Risk-Focused e-Commerce Adoption Model - MISRC · Risk-Focused e-Commerce Adoption Model ... but also suggest that online firms should consider ... (Agarwal and Prasad, 1998;

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: [email protected]

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 consumer’s 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, Cross-Country Study.

- 1 -

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).

- 2 -

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 consumer’s

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 consumer’s

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 consumer’s 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 consumer’s

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

- 3 -

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 consumer’s 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 --

- 4 -

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 individual’s 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 TAM’s 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

- 5 -

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.


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 store’s 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

consumer’s 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

- 6 -

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 customer’s 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.

- 7 -

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 consumer’s

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 consumer’s 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.


- 8 -

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 consumer’s perceived risk as well as

TAM.

3. Research Framework

3.1. Research Hypotheses

Many factors positively or negatively influence a consumer’s 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 e-

Commerce transactions, the lower the usage of e-Commerce will be for the consumer. In this context, the

- 9 -

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 consumer’s 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 consumer’s

adoption of e-Commerce. Therefore, we propose the following hypothesis regarding PRP.

H-2: Perceived risk with product/service (PRP) negatively affects consumer’s purchasing behavior in e-

Commerce (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).

- 10 -

H-5: Perceived ease of use (PEU) positively affects consumer’s 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: Perceived usefulness (PU) positively affects consumer’s purchasing behavior in e-Commerce (PB).

3.2. 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 e-

Commerce 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 consumer’s

adoption of e-Commerce. The relationships constituting the model also have support from prior theoretical

and empirical work in the previous research.


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 Davis’s 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 e-

Commerce 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

- 11 -

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 consumer’s 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

- 12 -

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 GVU’s 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 GVU’s surveys (87.8%).


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

- 13 -

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

reported in Table 4.


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.


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 Cronbach’s alpha (Cronbach and

Meehl, 1955) test to the individual scales and the overall measure. As the Cronbach’s 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

- 14 -

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.


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.


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;

- 15 -

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 model’s 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.


A Chi-square (χ2) statistic indicates that our model does not fit the U.S. data (χ2 = 243.11; p≤ .05) and

Korean data (χ2=372.80; p ≤ .05). However, as χ2 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 χ2 /degree of freedom (DF), goodness-

of-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 χ2/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).

- 16 -

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 e-

Commerce Adoption, i.e. purchasing behavior, as reported in Table 9.


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: β = -.293, p = .014; PRP: β = -.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 (β = .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: β

= .932, p = .000), yet the direct effect of PEU on the e-Commerce adoption is insignificant (H-5: β = .179,

p = .486). However, it should be noted that PEU has significant indirect effect (.465) on the adoption of e-

Commerce as reported in Table 9. Table 10 summarizes the test results and regression statistics in the U.S.

dataset to test the hypotheses.


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

consumer’s purchasing behavior. The data show that PRT (H-1) and PRP (H-2) has little effect on PB

- 17 -

(PRT: β = -.132, p = .113; PRP: β = -.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 (β = .072, p = .513) while PEU has significantly direct/indirect effects on the

consumer’s purchasing behavior. As represented in Table 11, PEU has strong positive effects on PU (H-4:

β = .824, p = .000), and the direct effect of PEU on the e-Commerce adoption is also significant (H-5: β

= .327, p = .012). Therefore, it should be noted that PEU has significant direct effect on the adoption of e-

Commerce 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 consumer’s adoption of e-Commerce. On the contrary, only the PEU construct of the

model in Korean dataset has a significant direct effect on the consumer’s purchasing behavior as illustrated

in Figure 6. With regard to the PEU construct, direct effect (β = .179) on the adoption is not significant in

the U.S. dataset while significant direct effect (β = .327) is shown in the Korean dataset. However, PEU in

the U.S. dataset mostly has indirect effect (β = .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 consumer’s 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

- 18 -

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 e-

Commerce. 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 consumer’s adoption of e-Commerce while PEU construct only directly affect the

adoption in the Korean dataset.



6. Discussion

In this research, we attempt to identify valid factors that predict a consumer’s online purchasing behavior

based on the e-CAM model, which integrates TAM with the theories of perceived risk to explain the

consumer’s adoption of e-Commerce. The results of this study suggest that a maturity level of e-

Commerce 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

- 19 -

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 consumer’s 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 consumer’s perceived risk, the consumer is willing to purchase

product/service from an online vendor that is perceived low risk, even if the consumer’s 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

- 20 -

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 consumer’s 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

- 21 -

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.

- 22 -

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.

- 23 -

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 Task-

technology 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.

- 24 -

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-of-

use," 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.

- 25 -

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-to-

Consumer 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.

- 26 -

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.

- 27 -

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.

- 28 -

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, PEU→U Validation of PU and PEU, primary determinants of user acceptance.

Davis et al. (1989)

PEU, PU, A, BI, U

Word PEU→PU / PU, PEU→A A, PU→BI→U

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, PEU→U 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 PEU↔PU / PU, PEU→U Replication of Davis’s previous work on TAM.

Hendrickson et al. (1993)

PEU, PU Spreadsheet, DB

Reliability of PEU/PU→High Test-retest reliability of PU and PEU.

Segars & Grover (1993)

PEU, PU, EF E-mail, V-mail PEU↔EF↔PU↔PEU Three factor model with eight items: PEU, PU, & EF.

Subramanian (1994)

PEU, PU, U V-mail, Dial up system

PU→U Replication of the impact of PU and PEU constructs on predicted future usage.

Szajna (1994, 1996)

PEU, PU, BI, U

PC-based DBMS

PEU→PU→BI→U Measurement of actual user choice rather than self-reported intentions.

Chin & Todd (1995)

PEU, PU, EF V-Mail Davis’s PU→ Unidimensional Re-examining PU and EF used in (Segars and Grover, 1993).

Straub et al. (1995)

PEU, PU, SPIR, U

V-mail SPIR→PU PU, PEU→U

Comparison of subjective and objective measures of system usage.

Taylor & Todd (1995)

PEU, PU, A, SN, PBC, BI, B

Computing resource center

PEU→PU / PU, PEU→A A, SN, PBC→BI / BI, PBC→B

Differences in the relative influence of the determinants of usage depending on experience.

Venkatesh & Davis (1996)

SE, OU, PEU, DE

Word, Lotus, Graphics, etc.

SE→PEU / OU + DE→PEU Focus on understanding the determinants of PEU: SE, OU, and DE.

Straub et al. (1997)

PEU, PU, U E-mail PEU, PU→U Comparison of the TAM across cultures.

Morris & Dillon (1997)

PEU, PU, A, BI, U

Netscape PEU→PU, A / PU→A PU, A→ BI→U

Significant influence of initial perceptions of Netscape’s PU and PEU on user’s attitude and BI.

Igbaria et al. (1997)

IV, EV, PEU, PU, U

Personal Computer

IV, EV, PEU→PU EV→PEU / PEU, PU→U

Significant effect of user characteristics, system quality, and organizational support on PEU & PU.

Gefen & Straub (1997)

G, PEU, PU, SPIR, U

E-mail G→SPIR, PU, PEU SPIR→PU→U

Extension of the TAM and SPIR addendum by adding gender to an IT diffusion model

Fenech (1998) PEU, PU, A, SE, U

WWW PEU→PU / PEU, PU, SE→U Inclusion of Consumer Self-Efficacy to improve the prediction for WWW.

Gefen & Keil (1998)

PDR, PEU, PU, U

Expert System PDR→PEU, PU PEU→PU→U

An extension of TAM based on Social Exchange Theory.

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

Graphics, Word, DB, etc.

G→PU / DE→PEU Type of IS→PEU

Support for the validity and reliability of Davis’s PU and PEU instrument.

Teo et al. (1999) PEU, PU, PE, U

Internet PEU→PU, U, PE PU, PE→U

Focus on both intrinsic (PE) and extrinsic (PU) motivation for the use of Internet.

Agarwal & Prasad (1999)

ID, PEU, PU, A, BI

Personal Computer

ID→PEU, PU / PEU→PU, A PU→A, BI / A→BI

Intervention of beliefs or perceptions (in TAM) between individual difference and IT acceptance.

Hu et al. (1999) PEU, PU, A, BI

Telemedicine PU→A→BI / PU→BI Examinations of physician’s decisions to accept telemedicine IT in the health-care context.

Venkatesh (1999) PEU, PU, BI, TI

Internet-based application

TI→PEU / TI, PEU, PU→BI 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 ACC→PEU / PEU, SP, SI→PUPEU, PU→U

Explanation for the psychological origin of PU and PEU based on TRA, SP/SI theory.

Lederer et al. (2000)

PEU, PU, U WWW PEU, PU→U Validation of TAM in the context of the WWW.

Venkatesh & Morris (2000)

PEU, PU, SN, BI, G, DE

New SW system

PU, PEU, SN→BI Longitudinal examination with the integration of SN into TAM using gender as a moderator.

- 29 -

Research Constructs Type of IS Results Characteristics

Venkatesh & Davis (2000)

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

New SW system

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

A theoretical extension of TAM that explains PU and usage BI in terms of social influence and cognitive instrumental processes.

Lin & Lu (2000) Q, PU, PEU, A, BI

WWW Q→PEU, PU→A→BI PEU→PU→BI

Investigation on user’s perception about WWW in view of IS quality.

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 consumer’s self image. Physical Risk The risk to the buyer’s or other’s 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 one’s 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 USA (N = 175) Korea (N = 267)

Gender • Male 117 (66.5%) 183 (68.5%) • Female 59 (33.5%) 84 (31.5%)

Age • 16 – 25 104 (59.1%) 182 (68.2%) • 26 – 35 53 (30.1%) 76 (28.5%) • 36 – 45 14 (8.0%) 6 (2.2%) • 46 – 55 4 (2.2%) 2 (0.7%) • Over 55 1 (0.6%) 1 (0.4%)

Education • High School or equivalent 13 (7.4%) 6 (2.2%) • Vocational/Technical School 1 (0.6%) 1 (0.4%) • Some College 98 (55.7%) 202 (75.7%) • College Graduate 32 (18.2%) 30 (11.2%) • Master’s Degree 25 (14.2%) 24 (9.0%) • Doctoral Degree 6 (3.4%) 1 (0.4%) • Other 1 (0.6%) 3 (1.1%)

- 30 -

Table 4: Internet Usage Statistics


Primary Internet Connection • 28.8-33.6k Modem 17 (9.7%) 7 (2.6%) • 56k Modem 70 (39.8%) 23 (8.6%) • Cable Modem 11 (6.3%) 29 (10.9%) • ISDN Line 4 (2.3%) 12 (4.5%) • ADSL/DSL 19 (10.8%) 92 (34.5%) • T1/T3 Line 40 (22.7%) 50 (18.7%) • Other 15 (8.5%) 64 (20.4%)

Primary Place to Use the Internet • Home 99 (56.3%) 150 (56.2%) • School 51 (29.0%) 42 (15.7%) • Work 26 (14.8%) 57 (21.3%) • PC Parlors 0 (0.0%) 18 (6.7%)

Internet Using Experience • Less than 3 months 0 (0.0%) 4 (1.5%) • 3–6 months 0 (0.0%) 18 (6.7%) • 6–12 Months 0 (0.0%) 58 (21.7%) • 1–2 years 11 (6.3%) 73 (27.3%) • More than 2 years 165 (93.7%) 114 (42.7%)

Internet Using Frequency • A few times a month 0 (0.0%) 1 (0.4%) • Once a week 0 (0.0%) 1 (0.4%) • A few times a week 5 (2.8%) 18 (6.7%) • 1–2 times a day 37 (21.1%) 113 (42.3%) • More than twice a day 134 (76.1%) 134 (50.2%)

Internet Using Hours per Week • Less than 1 hours 2 (1.1%) 17 (6.4%) • 1–5 hours 42 (23.9%) 78 (29.2%) • 6–10 hours 45 (25.6%) 68 (25.5%) • 11–20 hours 36 (20.5%) 46 (17.2%) • More than 20 hours 51 (29.0%) 58 (21.7%)

Table 5: On-line Purchasing Statistics


Electronic Shopping-Mall Visiting Experience • Yes 164 (93.2%) 225 (84.3%) • No 12 (6.8%) 42 (15.7%)

Electronic Purchasing Amount (Last 6 months) • None 22 (12.5%) 115 (43.1%) • Less than $50 19 (10.8%) 67 (25.1%) • $50–$99 21 (11.9%) 20 (7.5%) • $100–$199 24 (13.6%) 23 (8.6%) • $200–$499 37 (21.0%) 22 (8.2%) • $500–$999 21 (11.9%) 9 (3.4%) • $1,000–$2,000 8 (4.5%) 2 (0.7%) • More than $2,000 24 (13.6%) 9 (3.4%)

Electronic Purchasing Times (Last 6 months) • None 22 (12.5%) 115 (43.1%) • 1–2 times 50 (28.4%) 85 (31.8%) • 3–5 times 64 (36.4%) 40 (15.0%) • 6–10 times 20 (11.4%) 18 (6.7%) • More than 10 times 20 (11.4%) 9 (3.4%)

- 31 -

Table 6: Cronbach’s Alpha for the Scales

USA KOREA

Factors and Scale Items Mean / SD Cronbach αααα Mean / SD Cronbach ααααPerceived Ease of Use * .779 .754• Easy to search and locate desired information 5.39 / 1.36 4.43 / 1.41 • Easy to use from any location at any time 5.89 / 1.40 5.20 / 1.48 • Easy to use the customer service 4.39 / 1.50 3.90 / 1.52 • Overall PEU 5.55 / 1.18 4.37 / 1.28

Perceived Usefulness * .843 .824• Save money 5.11 / 1.09 4.67 / 1.44 • Save time 5.73 / 1.19 4.80 / 1.53 • Provide wide variety of products/ services 5.44 / 1.29 4.06 / 1.16 • Overall PU 5.82 / 1.08 4.72 / 1.12 Perceived Risk with Product/Service * .811 .812• Functional loss 4.98 / 1.26 4.55 / 1.29 • Time loss 5.32 / 1.34 5.03 / 1.19 • Financial loss 5.24 / 1.57 4.86 / 1.25 • Opportunity loss 4.82 / 1.35 5.02 / 0.98 • Overall PRP 4.80 / 1.06 4.92 / 1.07 Perceived Risk in the Context of Transaction * .875 .810• Privacy 5.14 / 1.48 5.85 / 1.27 • Security (Credit card) 4.30 / 1.82 5.29 / 1.56 • Non-repudiation 4.09 / 1.68 4.75 / 1.35 • Overall PRT 4.02 / 1.63 4.30 / 1.51 Purchase Behavior .827 .907• Total Amount of Online Purchasing ** 4.53 / 2.12 2.44 / 1.83 • Frequency of Online Purchasing *** 2.91 / 1.16 1.96 / 1.08

*: 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 USA KOREA Measured Items

PRT PRP PU PEU PB PU PRT PRP PEU PB Ease of search -.15 -.01 .28 .69 -.05 .34 -.12 .07 .66 .07Ease of ordering .04 -.15 .40 .67 .15 .29 .00 .07 .58 .21Customer service -.19 .04 .01 .72 .22 .02 -.01 -.09 .82 -.01Overall ease of use -.07 -.04 .43 .75 .14 .48 -.12 -.01 .67 .152Save money -.15 .06 .53 .34 .44 .75 -.10 .08 .23 .04Save time -.10 -.11 .73 .33 .11 .80 .17 .07 -.02 .128Variety of products -.26 -.08 .74 .16 .05 .69 -.08 -.02 .30 .01Overall usefulness -.18 -.05 .78 .36 .25 .84 -.05 .10 .29 .127Functional loss .06 .79 .00 .00 -.13 .02 -.09 .77 .07 -.117Time loss .15 .80 -.05 -.10 .04 .11 .11 .81 .05 .06Financial loss .34 .51 .25 -.15 -.26 .06 .26 .77 -.12 .03Opportunity loss .24 .65 -.30 .20 -.13 .03 .15 .65 .05 -.03Overall PRP .32 .84 -.09 -.08 -.17 -.02 .49 .66 -.13 -.14Privacy .76 .26 .05 -.16 -.03 .03 .72 .15 .06 .01Security (Credit card) .81 .16 -.25 -.01 -.10 .04 .86 .09 -.04 .04Non-repudiation .78 .22 -.24 -.11 -.19 .00 .77 .20 -.14 -.14Overall PRT .76 .21 -.26 -.17 -.25 -.19 .74 .08 -.10 -.22Purchasing Times -.15 -.21 .18 .13 .87 .13 -.11 -.08 .17 .92Purchasing Amount -.23 -.20 .18 .16 .86 .13 -.12 -.04 .08 .92Eigenvalue 6.87 2.96 1.36 1.20 1.02 4.65 3.81 1.66 1.48 1.09Cum. % of Variance 36.1 51.7 58.9 65.2 70.6 24.5 44.5 53.2 61.0 66.8

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

- 32 -

Table 8: Fit Measures for the e-CAM Model

Fit Measures Fit Guidelines USA KOREA

χ2 p ≥ .05 243.11 (p=.00) 372.80 (p=.00) χ2 / DF ≤ 3.0 1.665 2.553 GFI ≥ .90 .880 .863 AGFI ≥ .80 .844 .822 NFI ≥ .90 .877 .841 NNFI ≥ .90 .937 .901 CFI ≥ .90 .946 .903 RMSEA ≤ .08 .062 .076

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

Perceived Usefulness (PU) Purchasing Behavior (PB) USA KOREA USA KOREA

Direct Total Direct Total Direct Indirect Total Direct Indirect Total

PRT1) -.293 -.293 -.150 -.150 PRP2) -.442 -.442 -.132 -.132 PEU .932 .932 .824 .824 .179 .465 .643 .327 .059 .386 PU .498 .498 .072 .072 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 Causal

Relationship β1) S.E.2) C. R.3) P4) Result H-1 PRT→→→→PB (−−−−) -.293 .119 -2.459 .014 Supported H-2 PRP→→→→PB (−−−−) -.442 .175 -2.533 .011 Supported H-3 PRT↔↔↔↔PRP (++++) .581 .141 5.391 .000 Supported H-4 PEU→→→→PU (++++) .932 .122 7.631 .000 Supported H-5 PEU→PB (+) .179 .257 .696 .486 Rejected H-6 PU→→→→PB (++++) .498 .213 2.339 .019 Supported

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

Table 11: Hypotheses Testing Results (Korea)

Hypothesis Causal

Relationship β1) S.E.2) C. R.3) P4) Result H-1 PRT→PB (−) -.132 .083 -1.586 .113 Rejected H-2 PRP→PB (−) -.150 .122 -1.233 .217 Rejected H-3 PRT↔↔↔↔PRP (++++) .538 .077 5.315 .000 Supported H-4 PEU→→→→PU (++++) .824 .084 9.812 .000 Supported H-5 PEU→→→→PB (++++) .327 .130 2.519 .012 Supported H-6 PU→PB (+) .072 .110 .654 .513 Rejected

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

- 33 -

Figure 1: Davis’s Technology Acceptance Model

ExternalVariables

PerceivedUsefulness (PU)

Perceived Ease ofUse (PEU)

Attitude TowardUsing (A)

BehavioralIntention to

Use (BI)

ActualSystem Use

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

Perceived Ease ofUse (PEU)


PurchasingBehavior (PB)

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

Perceived Risk inthe context of

Transaction (PRT)

Perceived Riskwith Product/Service (PRP)


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

Perceived Easeof Use (PEU)



Perceived Risk withProduct/Service

(PRP)

Perceived Risk inthe context of

Transaction (PRT)

Adapted TAM Model

- 34 -

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

PEU

PU(R2=.560)

PB(R2=.308)

PRP

PRT

Siginificant Path ( ***: p<.001; **: p < .01; *: p<.05)

Non-significant Path (p > .05)

-.293*(p=.014)

-.442*(p=.011)

.932***(p=.000)

.498*(p=.019)

.179(p=.486)

.581***(p=.000)

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

PEU

PU(R2=.556)

PB(R2=.168)

PRP

PRT

Siginificant Path ( ***: p<.001; **: p < .01; *: p<.05)

Non-significant Path (p > .05)

-.132(p=.113)

-.150(p=.217)

.824***(p=.000)

.072(p=.513)

.327*(p=.012)

.538***(p=.000)

Documents

Risk-Focused e-Commerce Adoption Model - MISRC · Risk-Focused e-Commerce Adoption Model ... but also suggest that online firms should consider ... (Agarwal and Prasad, 1998;