of internal and external constraints (e.g., skills, information searching, available budget, time and places for OBU installation) on behavior (Ajzen, 1987, 1991). Likewise, we suggest that government policymakers and BOT vendors could lower the
OBU price and establish more service locations for OBU installation, thereby increasing the convenience and
willingness to use ETC service and lessen motorists’ possible external constraints. Next, advertising and marketing plans containing ETC system performance reports or adopters’ successful experience toward ETC service might also serve as possible and effective ways to attract potential or non-adopters.
To conclude, the results of this study demonstrate that while TAM by itself is useful in predicting ETC service adoption, integration with TPB could provide a more complete understanding of behavioral intention.
This should help to better manage the system implementation process by focusing attention on several control
factors such as system attributes, individual preference, social influences, and individual constraints in the
ETC service and related e-government issues.
Appendix A. Measurements
Constructs
Standardized
loadings
Attitude (Taylor and Todd, 1995)
• Using the ETC service is a good idea
• Using the ETC service is a wise idea
• I like the idea of using the ETC service
• Using the ETC service would be pleasant
0.9
0.86
0.93
0.81
Subjective norm (Taylor and Todd, 1995)
• People who influence my behavior would think that I should use the ETC service
• People who are important to me would think that I should use the ETC service
0.96
0.97
Percevied behavioral control (Taylor and Todd, 1995)
• It is not troublesome for me to replenish the stored value of smartcard of OBU
0.85
• I have available time to install the related equipment of ETC system (e.g., OBU)
0.84
• I have the resources and the knowledge and the ability to replenish the stored value of 0.8
smartcard of OBU
• I can afford the application fee of ETC system service
0.67
Perceived usefulness (Moore and Benbasat, 1991)
• Using the ETC service enables me to pass toll plazas more quickly
• Using the ETC service improves the efficiency of using highways
• Using the ETC service makes it easier to pay toll
• Using the ETC service enables me to reach my destination more quickly
• Using the ETC service makes me effectively control my travel time on highways
Dropped
0.92
0.79
0.92
Dropped
Perceived ease of use (Bhattacherjee, 2000)
• It is easy for me to learn how to replenish the stored value of smartcard of OBU
• It is easy for me to use the function of smartcard balance replenishment
• It is easy for me to be accustomed to use ETC service
• ETC system is easy to use for me
0.86
0.87
0.96
0.95
Intention of etc system adoption (Yu et al., 2005)
• I will intend to apply the ETC service as soon as possible
• I will use the ETC service soon after it is launched
0.98
0.97
310
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
References
Ajzen, I., 1987. Attitudes, traits, and actions: dispositional prediction of behavior in social psychology. Advances in Experimental Social
Psychology 20, 1–63.
Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50 (2), 179–211.
Ajzen, I., Fishbein, M., 1980. Understanding Attitudes and Predicting Social Behavior. Prentice-Hall Inc., Englewood Cliffs, NJ.
Ajzen, I., Madden, T.J., 1986. Prediction of goal directed behavior: attitudes, intentions and perceived behavioral control. Journal of
Experimental Social Psychology 22, 453–474.
Allison, P.D., 1999. Multiple Regression: A Primer. Pine Forge Press, Thousand Oaks, CA.
Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a review and recommended two-step approach.
Psychological Bulletin 103 (3), 411–423.
Armitage, C.J., Conner, M., 2001. Efficacy of the theory of planned behavior: a meta-analytic review. British Journal of Social Psychology
40 (3), 471–499.
Bentler, P.M., Chou, C.P., 1987. Practical issues in structural modeling. Sociological Methods and Research 16 (1), 78–117.
Bernadette, S., 1996. Empirical evaluation of the revised technology acceptance model. Management Science 42 (1), 85–93.
Bhattacherjee, A., 2000. Acceptance of e-commerce services: the case of electronic brokerages. IEEE Transaction on Systems, Man, and
Cybernetics – Part A: Systems and Humans 30 (4), 411–420.
Bush, A.J., Hair, J.F., 1985. An assessment of the mall intercept as a data collection method. Journal of Marketing Research 22 (2), 158–
167.
Chau, Y.K., 1996. An empirical assessment of a modified technology acceptance model. Journal of Management Information Systems 13
(2), 185–204.
Chau, P.Y.K., Hu, P.J.H., 2002. Investigating healthcare professional’s decisions to accept telemedicine technology: an empirical test of
competing theories. Information and Management 39, 297–311.
Chen, L.D., Gillenson, M.L., Sherrell, D.L., 2002. Enticing online consumers: an extended technology acceptance perspective.
Information and Management 39, 705–719.
Chin, W.W., 1998. Issues and opinion on structural equation modeling. MIS Quarterly 22 (1), 7–16.
Davis, F.D., 1989. Perceived usefulness, perceived ease of use and user acceptance of information technology. MIS Quarterly 13 (3), 319–
340.
Davis, F.D., 1993. User acceptance of information technology: system characteristics, user perceptions and behavioral impacts.
International Journal of Man Machine Studies 38 (3), 475–487.
Davis, F.D., Bagozzi, R.P., Warshaw, P.R., 1989. User acceptance of computer technology: a comparison of two theoretical models.
Management Science 35 (8), 982–1003.
Decorla-Souza, P., Kane, A.R., 1992. Peak period tolls: precepts and prospects. Transportation 19 (4), 293–311.
Dishaw, M.T., Strong, D.M., 1999. Extending the technology acceptance model with task-technology fit constructs. Information and
Management 36 (1), 9–21.
Everitt, B.S., 1996. Making Sense of Statistics in Psychology. Oxford University Press, Oxford, England.
Fishbein, M., Ajzen, I., 1975. Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addision-Wesley,
Reading, MA.
Fornell, C.R., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of
Marketing Research 18 (1), 39–50.
Fu, J.R., Farn, C.K., Chao, W.P., 2006. Acceptance of electronic tax filing: a study of taxpayer intentions. Information and Management
43 (1), 109–126.
Golob, T.F., Regan, A.C., 2001. Impacts of information technology on personal travel and commercial vehicle operations: research
challenges and opportunities. Transportation Research Part C. Emerging Technologies 9 (2), 87–121.
Karahanna, E., Straub, D.W., Chervany, N.L., 1999. Information technology adoption across time: a cross-sectional comparison of preadoption and post-adoption beliefs. MIS Quarterly 23 (2), 183–213.
Keil, M., Beranek, P.M., Konsynski, B.R., 1995. Usefulness and ease of use: field study evidence regarding task considerations. Decision
Support Systems 13 (1), 75–91.
Kwon, T.H., Zmud, R.W., 1987. Unifying the fragmented models of information systems implementation. In: Boland, R.J., Hirschheim,
R.A. (Eds.), Critical Issues in Information Systems Research. Wiley, New York, pp. 27–251.
Mathieson, K., 1991. Predicting user intentions: comparing the technology acceptance model with the theory of planned behavior.
Information Systems Research 2 (3), 173–191.
Miles, J.N.V., Shevlin, M.E., 2001. Applying Regression and Correlation: A Guide for Students and Researchers. Sage, London.
Moore, G.C., Benbasat, I., 1991. Development of an instrument to measure the perceptions of adoption an information technology
innovation. Information Systems Research 2 (3), 192–222.
Nunnally, J.C., 1978. Psychometric Theory, second ed. McGraw-Hill, New York.
Pavlou, P.A., 2003. Consumer acceptance of electronic commerce: integrating trust and risk with the technology acceptance model.
International Journal of Electronic Commerce 7 (3), 101–134.
Rogers, E.M., 1995. The Diffusion of Innovation, fourth ed. Free Press, New York.
Taylor, S., Todd, P.A., 1995. Understanding information technology usage: a test of competing models. Information Systems Research 6
(2), 144–176.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
311
Venkatesh, V., Davis, F.D., 2000. A theoretical extension of the technology acceptance model: four longitudinal field studies. Management
Science 46 (2), 186–204.
Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of information technology: toward a unified view. MIS
Quarterly 27 (3), 425–478.
Venkatraman, N., Grant, J.H., 1986. Construct measurement in strategy research: a critique and proposal. Academy Management Review
11 (1), 71–86.
Yu, J., Ha, I., Choi, M., Rho, J., 2005. Extending the TAM for a t-commerce. Information and Management 42 (7), 965–976.
www.elsevier.com/locate/trc
Predicting electronic toll collection service adoption:
An integration of the technology acceptance model
and the theory of planned behavior
Chun-Der Chen *, Yi-Wen Fan, Cheng-Kiang Farn
Department of Information Management, National Central University, ChungLi, Taiwan, ROC
Received 11 August 2006; received in revised form 12 April 2007; accepted 12 April 2007
Abstract
In order to reduce the number of vehicles stuck in congestion, especially for stop-and-go traffic at toll plazas, the establishment of electronic toll collection (ETC) systems has been a hot issue and dominant trend in many countries. Taiwan has
joined the crowd, adding an ETC system to its toll roads in early 2006. However, despite the potential benefits for motorists, the utilization rate has been lower than expected during the introductory stage. The objective of this study is to
advance our understanding on the critical antecedents of motorists’ intention of ETC service adoption by integrating both
technology acceptance model (TAM) and theory of planned behavior (TPB) perspectives. Through empirical data collection and analysis from highway motorists who had not installed on-board units (OBU) for ETC service in Taiwan, we
found that system attributes, perceived usefulness and perceived ease of use, indeed, positively engender motorists’ attitudes towards ETC service adoption. Moreover, results also reveal that attitude, subjective norm and perceived behavioral
control positively influence the intention of ETC system adoption. Implications for practitioners and researchers, and suggestions for future research are also addressed in this study.
2007 Elsevier Ltd. All rights reserved.
Keywords: Electronic toll collection; Intelligent transportation systems; Technology acceptance model; Theory of planned behavior
1. Introduction
Faced with annually increasing demand for travel and transport of goods, transportation systems are
reaching the limits of their existing capacity. Heavy highway congestion has become one of most serious urban
problems for many countries worldwide. As congestion and service disruptions in highways continue to escalate, especially for stop-and-go traffic at toll plazas, government officials are beginning to realize that traditional transportation strategies are no longer effective. Faster, more convenient, and information
*
Corresponding author. Tel.: +886 3 422 7151x66548; fax: +886 3 425 4604.
E-mail addresses: [email protected] (C.-D. Chen), [email protected] (Y.-W. Fan), [email protected] (C.-K.
Farn).
0968-090X/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.trc.2007.04.004
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
301
technology-related solutions must be implemented in order to meet the demand for expansion of current highway systems. Consequently, the establishment of electronic toll collection (ETC) systems has been a hot issue
and dominant trend in many countries.
ETC is part of an intelligent transportation system and it is also a fairly mature technology that allows for
electronic payment of highway tolls by using vehicle-to-roadside communication technologies (e.g., microwave, infrared communication, or GPS technology). Motorists purchase on-board units (OBUs) that electronically identify vehicles as they pass through a toll plaza without stopping or even slowing down, with tolls
automatically charged to or debited from ‘‘smartcards’’ inserted in their OBUs. The removable credit-cardsized electronic purse contains stored value which can be periodically replenished when the balance is low
(Decorla-Souza and Kane, 1992). ETC not only eliminates the traffic queue at tollbooths and improves safety
for the motoring public, but is also coupled with potential impacts on personal travel behavior, commercial
vehicle operations, and greater electronic commerce opportunities in particular (Golob and Regan, 2001).
In Taiwan, Far Eastern Electronic Toll Collection Co. (FETC), the build-operate-transfer project contractor, was commissioned by the Taiwan Area National Freeway Bureau to install the nation’s first ETC system
in 22 toll plazas along two North-South highways, which carry 5–6 million vehicles per year. After being tested
in a variety of conditions, the construction of ETC was completed by end 2005 and launched on February 10,
2006. Furthermore, government officials announced that a satellite-based vehicle positioning system (VPS)
would be implemented in 2008, with the entire system being operational by July 2010. Taiwan authorities envision that all manual tolls will eventually be replaced.
However, during the initial stage of the first three months – and despite potential benefits for motorists,
large-scale TV commercials, and diverse promotion activities for the ETC system – the utilization rate of large
vehicles (e.g., tourist coaches, trucks, container vehicles) was only around 20%. Furthermore, the utilization
rate of ETC lanes among small vehicles (e.g., private cars) was only about 6.5% by May 6, 2006, showing that
the individual use of ETC service has been lower than expected (http://www.fetc.net.tw, 2006).
The objective of our study is to uncover the important factors affecting the intention of ETC service adoption. Although the application of ETC service is still in an early stage and why motorists choose (or do not
choose) ETC is far from being completely understood, comprehending the determinants of ETC service adoption would help our government officials better deploy and manage their information technology resources
and enhance overall effectiveness. Moreover, the initial adoption of an e-service is also one of the crucial driving forces to further influence continued use of the service (Kwon and Zmud, 1987). In order to provide a solid
theoretical basis for examining the important antecedents for ETC service adoption, this paper integrates two
important streams of literature under the nomological structure of the theory of reasoned action (TRA): (1)
the technology acceptance model (TAM) (Davis, 1989; Davis et al., 1989), and the theory of planned behavior
(TPB) (Ajzen, 1991).
In TAM, behavioral intention is determined by attitude towards usage as well as by the direct and indirect
effects of two system features: perceived usefulness and perceived ease of use (Davis, 1989, 1993). The value of
TAM in technology-driven contexts has been consistently important and widely accepted (Bernadette, 1996;
Venkatesh and Davis, 2000). However, as Rogers (1995) argued, diffusion of innovative technology is highly
related to communication channels, individuals, organizational members, and social system in addition to the
technology itself. As with most information systems, ETC service adoption could only be partially explained
by TAM since both human and social factors should also be incorporated and considered simultaneously.
Likewise, together with TAM, TPB was selected to provide a necessary theoretical premise for the research
model examined in this study.
TPB is a well-researched model which is widely used in predicting and explaining human behavior across a
variety of settings while also considering the roles of individual and social systems in the process (Ajzen, 1991).
TPB identifies three attitudinal antecedents of behavioral intention. Two reflect the perceived desirability of
performing the behavior: attitude toward outcomes of the behavior and subjective norm. The third, perceived
behavioral control, reflects perceptions that the behavior is personally controllable (Ajzen, 1987, 1991). As the
focus of this study is on the ETC service adoption setting, which is considered as an instance of the acceptance
of innovative technology intertwined with social systems and personal characteristics, the integration of TAM
and TPB for our research framework should be in a more comprehensive manner to examine the intention and
acceptance of ETC service.
302
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
This paper proceeds as follows: Section 2 reviews the theoretical foundation from previous literature and
then advances a research model and hypotheses. Section 3 details the methodology and research design, and
Section 4 presents the data analysis and hypotheses testing results. Section 5 discusses our research findings,
and finally, Section 6 concludes with limitations, implications, and potential topics for future research.
2. Theoretical framework and hypotheses
Fig. 1 identifies the key constructs and main relationships examined in the study. As shown, the dependent
variable – intention of ETC service adoption – is posited as the primary construct to determine motorist acceptance of ETC service. In addition, TAM variables (perceived usefulness and perceived ease of use) are posited
as key drivers of motorists’ attitudes. Lastly, three key building blocks of TPB, attitude, social norms, and
perceived behavioral control, are hypothesized to affect the intention of ETC service adoption. The following
section elaborates on these relationships and explains the theoretical underpinning of these hypotheses.
2.1. The theory of planned behavior
The theory of planned behavior (TPB) is an extension of the theory of reasoned action (TRA) (Fishbein
and Ajzen, 1975), which has as its main goals the prediction and understanding of human behavior. It postulates that individuals’ behavioral intentions are determinants of their actual behavior. Behavioral intention in
TPB is a function of three determinants: attitude, subjective norm, and perceived behavioral control.
Attitude refers to an individual’s positive or negative feelings about performing the target behavior. TPB
predicts that the more favorable an individual evaluates a particular behavior, the more likely he or she will
intend to perform that behavior (Ajzen, 1987). Subjective norm reflects a person’s perception that most people
who are important to him think he should or should not perform the behavior in question. The more an individual perceives that significant others think he or she should engage in the behavior, the greater an individual’s level of motivation to comply with those others (Ajzen and Fishbein, 1980). Perceived behavioral control
reflects perceptions of internal and external constraints on behavior (Taylor and Todd, 1995). This perception
of volitional control or the perceived difficulty towards completion of the act will affect an individual’s intent
as well as the successful performance of that behavior (Ajzen and Madden, 1986). TPB has shown strong pre-
Fig. 1. The proposed conceptual model and research hypotheses.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
303
dictive utility for a wide range of behavioral intentions and actual behaviors, and a recent meta-analysis conducted by Armitage and Conner (2001) also provides support for the efficacy of the TPB.
2.2. The technology acceptance model
The technology acceptance model (TAM) (Davis, 1989, 1993; Davis et al., 1989) is grounded in both TRA
and TPB. TAM was specifically tailored for modeling user acceptance of an information system with the aim
of explaining the behavioral intention to use the system. In this model, perceived usefulness and perceived ease
of use were considered as two predecessors affecting attitude toward a technology, which affects behavioral
intention to use that technology (which, in turn, leads to its actual use).
Perceived usefulness refers to the degree to which a person believes that using a particular system would
enhance his or her job performance, and perceived ease of use is defined as the degree to which a person
believes that using a particular system would be free of effort (Davis, 1989). The effects of external variables
(e.g., system design characteristics) on behavioral intention are mediated by these beliefs. Moreover, perceived
ease of use has a direct effect on perceived usefulness and perceived usefulness could affect behavioral intention
directly (Davis et al., 1989; Venkatesh and Davis, 2000). Based on TAM and applied to the context of ETC
service adoption, an ETC system’s OBU that is perceived as facilitating the transaction process and being easy
to operate is likely to be accepted by consumers. Research following TAM consistently showed that there is a
positive relationship between PU and PEOU with acceptance of information technology (e.g., Dishaw and
Strong, 1999; Venkatesh and Davis, 2000).
TAM has received substantial attention in information systems literature because it focuses on system use,
has reliable instruments with excellent measurement properties, is parsimonious, and is empirically sound
(Pavlou, 2003). For evaluating technologies or making comparisons between user groups or applications
within and across organizations, TAM is useful and robust. However, TAM is criticized for ignoring the social
influence on technology acceptance so it has limitations in being applied beyond the workplace (Fu et al.,
2006). Several empirical studies have suggested that TAM could be integrated with other acceptance and diffusion theories, thereby including variables related to both human and social factors and improving its predictive and explanatory power (e.g., Taylor and Todd, 1995; Venkatesh and Davis, 2000).
2.3. Hypotheses development
Neither TAM nor TPB have been found to provide consistently superior explanations or predictions of
behavior (Taylor and Todd, 1995; Venkatesh et al., 2003) since factors influencing consumers’ adoption of
technology can vary, depending on the technology, target consumers, and context. As Chau and Hu (2002)
stated, theory testing follows replication logic and hence makes theory comparison an attractive approach,
i.e. generating support for a theory (or some theories) and, at the same time, singling out the potential limitation of others (p. 298). Dishaw and Strong (1999) also suggested that an integrated model may provide more
explanatory power than can either model alone. Therefore, we proposed a research framework integrating elements derived from both TAM and TPB since they are leading theoretical models and have accumulated fairly
strong empirical support involving various end-users and business managers.
In the context of ETC service adoption, TPB suggests that a motorist is more willing to adopt ETC service if he
or she has positive attitude towards using ETC, wants to comply with other important people’s opinions on the
use of ETC, and has the requisite resources, skills, or opportunities. In addition, by using TAM as a basic structure
for our research context, favorable system attributes of the ETC system, perceived usefulness and perceived ease
of use, will positively influence motorists’ attitudes toward ETC system service. Moreover, perceived usefulness
has a direct effect on the intention of ETC service adoption, and perceived ease of use has a direct effect on perceived usefulness of an ETC system. Accordingly, the hypotheses are presented below and as indicated in Fig. 1.
Hypothesis 1 (H1): Attitude towards ETC system service positively increases the intention of ETC service
adoption.
Hypothesis 2 (H2): Subjective norms towards ETC system service positively increases the intention of ETC
service adoption.
304
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
Hypothesis 3 (H3): Motorist’s perceived behavioral control towards ETC system service positively increases
the intention of ETC service adoption.
Hypothesis 4 (H4): Perceived usefulness of ETC system service positively increases the attitude towards
ETC system service.
Hypothesis 5 (H5): Perceived usefulness of ETC system service positively increases the intention of ETC
service adoption.
Hypothesis 6 (H6): Perceived ease of use of ETC system service positively increases the attitude towards
ETC system service.
Hypothesis 7 (H7): Perceived ease of use of ETC system service positively increases the perceived usefulness
of ETC system service.
3. Methodology and research design
3.1. Sample and data collection
A self-administered questionnaire was used for our research purpose. Since this paper aimed to examine the
effects of TAM and TPB variables on the intention of ETC service adoption, we surveyed individual motorists
of private vehicles who had not yet installed OBU. We employed trained doctoral students as interviewers to
conduct the study and mall-intercept personal interviews (Bush and Hair, 1985), were also administered in several major rest areas along two North–South highways in Taiwan.
Participants were first asked whether they had installed an OBU for using ETC service. If so, after briefly
elucidating our research purpose, they were invited to participate and complete the survey questionnaire. A
total of 300 questionnaires were distributed and a total of 264 completed questionnaires were returned. Since
9 questionnaires were invalid, 255 responses were obtained and valid (85.00% response rate). Specific demographic information is shown in Table 1.
Table 1
Demographic profile of the respondents (N = 255)
Demographic variables
Frequency
Percentage
Gender
Male
Female
197
58
77.25
22.75
Age
18–29
30–39
40–49
50–59
More than 59
20
113
93
27
2
7.84
44.31
36.47
10.59
0.78
Education
Elementary school or less
Junior high school
High school
University
Postgraduate degree
1
5
28
130
91
0.39
1.96
10.98
50.98
35.69
1
35
70
71
65
13
0.39
13.73
27.45
27.84
25.49
5.10
Frequencies of passing toll plazas
Ten times or above per day
One time or above per day
Several times per week
Several times per month
One time for several months
Few/barely
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
305
As shown in Table 1, 77.25% are males and the majority of respondents (80.78%) are in the age group of
30–39 and 40–49 years. In addition, most respondents are highly educated with 86.67% of them having
attained university diplomas or postgraduate degrees. Moreover, the frequency of passing toll plazas show
that 41.57% of our respondents are frequent highway motorists. Despite passing through toll plazas quite
often (10 times per day, one time or above per day, and several times per week), they have still not installed
OBUs.
3.2. Measurement development
The operationalization, sources, and standardized loadings of measurement items are shown in Appendix
A. To test the framework, we paid particular attention to issues of operationalization and measurement in this
study, following Venkatraman and Grant (1986). We operationalized the variables in two ways: (1) for those
variables that have been previously employed in a research setting, we adopted the measures with acceptable
measurement quality; and (2) for those variables that were unique to our conceptual model, we conducted
belief elicitation interviews with motorists and developed the operational measures. In addition, respecting
Nunnally’s (1978) recommended procedures, we developed multiple items for each construct when possible.
The preliminary instrument was pilot tested and reviewed by motorists for clarity again, thereby confirming
the content validity of our instrument. All items were five-point, Likert-type scales anchored at ‘‘strongly disagree’’ (1), ‘‘strongly agree’’ (5), and ‘‘neither agree nor disagree’’ (3). The scales of the intention of ETC service adoption are borrowed from Yu et al. (2005). For measuring attitude, subjective norm and perceived
behavioral control, the items are adopted from Taylor and Todd (1995). The perceived usefulness scales
are adopted from Moore and Benbasat (1991). Finally, as for perceived ease of use, we adopted this from
Bhattacherjee (2000).
4. Data analysis and results
4.1. Convergent validity and discriminant validity
In analyzing the collected data, this study followed the two-step procedure suggested by Anderson and Gerbing (1988). We estimated and respecified the measurement model prior to incorporating the structural restrictions. Convergent and discriminant validity of the remaining items and scales were tested with confirmatory
factory analysis (CFA) using the LISREL 8.50 program (Bentler and Chou, 1987). The result of the CFA
indicated that the measurement model provided a very good fit to the data: x2(131) = 328.37, Bentler
Bonett Normed Fit Index (NFI) = 0.95, Non-Normed Fit Index (NNFI) = 0.96, Comparative Fit Index
(CFI) = 0.97, Goodness-of-Fit Index (GFI) = 0.88, and Root Mean Square Error of Approximation
(RMSEA) = 0.077.
Convergent validity was assessed based on the criteria that the indicator’s estimated pattern coefficient was
significant on its posited underlying construct factor. We evaluated for the measurement scales using the three
criteria suggested by Fornell and Larcker (1981):
(1) All indicator factor loadings (k) should be significant and exceed 0.5.
(2) Construct reliabilities should exceed 0.8.
(3) Average variance extracted (AVE) by each construct should exceed the variance due to measurement
error for that construct (e.g., AVE should exceed 0.5).
All k values in the CFA model exceeded 0.5 and were significant at p = 0.001 except two indicators for perceived usefulness which were dropped due to their cross-loadings. We dropped these items from further analysis. Composite reliabilities of constructs ranged from 0.87 to 0.97 (see Table 2). AVE, ranging from 0.63 to
0.95, was greater than the variance due to measurement error. Therefore, all three conditions for convergent
validity were met.
Before examining the discriminant validity, however, we found that some constructs – attitude, perceived
behavioral control and perceived usefulness – may cause a multicollinearity problem because of their high
306
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
Table 2
Reliability, correlation coefficients and AVE results
Variable
Mean
SD
Cronbach’s alpha
Composite reliability
AVE
(1)
(2)
(3)
(4)
(5)
(6)
ATT
SN
PBC
INTENT
PU
PEOU
12.81
4.94
12.58
5.05
15.95
13.39
4.37
2.41
4.04
2.48
5.37
4.34
0.94
0.97
0.89
0.98
0.93
0.96
0.93
0.96
0.87
0.97
0.91
0.95
0.77
0.93
0.63
0.95
0.77
0.83
0.88
0.69
0.74
0.69
0.72
0.62
0.97
0.62
0.67
0.64
0.49
0.79
0.68
0.61
0.76
0.98
0.63
0.57
0.88
0.55
0.91
Notes:
1. The main diagonal shows the square root of the AVE (average variance extracted).
2. ATT for attitude, SN for subjective norm, PBC for perceived behavioral control, INTENT for intention of ETC service adoption, PU
for perceived usefulness, and PEOU for perceived ease of use.
3. Significant at p < .01 level is shown in bold.
correlation. Multicollinearity may lead to artifactual outcomes because of the effects of two or more predictors
being substantially confounded (Miles and Shevlin, 2001). Multicollinearity is indexed by the variance inflation factor (VIF) statistic; there has been little agreement in the literature, however, as to a suitable VIF cutoff
point at which the problem should give cause for concern. Recommended VIF criteria for problematical levels
of multicollinearity encompass values of 2.50 (Allison, 1999), 4.00 (Miles and Shevlin, 2001), and 10.00 (Everitt, 1996). We followed the procedure suggested by Miles and Shevlin (2001) and the VIF values of these constructs are all less than 3.4, showing there is no serious multicollinearity problem.
Finally, discriminant validity was shown when the square root of each construct’s AVE is larger than its
correlations with other constructs (Chin, 1998). As illustrated in Table 2, the square root of the AVE is larger
than all other cross-correlations. Hence the latter test of discriminant validity was also met. In addition to reliability coefficients and AVE values, Table 2 also reports the correlation matrix, means, and standard deviations of the study’s principal constructs.
4.2. Hypothesis testing
The research model presented earlier was tested using the structural equation modeling (SEM) approach.
Overall, the goodness-of-fit of the structural model was comparable to that of the previous CFA model and
provided evidence of adequate fit. As shown in Fig. 2, attitude (b = 0.22, p < 0.05) is positively associated with
the intention of ETC service adoption, rendering support for H1. In addition, both subjective norm (b = 0.23,
p < 0.001) and perceived behavioral control (b = 0.36, p < 0.001) were also significant predictors of the intention of ETC service adoption, thereby supporting H2 and H3. Furthermore, among these three building
blocks of TPB, perceived behavioral control has the strongest effect on the intention of ETC service adoption.
The impact of perceived usefulness (b = 0.58) and perceived ease of use (b = 0.38) on attitude are significant
at p = 0.001. Therefore, H4 and H6 can be accepted. This shows that both system attributes – perceived usefulness and perceived ease of use – are important antecedents in forming the motorist’s attitude towards ETC
service adoption. In other words, an ETC system service that is not advantageous for motorists and does not
help them perform their toll collection is not likely to be received favorably nor will a positive attitude be
formed towards it in spite of careful implementation efforts.
Meanwhile, it shows that perceived usefulness has no significant impact on the intention of ETC service
adoption, and thus H5 cannot be accepted. For the result in perceived usefulness, previous empirical studies
on TAM have shown inconsistency with either significant influence (Moore and Benbasat, 1991; Chau, 1996)
or with insignificant influence on behavioral intention to use (Chen et al., 2002). It might imply an indirect
influence of perceived usefulness on the intention of ETC service adoption via the mediator (e.g., attitude)
towards ETC service adoption in our model.
Moreover, perceived ease of use has a strong effect on perceived usefulness (b = 0.61, p < 0.001), validating
H7 and allowing the inference that perceived ease of use fosters a motorist’s perceived usefulness towards ETC
system service. We will discuss these findings in detail in the next section.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
307
Fig. 2. Data analysis results.
5. Discussion
This study aims to shed light on the critical antecedents of motorists’ intentions of ETC service adoption by
integrating both TAM and TPB perspectives. The overall explanatory power of our research model was relatively high; a R-square of 61% for the intention of ETC service adoption, a R-square of 74% for attitude
towards ETC service adoption, and a R-square of 38% for perceived usefulness were obtained, suggesting that
both theories are capable of explaining a relatively high proportion of the variation in the intention to adopt
ETC service.
Several insightful results could be summarized from our research framework as follows. First, motorist’s
perceived behavioral control was the most important predictor of the intention of ETC service adoption. This
finding is consistent with Mathieson’s (1991) identification of perceived behavioral control as an important
determinant to behavioral intention. Our result also reinforces that the inclusion of perceived behavioral control significantly improved the prediction of intentions (Ajzen, 1991). Following TPB, therefore, it is expected
that motorist’s perceived constraints of carrying out the transaction of ETC affects whether or not that behavior will be carried out. For example, higher price and the installation fee of the OBU might increase the degree
of perceived external constraints of a motorist, thereby decreasing the willingness to adopt ETC service.
Second, subjective norm suggests that behavior is instigated by one’s desire to act as how others act or think
one should act, and it appeared to be the second most important determinant of a motorist’s intention to
adopt ETC service. Our finding is similar to those reported by Bhattacherjee (2000) and Karahanna et al.
(1999) which indicated that subjective norm influences behavioral intention towards system use and subjective
norm could influence intention to use as strongly as attitude does. Although some prior studies (e.g., Venkatesh and Davis, 2000; Mathieson, 1991; Davis et al., 1989) found that subjective norm could significantly
determine intention to use in a mandatory-usage context, its impact would be less significant where users were
in a voluntary-usage context as with the case of ETC service adoption in this study. However, subjective norm
has been found to be more important in the early stages of system development and was a more important
predictor of intention for people without prior experience (Taylor and Todd, 1995). In particular, as ETC service adoption in this study is at the introduction stage, there is a lack of references from prior adopters such as
friends, peers and superiors. Furthermore, after the formal launch of the ETC system in Taiwan, disputes,
public complaints, and allegations of impropriety have been rife (e.g., perceived social pressure from TV
308
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
news). Accordingly, it is reasonable to expect that the effect of subjective norm on the intention of ETC service
adoption should indicate significance and have as strong an effect as attitude does.
Third, our finding supports TPB and TAM, which both predict that attitude towards behavior is a significant determinant of behavioral intention. Attitude has long been shown to influence behavioral intentions
(Ajzen and Fishbein, 1980). This relationship has received substantial empirical support. For our research
context, the motorists’ attitudes could strongly determine their willingness to use OBU for ETC service.
Fourth, we also confirmed that both perceived usefulness and perceived ease of use determine attitude
towards ETC service adoption. The results reveal that motorists’ perceptions of the ease and effectiveness
of ETC service may indirectly lead to their intention to adopt ETC service via their attitude. In addition,
the effect of perceived usefulness on attitude is greater than that of perceived ease of use (0.58 > 0.38). As
Keil et al. (1995) concluded, in determining whether or not to use a technology, usefulness is more important
than ease of use. In our context, an essential factor for a motorist’s acceptance of ETC service is whether the
technology fills his or her service needs. In other words, ETC is acceptable is only when demonstrating proven or desired utility in its practices. Moreover, the significant positive effect of perceived ease of use on perceived usefulness is also confirmed, and this is consistent with the results from many prior studies that
examined TAM.
Lastly, perceived usefulness appeared to have no significant effect on the intention of ETC service adoption.
A plausible explanation is that perceived usefulness might often indicate its influence on the intention through
the mediator of attitude. Moreover, the context of this study is focused on the stage of initial adoption and
voluntary use of ETC service. Motorists’ positive perceived usefulness in using ETC service may not immediately lead to a behavioral intention to use, but rather to firstly form a favorable attitude toward using ETC
service. In other words, potential users such as motorists would need to take a period of time to carefully
change their psychological state in order to adopt ETC service.
6. Conclusions
Given the turbulence of many industries, understanding what facilitates the delivery of products and services to satisfy customers’ needs offers scholars and practitioners continuously and increasingly important
challenges. By integrating two important intention-based theoretical perspectives, TAM and TPB, we expect
to provide meaningful insights in a more comprehensive manner that jointly predicts user acceptance of ETC
service. This could be especially valuable for government policymakers and BOT vendors for their policy
development and marketing practices.
However, we also acknowledge that a number of research limitations exist in our research which should be
overcome in the future. First, the conclusions drawn from our study are based on cross-sectional data. As suggested by Ajzen (1991), the relative importance of attitude, subjective norm, and perceived behavioral control
in predicting behavioral intention may vary across behavior and situation. With our cross-sectional data, we
only took a snapshot of this model. A stricter test of our argument, however, would be to use a longitudinal
study to evaluate this aspect more critically since the implementation of ETC service is only just beginning in
Taiwan. By using a longitudinal study in the future, we could investigate our research model in different time
periods and make comparisons, thus providing more insights into ETC adoption phenomenon and more contributions for both TAM and TPB perspectives.
Second, although our model provides some insights to explain the intention of ETC service adoption, some
possible moderating effects between TPB variables (e.g., attitude, subjective norm, and perceived behavioral
control) and the intention of ETC service adoption are not well understood. Future studies may benefit from
articulating the possible moderating factor (e.g., government’s procedural justice) that enhance or impede such
ETC adoption intention that is most compatible with such purposes. In sum, these questions open up fertile
grounds for future research opportunities.
These findings might have policy implications for the implementation of ETC and e-government related
services. Our research results revealed that attitude, subjective norm, and perceived behavioral control could
positively influence motorist’s intention of ETC service adoption. Moreover, the impact of perceived behavioral control has the strongest effect on adoption. Accordingly, this highlights the need for several strategic
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
309
plans to be carried out synchronously for gaining motorist’s trust and engendering the intention of ETC service adoption as soon as possible.
First, as discussed earlier, perceived behavioral control refers to perceptions of internal and external constraints (e.g., skills, information searching, available budget, time and places for OBU installation) on behavior (Ajzen, 1987, 1991). Likewise, we suggest that government policymakers and BOT vendors could lower the
OBU price and establish more service locations for OBU installation, thereby increasing the convenience and
willingness to use ETC service and lessen motorists’ possible external constraints. Next, advertising and marketing plans containing ETC system performance reports or adopters’ successful experience toward ETC service might also serve as possible and effective ways to attract potential or non-adopters.
To conclude, the results of this study demonstrate that while TAM by itself is useful in predicting ETC service adoption, integration with TPB could provide a more complete understanding of behavioral intention.
This should help to better manage the system implementation process by focusing attention on several control
factors such as system attributes, individual preference, social influences, and individual constraints in the
ETC service and related e-government issues.
Appendix A. Measurements
Constructs
Standardized
loadings
Attitude (Taylor and Todd, 1995)
• Using the ETC service is a good idea
• Using the ETC service is a wise idea
• I like the idea of using the ETC service
• Using the ETC service would be pleasant
0.9
0.86
0.93
0.81
Subjective norm (Taylor and Todd, 1995)
• People who influence my behavior would think that I should use the ETC service
• People who are important to me would think that I should use the ETC service
0.96
0.97
Percevied behavioral control (Taylor and Todd, 1995)
• It is not troublesome for me to replenish the stored value of smartcard of OBU
0.85
• I have available time to install the related equipment of ETC system (e.g., OBU)
0.84
• I have the resources and the knowledge and the ability to replenish the stored value of 0.8
smartcard of OBU
• I can afford the application fee of ETC system service
0.67
Perceived usefulness (Moore and Benbasat, 1991)
• Using the ETC service enables me to pass toll plazas more quickly
• Using the ETC service improves the efficiency of using highways
• Using the ETC service makes it easier to pay toll
• Using the ETC service enables me to reach my destination more quickly
• Using the ETC service makes me effectively control my travel time on highways
Dropped
0.92
0.79
0.92
Dropped
Perceived ease of use (Bhattacherjee, 2000)
• It is easy for me to learn how to replenish the stored value of smartcard of OBU
• It is easy for me to use the function of smartcard balance replenishment
• It is easy for me to be accustomed to use ETC service
• ETC system is easy to use for me
0.86
0.87
0.96
0.95
Intention of etc system adoption (Yu et al., 2005)
• I will intend to apply the ETC service as soon as possible
• I will use the ETC service soon after it is launched
0.98
0.97
310
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
References
Ajzen, I., 1987. Attitudes, traits, and actions: dispositional prediction of behavior in social psychology. Advances in Experimental Social
Psychology 20, 1–63.
Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50 (2), 179–211.
Ajzen, I., Fishbein, M., 1980. Understanding Attitudes and Predicting Social Behavior. Prentice-Hall Inc., Englewood Cliffs, NJ.
Ajzen, I., Madden, T.J., 1986. Prediction of goal directed behavior: attitudes, intentions and perceived behavioral control. Journal of
Experimental Social Psychology 22, 453–474.
Allison, P.D., 1999. Multiple Regression: A Primer. Pine Forge Press, Thousand Oaks, CA.
Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a review and recommended two-step approach.
Psychological Bulletin 103 (3), 411–423.
Armitage, C.J., Conner, M., 2001. Efficacy of the theory of planned behavior: a meta-analytic review. British Journal of Social Psychology
40 (3), 471–499.
Bentler, P.M., Chou, C.P., 1987. Practical issues in structural modeling. Sociological Methods and Research 16 (1), 78–117.
Bernadette, S., 1996. Empirical evaluation of the revised technology acceptance model. Management Science 42 (1), 85–93.
Bhattacherjee, A., 2000. Acceptance of e-commerce services: the case of electronic brokerages. IEEE Transaction on Systems, Man, and
Cybernetics – Part A: Systems and Humans 30 (4), 411–420.
Bush, A.J., Hair, J.F., 1985. An assessment of the mall intercept as a data collection method. Journal of Marketing Research 22 (2), 158–
167.
Chau, Y.K., 1996. An empirical assessment of a modified technology acceptance model. Journal of Management Information Systems 13
(2), 185–204.
Chau, P.Y.K., Hu, P.J.H., 2002. Investigating healthcare professional’s decisions to accept telemedicine technology: an empirical test of
competing theories. Information and Management 39, 297–311.
Chen, L.D., Gillenson, M.L., Sherrell, D.L., 2002. Enticing online consumers: an extended technology acceptance perspective.
Information and Management 39, 705–719.
Chin, W.W., 1998. Issues and opinion on structural equation modeling. MIS Quarterly 22 (1), 7–16.
Davis, F.D., 1989. Perceived usefulness, perceived ease of use and user acceptance of information technology. MIS Quarterly 13 (3), 319–
340.
Davis, F.D., 1993. User acceptance of information technology: system characteristics, user perceptions and behavioral impacts.
International Journal of Man Machine Studies 38 (3), 475–487.
Davis, F.D., Bagozzi, R.P., Warshaw, P.R., 1989. User acceptance of computer technology: a comparison of two theoretical models.
Management Science 35 (8), 982–1003.
Decorla-Souza, P., Kane, A.R., 1992. Peak period tolls: precepts and prospects. Transportation 19 (4), 293–311.
Dishaw, M.T., Strong, D.M., 1999. Extending the technology acceptance model with task-technology fit constructs. Information and
Management 36 (1), 9–21.
Everitt, B.S., 1996. Making Sense of Statistics in Psychology. Oxford University Press, Oxford, England.
Fishbein, M., Ajzen, I., 1975. Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addision-Wesley,
Reading, MA.
Fornell, C.R., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of
Marketing Research 18 (1), 39–50.
Fu, J.R., Farn, C.K., Chao, W.P., 2006. Acceptance of electronic tax filing: a study of taxpayer intentions. Information and Management
43 (1), 109–126.
Golob, T.F., Regan, A.C., 2001. Impacts of information technology on personal travel and commercial vehicle operations: research
challenges and opportunities. Transportation Research Part C. Emerging Technologies 9 (2), 87–121.
Karahanna, E., Straub, D.W., Chervany, N.L., 1999. Information technology adoption across time: a cross-sectional comparison of preadoption and post-adoption beliefs. MIS Quarterly 23 (2), 183–213.
Keil, M., Beranek, P.M., Konsynski, B.R., 1995. Usefulness and ease of use: field study evidence regarding task considerations. Decision
Support Systems 13 (1), 75–91.
Kwon, T.H., Zmud, R.W., 1987. Unifying the fragmented models of information systems implementation. In: Boland, R.J., Hirschheim,
R.A. (Eds.), Critical Issues in Information Systems Research. Wiley, New York, pp. 27–251.
Mathieson, K., 1991. Predicting user intentions: comparing the technology acceptance model with the theory of planned behavior.
Information Systems Research 2 (3), 173–191.
Miles, J.N.V., Shevlin, M.E., 2001. Applying Regression and Correlation: A Guide for Students and Researchers. Sage, London.
Moore, G.C., Benbasat, I., 1991. Development of an instrument to measure the perceptions of adoption an information technology
innovation. Information Systems Research 2 (3), 192–222.
Nunnally, J.C., 1978. Psychometric Theory, second ed. McGraw-Hill, New York.
Pavlou, P.A., 2003. Consumer acceptance of electronic commerce: integrating trust and risk with the technology acceptance model.
International Journal of Electronic Commerce 7 (3), 101–134.
Rogers, E.M., 1995. The Diffusion of Innovation, fourth ed. Free Press, New York.
Taylor, S., Todd, P.A., 1995. Understanding information technology usage: a test of competing models. Information Systems Research 6
(2), 144–176.
C.-D. Chen et al. / Transportation Research Part C 15 (2007) 300–311
311
Venkatesh, V., Davis, F.D., 2000. A theoretical extension of the technology acceptance model: four longitudinal field studies. Management
Science 46 (2), 186–204.
Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of information technology: toward a unified view. MIS
Quarterly 27 (3), 425–478.
Venkatraman, N., Grant, J.H., 1986. Construct measurement in strategy research: a critique and proposal. Academy Management Review
11 (1), 71–86.
Yu, J., Ha, I., Choi, M., Rho, J., 2005. Extending the TAM for a t-commerce. Information and Management 42 (7), 965–976.
