Risk-Focused e-Commerce Adoption Model

- A Cross-Country Study

Joongho Ahn
College of Business Administration
Seoul National University
Seoul, 151-742, Korea

Jinsoo Park, Dongwon Lee

Information and Decision Sciences Department
Carlson School of Management
University of Minnesota
Minneapolis, MN 55455

(Contact Author)

Jinsoo Park
3-356, Information and Decision Sciences Department
Carlson School of Management
University of Minnesota
321-19th Avenue South
Minneapolis, MN 55455
Phone: (612) 624-1301
E-mail: park@umn.edu

Working Paper
Last revised on June 2001

Risk-Focused e-Commerce Adoption Model

- A Cross-Country Study

Abstract
While e-Commerce has proliferated with the growth of the Internet, there have been insufficient research efforts concerning its
status in Korea. The United States, in contrast, has made significant efforts in making empirical research on the consumers
adoption of e-Commerce. This paper validates the e-Commerce Adoption Model (e-CAM) on the two countries. e-CAM
integrates the technology acceptance model with the theories of perceived risk to explain the e-Commerce adoption. The study
findings not only provide interim support for the generalizability of e-CAM, but also suggest that online firms should consider
these contextual factors in order to facilitate consumers adoption behavior.

Keywords: e-Commerce, Technology Acceptance Model, Perceived Risk, Perceived Ease of Use, Perceived Usefulness, CrossCountry Study.

1. Introduction
The Internet has grown at a remarkable pace since the emergence of the World-Wide Web (WWW) in the
early 1990s (U.S. Department of Commerce, 2000; USIC and IITA, 2000). The unprecedented growth of
Internet technology has made numerous resources on the Web instantly accessible to various user
communities. The business community has also developed innovative strategies to expand their customer
base using this technology (Ram et al., 1999). Further, the explosive increase of Internet users has led to
dramatic shifts in the way of conducting business. From our daily lives to commercial transactions
between businesses, the Internet has profoundly impacted and changed the way we do business. Electronic
Commerce (e-Commerce) presents enormous opportunities for both consumers and businesses in the world.
For example, online firms, which implement e-Commerce, deliver their products/services to markets
through their ability to organize and maintain a business network, while traditional business models stress
the ability to manufacture products and deliver services (Zwass, 1996). In addition, wide choice ranges,
lower prices, and entirely new products have become available in many product categories, such as books,
CDs, and travel packages, to consumers across regions and national frontiers (ISPO, 1997; Jarvenpaa et al.,
2000). Furthermore, many small businesses today are using the Internet to attract new customers, build
relationships with suppliers, and cut the costs of serving established clients (U.S. Department of Commerce,
1999). Therefore, the Internet and e-Commerce have the potential to drive world economic growth for
many decades.
According to Nua Internet Surveys, Internet users have increased in all regions of the world from 171
million in 1999 to 304 million in March 2000, an increase of 78 percent (U.S. Department of Commerce,
2000; USIC and IITA, 2000). In addition, according to the U.S. Bureau of the Census (2000), the estimate
of U.S. retail e-Commerce sales for the second quarter in 2000 totaled $5.52 billion, an increase of 6.16
percent from the fourth quarter in 1999. Furthermore, recent forecasts show that Internet users could
exceed 1 billion by 2005, with 700 million located outside North America (U.S. Department of Commerce,
2000).

-1-

A vigorous private sector in developing countries, as in the United States, has the ability to bring the
benefits of the information revolution to consumers. Experience in many developing countries
demonstrates, however, that without the aid of policy reform and some initial public investment, private
sector action alone may not be able to provide these benefits. With the on-going efforts to build a global
information infrastructure as well as a nation-wide information infrastructure, Korea is also at the
beginning of turbulent shifts to the Information Age. At the end of 1994, however, the Korean government
set up a plan to construct a nation-wide information superhighway by the year 2015. This plan can be
compared to the NII project in the U.S. and the IT 2000 project in Singapore. After the announcement of
this plan, a committee was organized to launch sub-projects that implement the architectural model of the
national information infrastructure. These projects provide an alternative source that will increase the
number of users connected to the communications network.
While e-Commerce has become an important issue with the growth of the Internet, there have been
insufficient empirical research efforts concerning its status and consumer behavior over the Internet. We
believe that there may be some valid factors to explain the marked differences between the consumers
adoption of e-Commerce in developed countries (e.g., the U.S.) and developing countries (e.g., Korea).
In this study, we validate the e-Commerce Adoption Model (e-CAM), which is derived from the
theoretical foundations of prior research in the theories of perceived risk as well as the technology
acceptance model (TAM). Specifically, we examine the impact of the following factors on the consumers
purchasing behavior: perceived ease of use (PEU), perceived usefulness (PU), perceived risk with
products/services (PRP), and perceived risk in the context of online transaction (PRT). We demonstrate not
only what contextual constructs make a consumer adopt or reject e-Commerce as a purchasing vehicle of
products/services, but also how these contextual differences influence the consumers adoption behavior
between the two countries.
The remainder of this paper is organized into the following five sections. The first section provides a
brief review of the literature on the technology acceptance model and the theories of the consumers
perceived risk. Next, we present our research model (e-CAM) and a set of research hypotheses based on
the theories in the preceding section. Then, we discuss our research methods used to test the proposed

-2-

model and present the analysis and results of our study. Finally, we make conclusion by discussing the
implications of our study, followed by presenting limitations and future research direction.

2. Theoretical Background
In this section, we review the previous research efforts on the technology acceptance model (TAM) and the
theories of the consumers perceived risk, which are used as theoretical backgrounds of the e-Commerce
adoption model (e-CAM).

2.1.

Technology Acceptance Model

Information Systems (IS) researchers have made significant efforts in building theories to examine and
predict the determinant factors of information technology (IT) acceptance (Agarwal and Prasad, 1998;
Agarwal and Prasad, 1999). Existing models of IT acceptance have their foundations from several diverse
theories, most noticeably innovation diffusion theory, where individuals perceptions about using an
innovation are considered to affect their adoption behavior (Agarwal and Prasad, 1998; Moore and
Benbasat, 1991; Rogers, 1995). Other important theoretical models that attempt to explain the relationship
between user beliefs, attitudes, intentions, and actual system use include the theory of reasoned action
(TRA) (Ajzen and Fishbein, 1980), the theory of planned behavior (TPB) (Ajzen, 1991), and the
technology acceptance model (TAM) (Davis, 1989; Davis et al., 1989). Among these theories, TAM seems
the most widely accepted among IS researchers due to the richness of recent empirical support (Agarwal
and Prasad, 1997; Morris and Dillon, 1997).
According to the TRA model, beliefs influence attitudes, which consecutively lead to intentions, then
direct or make behaviors. The TAM model, originally developed by Davis from the theoretical foundation
of TRA, adapts this belief-attitude-intention-behavior relationship to an IT user acceptance. Thus, the
purpose of TAM is to explain and predict IT acceptance and facilitate design changes before users have
experience with a system (Davis, 1989).

-- Insert Figure 1 here --

-3-

As illustrated in Figure 1, TAM predicts user acceptance based on two specific behavioral beliefs:
perceived ease of use (PEU) and perceived usefulness (PU), which determine an individuals behavior
intention (BI) to use an information technology (Davis et al., 1989). In addition, the effects of external
variables on behavioral intention are mediated by these two factors.
Significant empirical researches, as presented in Table 1, have examined the TAMs overall
explanatory power and measurement validity in different settings characterized by constructs, type of IS,
etc. Originally investigating e-mail, word processing and graphics software (Davis, 1989; Davis et al.,
1989), TAM has been extended its application to diverse types of IS, such as spreadsheets (Doll et al.,
1998; Hendrickson et al., 1993; Mathieson, 1991; Venkatesh and Davis, 1996), voice mail (Adams et al.,
1992; Chin and Todd, 1995; Segars and Grover, 1993; Straub et al., 1995; Subramanian, 1994), DBMS
(Doll et al., 1998; Hendrickson et al., 1993; Szajna, 1994; Szajna, 1996), personal computing (Agarwal and
Prasad, 1999; Igbaria et al., 1997), telemedicine (Hu et al., 1999), expert system (Gefen and Keil, 1998),
and some other software (Venkatesh, 1999; Venkatesh and Morris, 2000; Venkatesh and Davis, 2000).
Furthermore, several recent studies (Fenech, 1998; Lederer et al., 2000; Lin and Lu, 2000; Morris and
Dillon, 1997; Teo et al., 1999) have examined TAM to analyze users behavior on the Internet, specifically
the WWW.

--Insert Table 1 here --

Based on empirical evidence, the attitude construct (A) was left out from the original TAM model
(Davis, 1989; Davis et al., 1989) because it did not fully mediate the effect of PU on behavioral intention
(BI) (Venkatesh, 1999). In addition, several studies (Adams et al., 1992; Fenech, 1998; Gefen and Straub,
1997; Gefen and Keil, 1998; Igbaria et al., 1997; Karahanna and Straub, 1999; Lederer et al., 2000;
Mathieson, 1991; Straub et al., 1995; Teo et al., 1999; Venkatesh and Morris, 2000) have disregarded the
effect of PEU/PU on the attitude (A) and/or BI. Instead, they focus on the impact of PEU and/or PU

-4-

directly on the actual system usage. As our research focuses on consumers actual usage on e-Commerce,
we adapted the TAM model by dropping some constructs (i.e., A and BI), as illustrated in Figure 2.

-- Insert Figure 2 here --

Consumers can access thousands of online sites and purchase anything from groceries to books and
insurance policies without traveling to a store site or adjusting his schedules around the stores hours.
Recognizing that customers may want products/services delivered as soon as possible, many e-Commerce
sites offer next-day or second-day delivery. Furthermore, e-Commerce consumers can view catalogs of
different products/services and read extensive information detailing their features and performance while
information acquisition was time-consuming and difficult prior to the outset of the Internet. Therefore, we
recognize ease of information search, ease of ordering (any time, any location), ease of using customer
service, and overall ease of use as consumers perceived ease of use (PEU). In addition, we measure
perceived usefulness (PU) by the following factors: saving of money, saving of time, vast selection of
products/services, and overall usefulness.

2.2.

Theories of Perceived Risk

Since Bauer (1960) first formally proposed that consumer behavior be seen as risk taking,
valuable empirical research have attempted to identify various types of perceived risk in the context of the
consumers purchase behavior.

2.2.1.

Perceived Risk with Product/Service

Bauer (1960) mentioned that the belief of perceived risk as a key determinant to consumer behavior might
be a primary factor influencing the conversion of browsers to real buyers. Cox and Rich (1964) refer to
perceived risk as the overall amount of uncertainty perceived by a consumer in a particular purchase
situation. Cunningham (1967) recognized the risk resulting from poor performance, danger, health hazards,
and costs. Roselius (1971) identified four types of losses that related to the types of risk: time, hazard, ego,
and money. Jacoby and Kaplan (1972) classified consumers perceived risk into the following five types of

-5-

risk: physical, psychological, social, financial, and performance (functional) as listed in Table 2. Taylor
(1974) suggests that the uncertainty or perception of risk may generate anxiety that influences the consumer
decision-making process. Murphy and Enis (1986) define perceived risk as customers subjective
assessment of the consequence of making a purchasing mistake.

-- Insert Table 2 here --

As we cannot directly see or touch product/service in the electronic market (i.e., intangibility
characteristic), consumers may feel anxiety or uncertainty when they have transactions with online vendors.
For example, product/service delivered to consumers may not perform as expected. In addition, consumers
may be also required to bear the expenses such as shipping and handling, when returning or exchanging the
product/service. Among the five types of risk that Jacoby and Kaplan (1972) propose (refer to Table 2),
therefore, we recognized functional loss and financial loss as risk types related to product/service to
discourage consumers from doing online transactions. Further, when purchased products/services fail, we
may waste time, convenience, and effort getting it adjusted or replaced. Although time is non-monetary
effort and varies among individuals, we recognize time as a cost that consumers must pay for
products/services. Thus, we identified time loss as an additional risk with the product/service. After
purchasing product/service over the Internet, consumers may find a product/service of equal or higher
quality at a lower price. Hence we recognized another perceived risk, opportunity loss, which is the risk
that by taking one action a consumer will miss out on doing something else he/she would really prefer to
do.
Therefore, we define perceived risk with product/service (PRP) as the overall amount of uncertainty
or anxiety perceived by a consumer in a particular product/service when the consumer purchase online.
Finally, we identify five types of PRP as follows: functional loss, financial loss, time loss, opportunity loss,
and overall perceived risk with product/service.

-6-

2.2.2.

Perceived Risk in the Context of Online Transaction

Several research on the context of online transaction (Hoffman et al., 1999; Jarvenpaa and Tractinsky,
1999; Jarvenpaa et al., 2000; Ratnasingham, 1998; Swaminathan et al., 1999) suggest that consumers
confidence or trust will be improved by increasing the transparency of the transaction process (for example,
fully disclosing the identity, origin, and liability of the supplier), keeping to a minimum the personal data
required from the consumer, and by making clear the legal status of any information provided. Bhimani
(1996) points out the threats to the adoption of e-Commerce that could manifest from such illegal activities
as eavesdropping, password sniffing, data modification, spoofing, and repudiation. Therefore, Bhimani
(1996) and Ratnasingham (1998) suggest the fundamental requirements for e-Commerce that satisfies the
following security issues: authentication, authorization, availability, confidentiality, data integrity,
nonrepudiation, and selective application services. Swaminathan et al. (1999) assert that consumers
evaluate online vendors before they enter into online transaction and therefore the characteristics of the
vendors play an important role in facilitating the transaction. Rose et al. (1999) identify the technical
impediments and their associated costs and limitations specific to B2C e-Commerce, which include
download delays, limitations of the interface, search problems, inadequate measurement of Web application
success, security weakness, and lack of Internet standards. Further, they state that if people do their
business transactions with dishonest merchants or if sensitive information is stored on unsecured databases,
security threats exist even where data is perfectly secure in transmission.
Therefore, we define perceived risk in the context of online transaction (PRT) as a possible
transaction risk that consumers can face when exposed to electronic means of doing commerce. Finally,
four types of PRT are identified as follows: privacy, security (authentication), nonrepudiation, and overall
perceived risk on online transaction.

Based on previous research on the consumers perceived risk, we proposed a theoretical model that
postulates perceived risks as the antecedents to the adoption of e-Commerce as represented in Figure 3.

-- Insert Figure 3 here --

-7-

In summary, our research model, which will be described in the following section, derives its
theoretical foundations from prior research in the theories of the consumers perceived risk as well as
TAM.

3. Research Framework
3.1.

Research Hypotheses

Many factors positively or negatively influence a consumers decision to adopt e-Commerce as a

purchasing means of products/services. In this section, we propose several hypotheses regarding
consumers adoption of e-Commerce based on each construct that was derived from the pervious literature.

3.1.1.

Perceived Risk in the Context of Online Transaction

It is important that consumers feel confidence in the e-Commerce process itself. As suggested in numerous
studies (Bhimani, 1996; Hoffman et al., 1999; Ratnasingham, 1998; Swaminathan et al., 1999; Tan and Teo,
2000), a commonly recognized barrier to the diffusion and adoption of e-Commerce has been the lack of
security and privacy over the Internet. Messages on the Internet are being passed in a shared domain, and
therefore consumers are not yet comfortable with sending personal information across the Internet (Rose et
al., 1999). Most online vendors allow consumers to pay through credit card, which effectively limits the
number of consumers immediately. Security concerns with respect to exposure of credit card information
to hackers or unknown vendors is still a major anxiety for consumers (Sindhav and Balazs, 1999;
Swaminathan et al., 1999). According to the study of Hoffman et al. (1999), 95% of Web users have
declined to provide personal information to Web sites at one time or another when asked, and 40% who
have provided demographic data have gone to the trouble of fabricating it. Further, Bhimani (1996) states
that consumers may be afraid that online vendors can deny an agreement after the transaction.
Therefore, the higher concerns for privacy, security, and vendors trust a consumer perceives in the eCommerce transactions, the lower the usage of e-Commerce will be for the consumer. In this context, the

-8-

risk (PRT) would discourage consumers from their adoption on e-Commerce. Accordingly, we propose the
hypotheses with regard to the impact of PRT as follows:
H-1:

Perceived risk in the context of online transaction (PRT) negatively affects consumers purchasing
behavior in e-Commerce (PB).

3.1.2.

Perceived Risk with Product/Service

Online consumers may also feel anxiety when they do transactions on the Internet due to the
intangibility characteristics of the online products/services. The anxiety or uncertainty with product/service
drives consumers beliefs about the risk itself (PRP), and therefore this risk reduces the possibility that the
consumers obtain from shopping on the Internet. Finally, the risk (PRP) will put a damper on consumers
adoption of e-Commerce. Therefore, we propose the following hypothesis regarding PRP.
H-2:

Perceived risk with product/service (PRP) negatively affects consumers purchasing behavior in eCommerce (PB)
Further, a person will be generally expected to make a risky decision based on her own common

standard to any risks. Therefore, the following hypothesis is proposed to explain the relationship between
the risks.
H-3:

Perceived risk with product/service (PRP) is positively correlated with perceived risk in the context
of online transaction (PRT).

3.1.3.

Perceived Ease of Use

Information systems that users perceive easier to use and less complex will increase the likelihood of its
adoption and usage (Agarwal and Prasad, 1997; Teo et al., 1999). According to several researches on TAM
(Davis et al., 1989; Teo et al., 1999; Venkatesh and Morris, 2000), perceived ease of use (PEU) has been
shown to influence behavior (i.e., IT adoption) through two causal ways: (1) a direct effect on behavior and
(2) an indirect effect on behavior via perceived usefulness (PU). Therefore we propose the following
hypotheses.
H-4:

Perceived ease of use (PEU) is positively related to perceived usefulness (PU).

-9-

H-5:

Perceived ease of use (PEU) positively affects consumers purchasing behavior in e-Commerce
(PB).

3.1.4.

Perceived Usefulness

As represented in Table 1, valuable empirical studies have already validated the relationship between
perceived usefulness (PU) and user acceptance of information systems. It is expected that users will accept
information systems if they perceive the accepted systems would help them to attain desired performance.
In this context, we propose a hypothesis as follows.
H-6:

3.2.

Perceived usefulness (PU) positively affects consumers purchasing behavior in e-Commerce (PB).

Research Model (e-Commerce Adoption Model)

The research model to be empirically tested in the study is illustrated in Figure 4. Our model, called the eCommerce Adoption Model (e-CAM), is derived from the theories and hypotheses described in the
preceding sections. e-CAM suggests that PEU, PU, PRP, and PRT will have impact on the consumers
adoption of e-Commerce. The relationships constituting the model also have support from prior theoretical
and empirical work in the previous research.

-- Insert Figure 4 here --

4. Research Methods
4.1.

Questionnaires Generation

With regard to the independent variables, we derived questions to measure the PU and PEU constructs
based on Daviss TAM model and to measure the PRP and PRT constructs from the earlier research on
perceived risk. Measurement of the PEU construct stresses on how comfortable consumers will be with eCommerce as a purchasing medium as well as how easy it will be to use. For the measurement of the PU
construct, the influence of e-Commerce on consumers is emphasized. In order to measure the PRP

- 10 -

construct, we focus on the risk factors regarding product/service that discourage consumers from doing
online transaction. For the measurement of the PRT construct, transaction security and privacy on the
Internet are stressed.
The dependent variable of the study, the consumers online purchasing behavior (PB), was measured
in two ways as (i) the frequency of online purchases during the past six months and (ii) the total amount
spent on online purchases during the past six months. We referred to the previous research and survey on
the Internet usage, such as the GVU's 10th WWW User Surveys (1998), to develop the questions regarding
the dependent variable and other demographic variables.
The questionnaires were initially pretested with several graduate students who enrolled in a large
University and resulted in removing and rewording some unclear questions. Finally, the questionnaires
were mainly divided into five categories: e-Commerce Usage (3 items), PEU (4 items), PU (4 items), PRP
(5 items), and PRT (4 items). Each item of independent variables was assessed with a 7-point scale with
end points of strongly disagree and strongly agree. With regard to the dependent variables, the
frequency of online purchases was measured with 5-point scales ranging from none to more than 10
times, and the total amount spent on online purchases with 8-point scales ranging from none to more
than $2,000.

4.2.

Sample & Procedures

As Tan and Teo (2000) state, Web-based surveys are appropriate when the target subjects are Internet users
and a short time frame for responses is required. Thus, this study was carried out through the Web-based
survey methodology because our study mainly focuses on consumers who have ever experienced the
Internet. The questionnaires were posted on the Web using Java Servlet technology. In addition, we
utilized JavaScript language to check for missing responses and prompt users to answer them as suggested
in (Tan and Teo, 2000).
Subjects for the study were mostly undergraduate and graduate students at some major Universities in
the United States and Korea who enrolled in information system courses. Subjects received participation
credit, which fulfilled a course requirement. As an additional incentive for participation, we promised to
send an executive summary of the results to the respondents. Although numerous studies have used student
- 11 -

subject for developing and validating theories on IT diffusion and adoption, this subject may cause a
sampling bias (i.e., external validity problem). However, we believe that the students will finally grow to
be the most active Internet users and influential consumers in the online market, and therefore
understanding of potential consumers needs in e-Commerce is very imperative to predict future trends on
the adoption of e-Commerce.
The survey lasted about a month and elicited a total of 465 responses (USA: 183, Korea: 282).
Among these responses, 443 (USA: 176, Korea: 267) were valid for the analysis regardless of whether or
not the respondent has online purchasing experience.

4.3.

Data Analyses

4.3.1.

General Demographic Statistics

As shown in Table 3, gender composition of the two countries is almost similar to each other. This result is
also quite similar to that of the GVUs 10th WWW User Surveys (1998) (66.4%: 33.6%). The ages of the
overall majority of respondents are between 16 and 35 (USA: 89.2%, Korea: 96.7%). Further, all the
respondents in both countries are highly educated with at least some college experience, which is also
similar to the result of the GVUs surveys (87.8%).

-- Insert Table 3 here --

4.3.2.

Internet Usage Statistics

While 42% of the U.S. respondents have high connection speed faster than 56 kb/sec, 69% of the Korean
subjects have such high connection speed. This result shows that diffusion of the high speed Internet in
Korea is growing faster and the plan for the national information superhighway have been made in good
progress. More than half of the respondents in both countries access the Internet primarily at home. While
most of the U.S. respondents have more than 2 years of experience on the Internet, more than half of the
Korean subjects have less than 2 years experience. This statistic means that the information infrastructure

- 12 -

is fast growing but the Internet user base is still immature in Korea. Other Internet usage statistics are also
reported in Table 4.

-- Insert Table 4 here --

4.3.3.

Online Purchasing Statistics

As shown in Table 5, the majority of respondents (USA: 93.2%, Korea: 84.3%) have visited an online
shopping mall. While most of the visitors (87.5%) among the U.S respondents have purchased
product/service via the Internet at least once, 43.1% of Korean subjects have never purchased on the
Internet. This result also shows that the environment of e-Commerce is still immature in Korea in spite of
the advancement of the network infrastructure.

-- Insert Table 5 here --

4.4.

Validating the Instruments

4.4.1.

Reliability

Internal consistency reliability, also called reliability of components, looks at the extent to which the items
used to assess a construct reflect a true common score for the construct (Barki and Hartwick, 1994). In this
study, the internal consistency reliability is measured by applying the Cronbachs alpha (Cronbach and
Meehl, 1955) test to the individual scales and the overall measure. As the Cronbachs alpha values for all
the constructs (i.e., PEU, PU, PRP, PRT, and PB) in the datasets of the two countries are greater than the
guideline of .70 as specified by Nunally (1978), we conclude that the scales can be applied for the analysis
of the two datasets with acceptable reliability. Scale reliabilities along with mean and standard deviation of
each scale item are reported in Table 6. As represented in Table 6, the U.S. respondents feel much more
ease of use and usefulness on online shopping than Korean respondents. Comparing the two datasets, the

- 13 -

Korean respondents perceive more risk in the context of transaction. We can assume that this result is due
to the difference of physical infrastructure in both countries.

-- Insert Table 6 here --

4.4.2.

Construct Validity

Construct validity refers to the degree to which the test or questionnaire score is a measure of the
psychological characteristic of interest (Barki and Hartwick, 1994; Cronbach and Meehl, 1955; Rosential
and Rosnow, 1991). Implementing the datasets of the two countries, the data were examined using
principal components analysis as the extraction technique and Varimax as the method of rotation. The
factor analysis conducted in this study found a 5-factor structure with 19 scales loading with eigenvalues
greater than 1.0 that accounted for 70.6% (USA) and 66.8% (Korea) of the total variance respectively.
Items intended to measure the same construct demonstrated markedly higher factor loadings (>.50) on a
single component as represented in Table 7. Consequently, the instrument of our study shows adequate
validity in both datasets for further analysis.

-- Insert Table 7 here --

5. Results
We examined the e-CAM model using structural equation modeling (SEM) technique (Bollen, 1989; Hoyle,
1995; Loehlin, 1987). SEM technique not only allows researchers to analyze a set of latent constructs
much like independent and dependent variables in regression analysis (Segars and Grover, 1993), but also
provides researchers with a comprehensive means assessing and modifying theoretical models (Karahanna
and Straub, 1999; Subramanian, 1994). AMOS (Analysis of MOment Structures) with maximum
likelihood estimation (MLE) was used to analyze the data. AMOS is a particularly powerful SEM tool that
has been used in several previous literatures (Dishaw and Strong, 1999; Jarvenpaa and Tractinsky, 1999;

- 14 -

Jarvenpaa et al., 2000). The software package used to perform the analysis was AMOS 4.0 in SPSS 10.0
for Windows.

5.1.

Fitness of the e-CAM model

We also evaluated the construct validity of the models scales using confirmatory factor analysis (CFA).
As there is no single recommended measure of fit for the structural equation model (SEM), a variety of
measures are proposed in numerous literatures (Adams et al., 1992; Chau, 1997; Chin and Todd, 1995;
Hoyle, 1995; Hu et al., 1999; Loehlin, 1987; Segars and Grover, 1993; Subramanian, 1994) as represented
in Table 8.

-- Insert Table 8 here --

A Chi-square (2) statistic indicates that our model does not fit the U.S. data (2 = 243.11; p .05) and
Korean data (2=372.80; p .05). However, as 2 is a direct function of sample size, in small sample size
like our study, the statistic may lead to inaccurate probability value (Chau, 1997). Instead, we assessed our
model using other multiple fit criteria as reported in Table 8, such as 2 /degree of freedom (DF), goodnessof-fit index (GFI), adjusted goodness-of-fit index (AGFI), normed fit index (NFI), nonnormed fit index
(NNFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA). The values
of 2/DF are around 1.665 (USA) and 2.553 (Korea), which is below the desired cutoff value of 3.0 as
recommended by several literatures (Chau, 1997; Hu et al., 1999; Segars and Grover, 1993). In addition,
the GFI and AGFI values are .880 and .844 in the U.S. dataset, and .863 and .822 in Korean dataset
respectively, indicating a reasonable fit, while GFI is somewhat lower than the recommended value.
Further, the, NNFI (USA: .937, Korea: .901), CFI (USA: .946, Korea: .903), and RMSEA (USA: 0.062,
Korea: .076) are all adequate levels. Therefore, all these fit measures indicate that our model is reasonably
acceptable to assess the results for our structural model (e-CAM).

- 15 -

5.2.

Hypotheses Testing

The results of the multivariate test of the structural model in the two datasets are presented in Table 9. The
table shows the square multiple correlation (R2) as well as the path coefficients including direct and indirect
effects. The explained variances in perceived usefulness are .560 and .556 respectively in the datasets of
the two countries. Further, the models as a whole explain .308 and .168 of the variances respectively in eCommerce Adoption, i.e. purchasing behavior, as reported in Table 9.

-- Insert Table 9 here --

5.2.1.

Hypotheses Testing Results (USA)

As expected in Hypotheses H-1 and H-2, PRT and PRP both negatively affect the adoption of e-Commerce
(i.e., purchasing behavior). The data show that PRT (H-1) and PRP (H-2) are strongly associated with the
e-Commerce adoption (PRT: = -.293, p = .014; PRP: = -.442, p = .011). Furthermore, consistent with
Hypothesis H-3, PRT is highly correlated to PRP (correlation = .581, p = .000). In accordance with TAM,
PU has positive direct effects on the adoption of e-Commerce ( = .498, p = .019) while PEU is little
directly related to the adoption. As represented in Table 10, PEU has strong positive effects on PU (H-4:
= .932, p = .000), yet the direct effect of PEU on the e-Commerce adoption is insignificant (H-5: = .179,
p = .486). However, it should be noted that PEU has significant indirect effect (.465) on the adoption of eCommerce as reported in Table 9. Table 10 summarizes the test results and regression statistics in the U.S.
dataset to test the hypotheses.

-- Insert Table 10 here --

5.2.2.

Hypotheses Testing Results (Korea)

Inconsistent with Hypotheses H-1 and H-2, PRT and PRP both have no significant relationship with the
consumers purchasing behavior. The data show that PRT (H-1) and PRP (H-2) has little effect on PB

- 16 -

(PRT: = -.132, p = .113; PRP: = -.150, p = .217). Consistent with Hypothesis H-3, PRT is highly
correlated to PRP (correlation = .538, p = .000). Inconsistent with TAM, PU is little directly related to the
adoption of e-Commerce ( = .072, p = .513) while PEU has significantly direct/indirect effects on the
consumers purchasing behavior. As represented in Table 11, PEU has strong positive effects on PU (H-4:
= .824, p = .000), and the direct effect of PEU on the e-Commerce adoption is also significant (H-5:
= .327, p = .012). Therefore, it should be noted that PEU has significant direct effect on the adoption of eCommerce in Korean dataset. Table 11 summarizes the test results and regression statistics in Korean
dataset to test the hypotheses.

-- Insert Table 11 --

5.3.

Summary of the Results

As presented in Figure 5, most of causal relationships between the constructs postulated by the e-CAM
model are well supported in the U.S. dataset. The results suggest that perceived usefulness (PU), perceived
risk in the context of transaction (PRT), and perceived risk with product/service (PRP) have significant
direct effects on consumers adoption of e-Commerce. On the contrary, only the PEU construct of the
model in Korean dataset has a significant direct effect on the consumers purchasing behavior as illustrated
in Figure 6. With regard to the PEU construct, direct effect ( = .179) on the adoption is not significant in
the U.S. dataset while significant direct effect ( = .327) is shown in the Korean dataset. However, PEU in
the U.S. dataset mostly has indirect effect ( = .465) on the adoption through the mediating construct (i.e.,
PU) as reported in Table 9. The insignificant direct effect of PEU on the adoption of e-Commerce is not
unexpected but shows the same pattern as found in several prior TAM research (Davis, 1989; Gefen and
Straub, 1997; Keil et al., 1995; Morris and Dillon, 1997; Subramanian, 1994; Szajna, 1994; Szajna, 1996).
Therefore, we believe that all the determinant factors of the consumers purchasing behavior (i.e., PEU, PU,
PRT, and PRP) have significant total effects on the adoption of e-Commerce including direct/indirect
effects in the U.S. dataset. The effect of PEU on the adoption, however, remains controversial in that
several TAM studies show that PEU directly influences IT adoption (Gefen and Straub, 2000). Regarding
- 17 -

the impacts of PRT and PRP, both constructs in the U.S. dataset have strong direct effects on the adoption
of e-Commerce while the result from the Korean dataset shows no significant effects of both constructs on
the adoption. The result in the U.S. dataset is consistent with Rose et al.s study (1999), which describes
online transactional risk (i.e., security, privacy, etc.) as the most important impediment to B2C eCommerce. Unlike the United States, information industries in Korea have focused their efforts on
systems development, while paying little attention to related commercial problems. Although consumers in
Korea have shown more concerns for these risks (PRT and PRP) as reported in Table 6, they did not
consider the risks as significant decision factors. While it is true that there are some differences between
cultures, it is expected that these construct will eventually have negative effects on the e-Commerce
adoption in Korea.
In sum, in the U.S. dataset, all of the antecedent constructs (i.e. PRT, PRP, PEU and PU) directly
and/or indirectly affect consumers adoption of e-Commerce while PEU construct only directly affect the
adoption in the Korean dataset.

-- Insert Figure 5 here --

-- Insert Figure 6 here --

6. Discussion
In this research, we attempt to identify valid factors that predict a consumers online purchasing behavior
based on the e-CAM model, which integrates TAM with the theories of perceived risk to explain the
consumers adoption of e-Commerce. The results of this study suggest that a maturity level of eCommerce environment has an effect on overall customer models for the environment. For example, the
Internet and e-Commerce is growing yet still immature in Korea. Thus, the e-CAM model in Korea is
affected by an inexperienced user base as well as a still developing technology infrastructure. The United
States, in contrast, has a strong infrastructure and experienced user base, and thus the e-CAM model could

- 18 -

be better validated as expected. The difference between the two countries models will gradually decrease
as the Internet and e-Commerce infrastructure and user base in Korea grow and develop.
In conclusion, our e-CAM model holds much promise for helping researcher and practitioners better
understanding why consumers adopt or reject e-Commerce for a purchasing vehicle.

6.1.

Implications

The findings of this study have significant implications in the perspective of research on e-Commerce
consumer behavior. Our study provides further evidence on the appropriateness of using the TAM model
to measure the different dimensions of actual usage on e-Commerce. As expected from the previous TAM
research, two specific behavioral beliefs (i.e., PEU and PU) well explain the adoption of e-Commerce by
Internet users. In addition, the results from our study have shown that other factors with regard to
perceived risk (i.e., PRT and PRP) help us better understand the e-Commerce adoption. Therefore, in the
e-Commerce setting, the power of the TAM model will be greatly enhanced by taking into account the
impact of the perceived risks with regard to product/service itself as well as the context of online
transaction. In sum, our study has significant contributions to the theories of perceived risk as well as the
TAM model by providing empirical evidence that PEU, PU, PRT, and PRP are important factors that
influence the consumers adoption of e-Commerce.
The findings of the study also suggest important practical implications for businesses currently
providing products/services on the Internet as well as that are planning to do so. It is evident from this
study that to convert Internet users into real buyers, perceived ease of use (PEU) and perceived usefulness
(PU) must be enhanced and the perceived risk relating to product/service (PRP) and online transaction
(PRT) reduced. From the perspective of a consumers perceived risk, the consumer is willing to purchase
product/service from an online vendor that is perceived low risk, even if the consumers perceived ease of
use or usefulness on e-Commerce is relatively low (Jarvenpaa et al., 2000). As shown in the results of our
study, consumers consider the risk related to the online transaction (i.e., privacy, security, nonrepudiation,
etc.) as one of the important factors when they purchase on the Internet. Thus, diminishing such risk is
considerably important to online vendors. To lower the transactional risk, online vendors should establish a
robust mechanism, which utilize state-of-art Internet technologies. However, as criminals will always exist
- 19 -

in our society, if credit card is lost or stolen in a physical place, anyone can abuse the card information
(Ratnasingham, 1998). Hence, it is more important that online vendors enable consumers to place full trust
on the privacy, security, integrity and availability of vendor information. Further, to reduce consumers
risk with product/service, online vendors should build trust with consumers by giving them complete
confidence on the product/service that they provide. We believe this can be achieved when online vendors
improve the following service quality factors: (i) reliability: the ability to perform a promised service
reliably and accurately; (ii) responsiveness: the willingness to assist customers and to provide prompt
service; (iii) assurance: the knowledge and courtesy of employees and their ability to convey trust, and
confidence; (iv) empathy: the caring attitude which provides individualized attention to customers
(Kettinger and Lee, 1997). In sum, firms providing products/services through e-Commerce should consider
these contextual factors in order to facilitate consumer adoption behavior.

6.2.

Limitations and Future Research Direction

Our study has certain limitations like any other research. First, our sampling pool was restricted to mainly
academic circles for economic constraints, and therefore, most of the respondents were highly educated and
well experienced regarding the Internet. In addition, results of the study might also be biased given the
youth. Although the young respondents will finally grow to be influential consumers on the Internet,
examining with more diverse Web users, such as older, less educated, and less experienced on the Internet,
may enable us to construct and validate more generalized model. Second, the explanation power is
relatively low compared with previous TAM studies. In the U.S. dataset, the proposed model explains
only 30.8% of the total variance (R2 =.308) of consumers e-Commerce adoption behavior. Further, the
explanation power of the Korean model is only 16.8% of the total variance (R2 =.168). These results mean
our model might have missed some important factors to the consumers behavior on e-Commerce. In this
exploratory study, due to the sampling limitations, some factors regarding individual differences, such as
gender, education, income, and Internet experience, were deliberately left out. Further, we do not include
some characteristics of product/service, such as low asset specificity and ease of description
(standardization), which also differentiate the products/services correctly sold over the Internet, thereby can
draw consumers to purchase. Therefore, it is suggested that additional applicable factors including

- 20 -

individual differences and product/service characteristics be identified and included in future research for
better understanding of consumers behavior and the state of e-Commerce. Finally, the measures for the
construct of online purchasing behavior used in our study are self-reported; thereby the respondents may
not correctly answer their experience due to their limited memory. Although self-reported measures are
used in numerous studies, and the interchangeability of self-report and objective usage measures remains a
controversial point in IS research (Straub et al., 1995, Venkatesh and Davis, 2000), we believe that future
research can develop more objective and accurate measures, such as an analysis of consumers actual visit
and purchase log on real e-Commerce site.

- 21 -

References
Adams, D. A., R. R. Nelson, and P. A. Todd. (1992). "Perceived Usefulness, Ease of Use, and Usage of
Information Technology: A Replication," MIS Quarterly 16 (2), 227-250.
Agarwal, R. and J. Prasad. (1997). "The Role of Innovation Characteristics and Perceived Voluntariness in
the Acceptance of Information Technologies," Decision Sciences 28 (3), 557-581.
Agarwal, R. and J. Prasad. (1998). "A Conceptual and Operational Definition of Personal Innovativeness in
the Domain of Information Technology," Information Systems Research 9 (2), 204-215.
Agarwal, R. and J. Prasad. (1999). "Are individual differences germane to the acceptance of new
information technologies?," Decision Sciences 30 (2), 361-391.
Ajzen, I. (1991). "The Theory of Planned Behavior," Organizational Behavior and Human Decision
Processes 50 (2), 179-211.
Ajzen, I. and M. Fishbein. (1980). Understanding Attitudes and Predicting Social Behavior, Eaglewood
Cliffs, NJ: Prentice-Hall.
Barki, H. and J. Hartwick. (1994). "Measuring User Participation, User Involvement, and User Attitude,"
MIS Quarterly 18 (1), 59-82.
Bauer, R. (1960). "Consumer Behavior as Risk Taking," in Proceedings of the 43rd National Conference of
the American Marketing Association, December, 389-398.
Bhimani, A. (1996). "Securing the Commercial Internet," Communications of the ACM 39 (6), 29-35.
Bollen, K. A. (1989). Structural Equations with Latent Variables, New York, NY: John Wiley.
Chau, P. Y. K. (1997). "Reexamining a Model for Evaluating Information Center Success Using a
Structural Equation Modeling Approach," Decision Sciences 28 (2), 309-334.
Chin, W. W. and P. A. Todd. (1995). "On the Use, Usefulness, and Ease of Use of Structural Equation
Modeling in MIS Research: A Note of Caution," MIS Quarterly 19 (2), 237-246.
Cox, D. F. and S. U. Rich. (1964). "Perceived Risk and Consumer Decision Making - The Case of
Telephone Shopping," Journal of Marketing Research 1 (4), 32-39.
Cronbach, L. and P. Meehl. (1955). "Construct Validity in Psychological Tests," Psychological Bulletin 52,
281-302.

- 22 -

Cunningham, S. M. (1967). "The Major Dimensions of Perceived Risk," in Cox, D. F. (ed.), Risk Taking
and Information Handling in Consumer Behavior, Boston: Graduate School of Business
Administration, Harvard University, 82-108.
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., R. P. Bagozzi, and P. R. Warshaw. (1989). "User Acceptance of Computer Technology: A
Comparison of Two Theoretical Models," Management Science 35 (8), 982-1003.
Dishaw, M. T. and D. M. Strong. (1999). "Extending the Technology Acceptance Model with Tasktechnology Fit Constructs," Information & Management 36 (1), 9-21.
Doll, W. J., A. Hendrickson, and X. Deng. (1998). "Using Davis's Perceived Usefulness and Ease-of-Use
Instruments for Decision Making: A Confirmatory and Multigroup Invariance Analysis," Decision
Sciences 29 (4), 839-869.
Fenech, T. (1998). "Using Perceived ease of use and perceived usefulness to predict acceptance of the
World Wide Web," Computer Network and ISDN Systems 30 (1-7), 629-630.
Gefen, D. and M. Keil. (1998). "The Impact of Developer Responsiveness on Perceptions of Usefulness
and Ease of Use: An Extension of the Technology of the Technology Acceptance Model," Database
for Advances in Information Systems 29 (2), 35-49.
Gefen, D. and D. Straub. (2000). "The Relative Importance of Perceived Ease of Use in IS Adoption: A
Study of E-Commerce Adoption," Journal of the Association for Information Systems 1 (8), 1-28.
Gefen, D. and D. W. Straub. (1997). "Gender Differences in the Perception and Use of E-mail: An
Extension to the Technology Acceptance Model," MIS Quarterly 21 (4), 389-400.
GVU's WWW User Survey Team. (1998). "GVU's 10th WWW User Surveys." URL:
http://www.gvu.gatech.edu/user_surveys/survey-1998-10/ (Internet Publication).
Hendrickson, A. R., P. D. Massey, and T. P. Cronan. (1993). "On the Test-Retest Reliability of Perceived
Usefulness and Perceived Ease of Use Scale," MIS Quarterly 17 (2), 227-230.
Hoffman, D. L., T. P. Novak, and M. Peralta. (1999). "Building Consumer Trust Online," Communications
of the ACM 42 (4), 80-85.

- 23 -

Hoyle, R. (1995). Structural Equation Modeling: Concepts, Issues, and Applications, Thousand Oaks, CA:
Sage.
Hu, P. J.et al. (1999). "Examining the Technology Acceptance Model Using Physician Acceptance of
Telemedicine Technology," Journal of MIS 16 (2), 91-112.
Igbaria, M.et al. (1997). "Personal Computing Acceptance Factors in Small Firms: A Structural Equation
Model," MIS Quarterly 21 (3), 279-302.
ISPO. (1997). "A European Initiative in Electronic Commerce," COM (97) 157. URL:
http://www.cordis.lu/esprit/src/ecomcom.htm (Internet Publication), April 15.
Jacoby, J. and L. B. Kaplan. (1972). "The Components of Perceived Risk," in Proceedings of the 3rd
Annual Conference of the Association for Consumer Research, Association for Consumer Research,
382-393.
Jarvenpaa, S. L. and N. Tractinsky. (1999). "Consumer Trust in an Internet Store: A Cross-Cultural
Validation," Journal of Computer-Mediated Communication 5 (2). URL:
http://www.ascusc.org/jcmc/vol5/issue2/jarvenpaa.html (Internet Publication).
Jarvenpaa, S. L., N. Tractinsky, and M. Vitale. (2000). "Consumer Trust in an Internet Store," Information
Technology and Management 1 (1-2), 45-71.
Karahanna, E. and D. W. Straub. (1999). "The Psychological Origins of Perceived Usefulness and Ease-ofuse," Information & Management 35 (4), 237-250.
Keil, M., P. M. Beranek, and B. R. Konsynski. (1995). "Usefulness and Ease of Use: Field Study Evidence
Regarding Task Considerations," Decision Support Systems 13 (1), 75-91.
Kettinger, W. J. and C. C. Lee. (1997). "Pragmatic Perspective on Measurement of Information Service
Quality," MIS Quarterly 21 (2), 223-228.
Lederer, A. L.et al. (2000). "The Technology Acceptance Model and the World Wide Web," Decision
Support Systems 29 (3), 269-282.
Lin, J. C.-C. and H. Lu. (2000). "Towards an Understanding of Behavioral Intention to Use a Web Site,"
International Journal of Information Management 20 (3), 197-208.
Loehlin, J. (1987). Latent Variable Models - An Introduction to Factor, Path, and Structural Analysis,
Hillsdale, NJ: Lawrence Erlbaum Associates.

- 24 -

Mathieson, K. (1991). "Predicting User Intensions: Comparing the Technology Acceptance Model with
Theory Planned Behavior," Information Systems Research 2 (3), 192-222.
Moore, G. and I. Benbasat. (1991). "Development of an Instrument to Measure the Perceptions of Adopting
an Information Technology Innovation," Information Systems Research 2 (3), 192-222.
Morris, M. G. and A. Dillon. (1997). "How User Perceptions Influence Software Use," IEEE Software 14
(4), 58-65.
Murphy, P. E. and B. M. Enis. (1986). "Classifying Products Strategically," Journal of Marketing 50 (3),
24-42.
Nunally, J. (1978). Psychometric Theory, 2nd ed., New York, NY: McGraw-Hill.
Ram, S., J. Park, and D. Lee. (1999). "Digital Libraries for the Next Millennium: Challenges and Research
Directions," Information Systems Frontiers 1 (1), 75-93.
Ratnasingham, P. (1998). "The Importance of Trust in Electronic Commerce," Internet Research:
Electronic Networking Applications and Policy 8 (4), 313-321.
Rogers, E. (1995). Diffusion of Innovations, 4th ed., New York, NY: Free Press.
Rose, G., H. Khoo, and D. W. Straub. (1999). "Current Technological Impediments to Business-toConsumer Electronic Commerce," Communications of the AIS 1 (16), 1-74.
Roselius, T. (1971). "Consumer Ranking of Risk Reduction Methods," Journal of Marketing 35 (1), 56-61.
Rosential, R. and R. L. Rosnow. (1991). Essentials of Behavioral Research: Methods and Data Analysis,
2nd ed., Boston, MA: McGraw-Hill.
Segars, A. H. and V. Grover. (1993). "Re-Examining Perceived Ease of Use and Usefulness: A
Confirmatory Factor Analysis," MIS Quarterly 17 (4), 517-527.
Sindhav, B. and A. L. Balazs. (1999). "A Model of Factors Affecting the Growth of Retailing on the
Internet," Journal of Market Focused Management 4 (4), 319-339.
Straub, D., M. Keil, and W. Brenner. (1997). "Testing the Technology Acceptance Model Across Cultures:
A Three Country Study," Information & Management 33 (1), 1-11.
Straub, D., M. Limayem, and E. Karahanna-Evaristo. (1995). "Measuring System Usage: Implications for
IS Theory Testing," Management Science 41 (8), 1328-1342.

- 25 -

Subramanian, G. H. (1994). "A Replication of Perceived Usefulness and Perceived Ease of Use
Measurement," Decision Sciences 25 (5-6), 863-874.
Swaminathan, V., E. Lepkowska-White, and B. P. Rao. (1999). "Browsers or Buyers in Cyberspace? An
Investigation of Factors Influencing Electronic Exchange," Journal of Computer-Mediated
Communication 5 (2). URL: http://www.ascusc.org/jcmc/vol5/issue2/swaminathan.htm (Internet
Publication).
Szajna, B. (1994). "Software Evaluation and Choice: Predictive Validation of the Technology Acceptance
Instrument," MIS Quarterly 18 (3), 319-324.
Szajna, B. (1996). "Empirical Evaluation of the Revised Technology Acceptance Model," Management
Science 42 (1), 85-92.
Tan, M. and T. S. H. Teo. (2000). "Factors Influencing the Adoption of Internet Banking," Journal of the
AIS 1 (5), 1-42.
Taylor, J. W. (1974). "The Role of Risk in Consumer Behavior," Journal of Marketing 38 (2), 54-60.
Taylor, S. and P. Todd. (1995). "Assessing IT Usage: The Role of Prior Experience," MIS Quarterly 19 (4),
561-570.
Teo, T. S. H., V. K. G. Lim, and R. Y. C. Lai. (1999). "Intrinsic and Extrinsic Motivation in Internet
Usage," Omega 27 (1), 25-37.
U.S. Bureau of the Census. (2000). "Retail eCommerce sales in Second Quarter 2000 Increased 5.3 percent
from First Quarter 2000." URL: http://www.census.gov/mrts/www/current.html (Internet Publication),
August.
U.S. Department of Commerce. (1999). "The Emerging Digital Economy II." URL:
http://www.ecommerce.gov/ede/report.html (Internet Publication), June.
U.S. Department of Commerce. (2000). "Digital Economy 2000." URL: http://www.esa.doc.gov/de2k.htm
(Internet Publication), June.
USIC and IITA. (2000). "The State of the Internet 2000." URL: http://usic.wslogic.com/intro.html (Internet
Publication), September.
Venkatesh, V. (1999). "Creation of Favorable User Perceptions: Exploring the Role of Intrinsic
Motivation," MIS Quarterly 23 (2), 239-260.

- 26 -

Venkatesh, V. and F. D. Davis. (1996). "A Model of the Antecedents of Perceived Ease of Use:
Development and Test," Decision Sciences 27 (3), 451-481.
Venkatesh, V. and F. D. Davis. (2000). "A Theoretical Extension of the Technology Acceptance Model:
Four Longitudinal Field Studies," Management Science 46 (2), 186-204.
Venkatesh, V. and M. G. Morris. (2000). "Why Don't Men Ever Stop to Ask for Directions? Gender, Social
Influence, and Their Role in Technology Acceptance and Usage Behavior," MIS Quarterly 24 (1),
115-139.
Zwass, V. (1996). "Electronic Commerce: Structures and Issues," International Journal of Electronic
Commerce 1 (1), 3-23.

- 27 -

Table 1: Previous Research on Technology Adoption Model (TAM)

Research

Constructs

Type of IS

Results

Characteristics

Davis (1989)

PEU, PU, U

E-mail,
Graphics

PU, PEUU

Validation of PU and PEU, primary determinants

of user acceptance.

Davis et al.
(1989)

PEU, PU, A,
BI, U

Word

PEUPU / PU, PEUA

A, PUBIU

Explanation of users computer acceptance and

their intentions in terms of A, SN, PU, PEU, etc.

Mathieson
(1991)

EV, PEU, PU,

A, BI, U

Spreadsheet,
Calculator

PU, PEUU

Support for the reliability of PU/PEU and its

association with self-reported intention.

Adams et al.
(1992)

PEU, PU, U

E-mail/V-mail

PEUPU / PU, PEUU

Replication of Daviss previous work on TAM.

Hendrickson et al. PEU, PU

(1993)

Spreadsheet,
DB

Reliability of PEU/PUHigh

Test-retest reliability of PU and PEU.

Segars & Grover

(1993)

PEU, PU, EF

E-mail, V-mail PEUEFPUPEU

Three factor model with eight items: PEU, PU, &

EF.

Subramanian
(1994)

PEU, PU, U

V-mail, Dial up PUU

system

Replication of the impact of PU and PEU

constructs on predicted future usage.

Szajna (1994,
1996)

PEU, PU, BI,

U

PC-based
DBMS

PEUPUBIU

Measurement of actual user choice rather than

self-reported intentions.

Chin & Todd

(1995)

PEU, PU, EF

V-Mail

Daviss PU Unidimensional

Re-examining PU and EF used in (Segars and

Grover, 1993).

Straub et al.
(1995)

PEU, PU,
SPIR, U

V-mail

SPIRPU
PU, PEUU

Comparison of subjective and objective measures

of system usage.

Taylor & Todd

(1995)

PEU, PU, A,
Computing
Differences in the relative influence of the
PEUPU / PU, PEUA
SN, PBC, BI, B resource center A, SN, PBCBI / BI, PBCB determinants of usage depending on experience.

Venkatesh &
Davis (1996)

SE, OU, PEU,

DE

Word, Lotus,
Graphics, etc.

SEPEU / OU + DEPEU

Focus on understanding the determinants of PEU:

SE, OU, and DE.

Straub et al.
(1997)

PEU, PU, U

E-mail

PEU, PUU

Comparison of the TAM across cultures.

Morris & Dillon

(1997)

PEU, PU, A,
BI, U

Netscape

PEUPU, A / PUA
PU, A BIU

Significant influence of initial perceptions of

Netscapes PU and PEU on users attitude and BI.

Igbaria et al.
(1997)

IV, EV, PEU,

PU, U

Personal
Computer

IV, EV, PEUPU

EVPEU / PEU, PUU

Significant effect of user characteristics, system

quality, and organizational support on PEU & PU.

Gefen & Straub

(1997)

G, PEU, PU,
SPIR, U

E-mail

GSPIR, PU, PEU

SPIRPUU

Extension of the TAM and SPIR addendum by

adding gender to an IT diffusion model

Fenech (1998)

PEU, PU, A,
SE, U

WWW

PEUPU / PEU, PU, SEU

Inclusion of Consumer Self-Efficacy to improve

the prediction for WWW.

Gefen & Keil

(1998)

PDR, PEU, PU, Expert System

U

PDRPEU, PU
PEUPUU

An extension of TAM based on Social Exchange

Theory.

Doll et al. (1998) PEU, PU, G,

DE

Graphics,
GPU / DEPEU
Word, DB, etc. Type of ISPEU

Support for the validity and reliability of Daviss

PU and PEU instrument.

Teo et al. (1999)

PEU, PU, PE,

U

Internet

PEUPU, U, PE
PU, PEU

Focus on both intrinsic (PE) and extrinsic (PU)

motivation for the use of Internet.

Agarwal &
Prasad (1999)

ID, PEU, PU,

A, BI

Personal
Computer

IDPEU, PU / PEUPU, A
PUA, BI / ABI

Intervention of beliefs or perceptions (in TAM)

between individual difference and IT acceptance.

Hu et al. (1999)

PEU, PU, A,
BI

Telemedicine

PUABI / PUBI

Examinations of physicians decisions to accept

telemedicine IT in the health-care context.

Venkatesh (1999) PEU, PU, BI,

TI

Internet-based
application

TIPEU / TI, PEU, PUBI

Discovery of the effect of intrinsic motivation in

training on creating favorable user perceptions.

Karahanna &
Straub (1999)

PEU, PU, SP,

SI, ACC, U

E-mail

ACCPEU / PEU, SP, SIPU Explanation for the psychological origin of PU

and PEU based on TRA, SP/SI theory.
PEU, PUU

Lederer et al.
(2000)

PEU, PU, U

WWW

PEU, PUU

Validation of TAM in the context of the WWW.

Venkatesh &
Morris (2000)

PEU, PU, SN,

BI, G, DE

New SW
system

PU, PEU, SNBI

Longitudinal examination with the integration of

SN into TAM using gender as a moderator.

- 28 -

Research

Constructs

Type of IS

Results

Characteristics

Venkatesh &
Davis (2000)

SN, I, Q, RD,
JR, E, V, PU,
PEU, BI, U

New SW
system

A theoretical extension of TAM that explains PU

SN, I, JR, Q, RD, EPU
PEUPU / SN, V, EBI PEU, and usage BI in terms of social influence and
cognitive instrumental processes.
PUBIU

Lin & Lu (2000)

Q, PU, PEU, A, WWW

BI

Investigation on users perception about WWW in

view of IS quality.

QPEU, PUABI
PEUPUBI

ACRONYMS: A (Attitude); ACC (Physical Accessibility); B (Behavior); BI (Behavioral Intention); E (Experience); EF

(Effectiveness); EV (External Variables); G (Gender); IR (Information Richness); I (Image); IV (Internal Variables); JR (Job
Relevance); OU (Objective Usability); PBC (Perceived Behavioral Control); PDR (Perceived Developer Responsiveness); PE
(Perceived Enjoyment); PEU (Perceived Ease of Use); PU (Perceived Usefulness); Q (Quality); RD (Result Demonstrability); SE
(Self-Efficacy); SI (Social Influence); SN (Subjective Norm); SP (Social Presence); SPIR (SP + IR); TI (Training Intervention); U
(Usage); V (Voluntariness)

Table 2: Definition of Perceived Risk Type

Risk Type

Definition

Financial Risk

The risk that the product will not be worth the financial price

Psychological Risk

The risk that the product will lower the consumers self image.

Physical Risk

The risk to the buyers or others safety in using products

Functional Risk

The risk that the product will not perform as expected

Social Risk

The risk that a product choice may result in embarrassment before ones friends/family/work group

Time Risk
(Non-monetary)

The risk of time spent preparing shopping lists, traveling, seeking information, shopping, and
waiting for product delivery.

Table 3: General Demographic Statistics

Variable
Gender

Male

Female
Age

16 25

26 35

36 45

46 55

Over 55
Education

High School or equivalent

Vocational/Technical School

Some College

College Graduate

Masters Degree

Doctoral Degree

Other

- 29 -

USA (N = 175)

Korea (N = 267)

117 (66.5%)
59 (33.5%)

183 (68.5%)
84 (31.5%)

104 (59.1%)
53 (30.1%)
14 (8.0%)
4 (2.2%)
1 (0.6%)

182 (68.2%)
76 (28.5%)
6 (2.2%)
2 (0.7%)
1 (0.4%)

13 (7.4%)
1 (0.6%)
98 (55.7%)
32 (18.2%)
25 (14.2%)
6 (3.4%)
1 (0.6%)

6 (2.2%)
1 (0.4%)
202 (75.7%)
30 (11.2%)
24 (9.0%)
1 (0.4%)
3 (1.1%)

Table 4: Internet Usage Statistics

Variable

USA (N = 176)

Korea (N = 267)

17 (9.7%)
70 (39.8%)
11 (6.3%)
4 (2.3%)
19 (10.8%)
40 (22.7%)
15 (8.5%)

7 (2.6%)
23 (8.6%)
29 (10.9%)
12 (4.5%)
92 (34.5%)
50 (18.7%)
64 (20.4%)

99 (56.3%)
51 (29.0%)
26 (14.8%)
0 (0.0%)

150 (56.2%)
42 (15.7%)
57 (21.3%)
18 (6.7%)

0 (0.0%)
0 (0.0%)
0 (0.0%)
11 (6.3%)
165 (93.7%)

4 (1.5%)
18 (6.7%)
58 (21.7%)
73 (27.3%)
114 (42.7%)

0 (0.0%)
0 (0.0%)
5 (2.8%)
37 (21.1%)
134 (76.1%)

1 (0.4%)
1 (0.4%)
18 (6.7%)
113 (42.3%)
134 (50.2%)

2 (1.1%)
42 (23.9%)
45 (25.6%)
36 (20.5%)
51 (29.0%)

17 (6.4%)
78 (29.2%)
68 (25.5%)
46 (17.2%)
58 (21.7%)

Primary Internet Connection

28.8-33.6k Modem

56k Modem

Cable Modem

ISDN Line

ADSL/DSL

T1/T3 Line

Other
Primary Place to Use the Internet

Home

School

Work

PC Parlors
Internet Using Experience

Less than 3 months

36 months

612 Months

12 years

More than 2 years

Internet Using Frequency

A few times a month

Once a week

A few times a week

12 times a day

More than twice a day

Internet Using Hours per Week

Less than 1 hours

15 hours

610 hours

1120 hours

More than 20 hours

Table 5: On-line Purchasing Statistics

Variable
Electronic Shopping-Mall Visiting Experience

Yes

No
Electronic Purchasing Amount (Last 6 months)

None

Less than $50

$50$99

$100$199

$200$499

$500$999

$1,000$2,000

More than $2,000

Electronic Purchasing Times (Last 6 months)

None

12 times

35 times

610 times

More than 10 times

- 30 -

USA (N = 176)

Korea (N = 267)

164 (93.2%)
12 (6.8%)

225 (84.3%)
42 (15.7%)

22 (12.5%)
19 (10.8%)
21 (11.9%)
24 (13.6%)
37 (21.0%)
21 (11.9%)
8 (4.5%)
24 (13.6%)

115 (43.1%)
67 (25.1%)
20 (7.5%)
23 (8.6%)
22 (8.2%)
9 (3.4%)
2 (0.7%)
9 (3.4%)

22 (12.5%)
50 (28.4%)
64 (36.4%)
20 (11.4%)
20 (11.4%)

115 (43.1%)
85 (31.8%)
40 (15.0%)
18 (6.7%)
9 (3.4%)

Table 6: Cronbachs Alpha for the Scales

USA
Factors and Scale Items

Mean / SD

Perceived Ease of Use *

Easy to search and locate desired information
Easy to use from any location at any time
Easy to use the customer service
Overall PEU
Perceived Usefulness *
Save money
Save time
Provide wide variety of products/ services
Overall PU
Perceived Risk with Product/Service *
Functional loss
Time loss
Financial loss
Opportunity loss
Overall PRP
Perceived Risk in the Context of Transaction *
Privacy
Security (Credit card)
Non-repudiation
Overall PRT
Purchase Behavior
Total Amount of Online Purchasing **
Frequency of Online Purchasing ***

KOREA

Cronbach

Mean / SD

Cronbach

.779
5.39
5.89
4.39
5.55

/
/
/
/

1.36
1.40
1.50
1.18

5.11
5.73
5.44
5.82

/
/
/
/

1.09
1.19
1.29
1.08

4.98
5.32
5.24
4.82
4.80

/
/
/
/
/

1.26
1.34
1.57
1.35
1.06

5.14
4.30
4.09
4.02

/
/
/
/

1.48
1.82
1.68
1.63

.754
4.43
5.20
3.90
4.37

/
/
/
/

1.41
1.48
1.52
1.28

4.67
4.80
4.06
4.72

/
/
/
/

1.44
1.53
1.16
1.12

4.55
5.03
4.86
5.02
4.92

/
/
/
/
/

1.29
1.19
1.25
0.98
1.07

5.85
5.29
4.75
4.30

/
/
/
/

1.27
1.56
1.35
1.51

.843

.824

.811

.812

.875

.810

.827

.907
2.44 / 1.83
1.96 / 1.08

4.53 / 2.12
2.91 / 1.16

*: 7-point scales ranging from strongly disagree to strongly agree.

**: 8-point scales ranging from none to more than $2,000.
***: 5-point scales ranging from none to more than 10 times.

Table 7: Factor Analysis for Construct Validation

Factor Loading
Measured Items
Ease of search
Ease of ordering
Customer service
Overall ease of use
Save money
Save time
Variety of products
Overall usefulness
Functional loss
Time loss
Financial loss
Opportunity loss
Overall PRP
Privacy
Security (Credit card)
Non-repudiation
Overall PRT
Purchasing Times
Purchasing Amount
Eigenvalue
Cum. % of Variance

PRT
-.15
.04
-.19
-.07
-.15
-.10
-.26
-.18
.06
.15
.34
.24
.32
.76
.81
.78
.76
-.15
-.23
6.87
36.1

PRP
-.01
-.15
.04
-.04
.06
-.11
-.08
-.05
.79
.80
.51
.65
.84
.26
.16
.22
.21
-.21
-.20
2.96
51.7

USA
PU
.28
.40
.01
.43
.53
.73
.74
.78
.00
-.05
.25
-.30
-.09
.05
-.25
-.24
-.26
.18
.18
1.36
58.9

PEU
.69
.67
.72
.75
.34
.33
.16
.36
.00
-.10
-.15
.20
-.08
-.16
-.01
-.11
-.17
.13
.16
1.20
65.2

PB
-.05
.15
.22
.14
.44
.11
.05
.25
-.13
.04
-.26
-.13
-.17
-.03
-.10
-.19
-.25
.87
.86
1.02
70.6

PU
.34
.29
.02
.48
.75
.80
.69
.84
.02
.11
.06
.03
-.02
.03
.04
.00
-.19
.13
.13
4.65
24.5

PRT
-.12
.00
-.01
-.12
-.10
.17
-.08
-.05
-.09
.11
.26
.15
.49
.72
.86
.77
.74
-.11
-.12
3.81
44.5

KOREA
PRP
PEU
.07
.66
.07
.58
-.09
.82
-.01
.67
.08
.23
.07
-.02
-.02
.30
.10
.29
.07
.77
.05
.81
-.12
.77
.05
.65
-.13
.66
.15
.06
.09
-.04
.20
-.14
.08
-.10
-.08
.17
-.04
.08
1.66
1.48
53.2
61.0

* Extraction Method: Principal Component Analysis (Varimax Rotation with Kaiser Normalization)

- 31 -

PB
.07
.21
-.01
.152
.04
.128
.01
.127
-.117
.06
.03
-.03
-.14
.01
.04
-.14
-.22
.92
.92
1.09
66.8

Table 8: Fit Measures for the e-CAM Model

Fit Measures

Fit Guidelines

USA

p .05
3.0
.90
.80
.90
.90
.90
.08

243.11 (p=.00)
1.665
.880
.844
.877
.937
.946
.062

2
2 / DF
GFI
AGFI
NFI
NNFI
CFI
RMSEA

KOREA
372.80 (p=.00)
2.553
.863
.822
.841
.901
.903
.076

Table 9: Direct/Indirect Effects between the Constructs in e-CAM

Perceived Usefulness (PU)
USA
KOREA
Direct
Total
Direct
Total

Direct

Purchasing Behavior (PB)

USA
KOREA
Indirect
Total
Direct
Indirect

1)

Total

PRT
-.293
-.293
-.150
-.150
PRP2)
-.442
-.442
-.132
-.132
.932
.932
.824
.824
.179
.465
.643
.327
.059
.386
PEU
.498
.498
.072
.072
PU
R2 3)
.560
.556
.308
.168
1)
Perceived Risk in the context of Transaction; 2) Perceived Risk with Product/service; 3) Squared Multiple Correlation

Table 10: Hypotheses Testing Results (USA)

Hypothesis
H-1
H-2
H-3
H-4
H-5
H-6
1)

Causal
Relationship
PRT
PB (
)
PRP
PB (
)
PRT
PRP (+
+)
PEU
PU (+
+)
PEUPB (+)
PU
PB (+
+)

1)
-.293
-.442
.581
.932
.179
.498

S.E.2)
.119
.175
.141
.122
.257
.213

C. R.3)
-2.459
-2.533
5.391
7.631
.696
2.339

P4)
.014
.011
.000
.000
.486
.019

Result
Supported
Supported
Supported
Supported
Rejected
Supported

Regression Coefficient; 2) Standard Error of ; 3) Critical Ratio (= / S.E.); 4) Statistical Significance of the Test

Table 11: Hypotheses Testing Results (Korea)

Hypothesis
H-1
H-2
H-3
H-4
H-5
H-6
1)

Causal
Relationship
PRTPB ()
PRPPB ()
PRT
PRP (+
+)
PEU
PU (+
+)
PEU
PB (+
+)
PUPB (+)

1)
-.132
-.150
.538
.824
.327
.072

S.E.2)
.083
.122
.077
.084
.130
.110

C. R.3)
-1.586
-1.233
5.315
9.812
2.519
.654

P4)
.113
.217
.000
.000
.012
.513

Result
Rejected
Rejected
Supported
Supported
Supported
Rejected

Regression Coefficient; 2) Standard Error of ; 3) Critical Ratio (= / S.E.); 4) Statistical Significance of the Test

- 32 -

Figure 1: Daviss Technology Acceptance Model

Perceived
Usefulness (PU)
External
Variables

Attitude Toward
Using (A)

Behavioral
Intention to
Use (BI)

Actual
System Use

Perceived Ease of
Use (PEU)

Figure 2: Adapted TAM Model on the Adoption of e-Commerce

Perceived Ease of
Use (PEU)

Purchasing
Behavior (PB)
Perceived
Usefulness (PU)

Figure 3: Perceived Risk Model on the Adoption of e-Commerce

Perceived Risk in
the context of
Transaction (PRT)

Purchasing
Behavior (PB)
Perceived Risk
with Product/
Service (PRP)

Figure 4: e-Commerce Adoption Model (e-CAM)

Perceived Risk in
the context of
Transaction (PRT)

Adapted TAM Model

Perceived Ease
of Use (PEU)

Purchasing
Behavior (PB)

Perceived Risk with

Product/Service
(PRP)

Perceived
Usefulness (PU)

- 33 -

Figure 5: Result of testing the e-CAM model (USA)

-.293*
(p=.014)

PRT
PEU

.179
(p=.486)

PB

.932***
(p=.000)

.581***
(p=.000)

PU
(R2=.560)

(R2=.308)
.498*
(p=.019)

PRP
-.442*
(p=.011)
Siginificant Path ( ***: p<.001; **: p < .01; *: p<.05)
Non-significant Path (p > .05)

Figure 6: Result of testing the e-CAM model (Korea)

-.132
(p=.113)

PRT
PEU

.538***
(p=.000)

.327*
(p=.012)

PB

.824***
(p=.000)

PU
(R2=.556)

(R2=.168)
.072
(p=.513)

PRP
-.150
(p=.217)
Siginificant Path ( ***: p<.001; **: p < .01; *: p<.05)
Non-significant Path (p > .05)

- 34 -