The role of experience and innovation characteristics in

the adoption and continued use of e-learning websites

Hsiu-Li Liao

, Hsi-Peng Lu

Department of Information Systems, National Taiwan University of Science and Technology, No. 43, Sec.4,

Keelung Road, Taipei 106, Taiwan, ROC

Received 21 August 2007; received in revised form 27 October 2007; accepted 13 November 2007

Abstract

With the advent of e-learning technologies in the past decade, the accessibility of training, teaching, and learning has

drastically increased. The challenge for the education enterprise now is how to attract learners to their e-learning services.

In this study, a technology adoption model is developed to predict the users’ intention of adoption and their continued use

behavior. The results show signiﬁcant evidence in support of the hypothesis. The ﬁndings indicate that perceptions of rel-

ative advantage and compatibility are signiﬁcantly related to users’ intention to use e-learning. Also, the intention is sig-

niﬁcantly related to their actual use of e-learning. Furthermore, the technology adoption of learners with prior e-learning

experience is diﬀerent from those without prior e-learning experience. These ﬁndings may contribute to deeper understand-

ing of e-learning users’ perceptions in terms of adoption and continued use behavior.

Keywords: Distance education and telelearning; Teaching/learning strategies; Lifelong learning

1. Introduction

E-learning has grown into a revolutionary way of learning due to the rapid development of information and

communication technologies (Cappel & Hayen, 2004). The technological innovations have made training,

teaching, and learning over the Internet possible, which is so-called Web-based instruction (WBI) in education

and training ﬁelds (Lee, 2001). E-learning has received considerable attention as a means of providing alter-

natives to traditional face-to-face, instructor-led education (Douglas & Van Der Vyver, 2004). IDC, the IT

intelligence analyst puts the total global e-learning market at $8 billion and set to grow to $13 billion in

the next ﬁve years (Flood, 2006). In Taiwan, the Institute for Inform ation Industry (III) reported that the

domestic market for e-learning services has grown from $17 million in 2001 to $209 million in 2005 (UDN,

2006). With the continuous growth of the e-learning market, however, there is a lack of discussion on the indi-

viduals’ behavior in the adoption and continued use of e-learning.

doi:10.1016/j.compedu.2007.11.006

Corresponding author. Tel.: +886 3 2655411; fax: +886 3 2655499.

E-mail addresses: wenlly.liao@msa.hinet.net, wenlly@mis.cycu.edu.tw (H.-L. Liao), hsipeng@cs.ntust.edu.tw (H.-P. Lu).

Available online at www.sciencedirect.com

Computers & Education 51 (2008) 1405–1416

www.elsevier.com/locate/compedu

E-learning, by this deﬁnition, can be considered an information technology (IT) innovation for many learn-

ers. According to Tornatzky and Fleischer’s (1990) deﬁnition, innovation is ‘‘the situationally new develop-

ment and introduction of knowledge-derived tools, artifacts, and devices by which people extend and

interact with their environment” (p. 10). Colleges, universities, and private training companies are converting

more and more courses online. With a PC con nected to the Web, e-learning allows students to attend courses

anywhere at any time. Before the telecommunication technologies became easily accessible, distance learning

was delivered using mail correspondence. For interactivity, learners would have to mail back the assignments

to receive feedback from the instructor or go to the tutoring center for face-to-face interaction experience.

Today, e-learning (or Web-based learning) has made distance learning highly interactive. Students may obtain

an intimate learning experience without attending a brick-and-mortar facility.

Many of the prior research on e-learning are concerned with learning eﬀectiveness and performance. David

and Glen described the results of a designed experiment to measure the learning eﬀectiveness of making all the

text’s multiple choice questions with answers, available to oﬀ-campus students enrolled in the database unit in

an e-learning environment. The approach improves performance on the ﬁnal examination not only on the

multiple choice questions but also on the theory questions (Douglas & Van Der Vyver, 2004). James and

Roger investigated the users’ previous online learning experiences, satisfaction, perceived eﬀectiveness and

quality of the online learning units in an online self-paced independent study courses as a project within a tra-

ditional, instructor-led graduate course (Cappel & Hayen, 2004). Huang and Cappel evaluated the use, satis-

faction, and perceived beneﬁts of online learning games ( Huang & Cappel, 2005). Robert, Kimberly, and Jean

sought to examine the eﬀects of cognitive load on performance outcomes of students using computer-based

instruction. The performance of students in a cooperative group interacting in a computer-based instructional

environment was compared to the performance of students working alone (Fuller, Vician, & Brown, 2006 ).

However, few previous studies have investigated how the users’ perceptions of innovation characteristics inﬂu-

ence the use of e-learning.

This study focuses on individual users’ perceptions of innovation characteristics (PCI) of e-learning as

explanatory and predictive variables for their adoption and continued use behavior. Two speciﬁc questions

therefore guided this research: First, can the perception variables of innovation characteristics (PCI) predict

individual’s intention to use an e-learning website? While some studies have used innovation characteristics to

explain users’ acceptance behavior in speciﬁc contexts successfully (Jurison, 2000; Van Slyke, Belanger, &

Comunale, 2004; Ilie, Van Slyke, Green, & Lou, 2005; Lin & Lee, 2006), few research have been done in rela-

tion to e-learning adoption factors. The second research question is whether the technology adoption model of

learners experienced in e-learning is diﬀerent from inexperienced learners. While PCI variables are examined

against the adoption of e-learning of inexperienced users and the continued use of the experienced users, it is

expected that perceptions diﬀer or change with experience (Fazio, 1989).

2. Theoretical background and research model

For research in technology adoption, the technology accepta nce model (TAM) has received considerable

attention. This model proposes two key beliefs in the adoption of technology: perceived usefulness (PU)

and perceived ease of use (PEU). While TAM focuses on the technology adoption process, later innovation

research further distinguish between initial adoption of innovation and continued/sustained use of innovation

(Rogers, 1983). This study pro poses a theoretical framework for user adoption behavior of e-learning based

on the theory of perceptions of innovation characteristics (PCI). Both e-learning adoption behavior and con-

tinued use will be discussed.

2.1. Perceptions of innovation characteristics

Rogers (1983) identiﬁed attributes of innovation that are key to acceptance behavior including relative

advantage, complexity, compatibility, trialability, and observability. Based on the work of Rogers and others,

Moore and Benbasat (1991) expanded the innovation characteristics into seven constructs and developed an

instrument to measure the perceptions. The constructs include relative advantage, ease of use, compatibility,

image, result demonstrability, visibility and trialability.

1406 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416

Rogers’ and Moore and Benbasat’s ﬁrst construct of PCI, relative advantage, is similar to the concept of

perceived usefulness of TAM (Davis, Bagozzi, & Warshaw, 1989). Relative advantage presents the degree to

which a potential adopter views the innovation as oﬀering an advantage over previous ways of pe rforming the

same task. The second construct, ease of use, which is also part of the TAM, is similar to deﬁnition to Rogers’

concept of complexity (Rogers, 1983) and captures the degree to which a potential adopter considers use of the

target system to be relatively free of eﬀort (Davis, 1993).

Moore and Benbasat’s construct of compatibility is inline with Rogers’: ‘‘the degree to which an innovation

is perceived as being consistent with the existing values, needs, and past exp eriences of potential adopters.”

Moore and Benbasat (1991) indicated that the image construct, which is as part of relative advantag e in Rog-

ers’ framework, can independently predict innovation use. The image construct refer s to the perception when

using an innovation in terms of enhancing the social status of the potential adopter. Further, Rogers’ attribute

of observability is divided into two constructs of result demonstrability and visibility, which refers to ‘‘the tan-

gibility of the results of using an innovation” and ‘‘the extent to which potential adopters see the innovation as

being visible in the adoption context” respectively. Finally, trialability indicates the perception of potential

adopters of an opportunity to try the innovation before committing to its use.

2.2. Perceptions of innovation characteristics and intention of use

Researches in PCI indicate that individuals’ perceptions about the characteristics of an innovation signif-

icantly aﬀect their acceptance behavior. Such discussion on percept ions has been persistent in research liter-

ature in system use (Moore & Benbasat, 1991; Davis, 1993) and use intentions (Agarwal & Prasad, 1997;

Van Slyke, Lou, & Day, 2002; Van Slyke et al., 2004; Lin & Lee , 2006; Ilie et al., 2005).

Innovation characteristics researches often include a hypothesized relationship between user perceptions

and use. There are also inconsistent empirical studies regarding the signiﬁcance of perceptions. For exam-

ple, Van Slyke et al. (2002) used Rogers’ diﬀusion of innovation theory to investigate users’ intention in

using a speciﬁc groupware application (Lotus Domino) discussion database. The ﬁndin gs indicate that per-

ceptions of relative advantage, complexity, compatibility and result demonstrability are signiﬁcantly related

to users’ intention to use the application. Lin and Lee (2006) , in a study of a knowledge management

system implementation, examined how perceived innovation characteristics aﬀect the intention to encour-

age knowledge sharing. In total 154 senior executives in Taiwan were included in the survey to test the

relationship between perceived innovation characteristics and intention. The results showed that perceived

relative advantage, compatibility, and complexity positively aﬀected the intention to encourage knowledge

sharing.

Based on the constructs Moore and Benbasat (1991) proposed, Van Slyke et al. (2004) studied factors that

may impact consumers’ decision to engage in Web-based shopping and found that perceived compatibilit y has

the strongest impact on intention of use, followed by perceived complexity, relative advantage and image. Ilie

et al. (2005) added to the understanding of instant messaging adoption and use by examining gender diﬀer-

ences in perceived innovation characteristics. Diﬀerent patterns were discussed in their study. For females, per-

ceived ease of use and visibility were signiﬁcant predictors of intention to use; for males, perceived relative

advantage, ease of use, and result demonstrability were signiﬁcant.

These studies conﬁrmed the innovation characteristics identiﬁed by Rogers (1983) as well as Moore and

Benbasat (1991). They explained technology adoption behavior in speciﬁc technology contexts, but the

results of salient perception factors were inconsistent. Tornatzky and Klein (1982), in a meta-analysis

of innovation characteristics research, found that only three innovation characteristics - perceived relative

advantage, perceived complexity, and perceived compatibility – are consistently related to innovation

adoption.

Few studies have addressed the adoption of e-learning in the perspective of perceived innovation charac-

teristics. As an innovative learning method, the adoption of e-learning involves the adoption of information

technology and change of learning approach. Learners have more control over selection of learning topics in

comparing to learning in the traditional classrooms. Therefore, this research represents an attempt to use

diﬀusion of innovation perspective (2003) to underst and factors that may impact the intention to use online

e-learning.

H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416 1407

2.3. Prior experience, Perceptions of innovation, and Intention to use

Studies have sho wn that prior experience can change intention of adoption or continued use of innovation.

For example, Web shoppers’ intention to con tinue shopping online is aﬀected by past Web-shopping experi-

ence and past experience with the Web (May So, Danny Wong, & Sculli, 2005). Agarwal and Prasad (1997)

focused on individual perceptions as explanatory and predictive variables of users’ World Wide Web accep-

tance and continued use behavior. Their study results indicated that the innovation characteristics of visibility,

compatibility, and trialability are relevant in explaining acceptance behavior. For continued use behavior,

their study suggested that only advantage and result demonstrability were relevant innovation characteristics.

Agarwal and Prasad also suggested that there were diﬀerences in perceived characteristics between acceptance

and continued use of innovation.

Based on the above review of literature, an adoption model for e-learning users is proposed. Given that the

learners experienced in e-learning are diﬀerent from inexperienced learners. The hypotheses are formulated as

below to address the diﬀerence between e-learning users’ innovation adoption and continued use:

H1: The perceptions of innovation characteristics of Web-based e-learning are positively related to users’

intention to adopt or continue to use e-learning.

H1a: The perception of ease of use is positively related to user’s intention about e-learning adoption/con-

tinued use.

H1b: The perception of relative advantage is positively related to user’s intention about e-learning adop-

tion/continued use.

H1c: The perception of compatibility about using a Web e-learning system is positively related to user’s

intentions about e-learning adoption/continued use.

H1d: The perception of trialability is positively related to user’s intentions about e-learning adoption/

continued use.

H1e: The result demonstrability is positively related to user’s intention about e-learning adoption/con-

tinued use.

H1f: The perception of visibility of the innovation is positively related to user’s intention about e-learn-

ing adoption/continued use.

H1g: The perception of image is positively related to user’s intention about e-learning adoption/contin-

ued use.

H2: The user’s intention is positively related to the actual use of a Web-based e-learning system.

H3: The eﬀect of percei ved innovation characteristics on user’s intention to use Web-based e-learning is dif-

ferent between learners with and without previous e-learning experience.

2.4. The research model

Fig. 1 represents the proposed research model drawn from the constructs of perceived innovation charac-

teristics, intention of adoption and continued use, and the role of experience in adoption and continued use of

innovation as discussed above. The research model is empirically tested in this study. In this model, perceived

innovation characteristics comprise of seven user perceptions-ease of use, relative advantage, compatibility ,

trialability, result demonstrability, visibility and image. It is proposed in this model that prior experience

be a potential determinant of e-learning website use as the independent variable for this study. The deﬁnitions

of the constructs are provided in Table 1 and discussed in the subsequent discussions.

3. Research methodology

3.1. Characteristics of the sample and study context

To test the research model, an e-learning website was chosen as a representative of the innovation of e-

learning. A survey was conducted on students who were enrolled in a project management (PM) course at

1408 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416

a comprehensive university in Taiwan. Digital materials related to conducting PM on Microsoft Project 2003

were developed and students can refer to them on the e-learning website.

The system used in the experiments was designed explicitly for this research. It ran on a Pentium IV PC with

a15

monitor. Subjects used Internet Explorer 6 to browse the teaching materials stored on a university server.

With this conﬁguration, retrieval of information, including video clips, was almost instantaneous. The e-

Learning website development was using the Wisdom M aster. Developed by SUN NET Technology Corpo-

ration, Wisdom Master is one of the most popularly used platforms of Learning Management System (LMS)

in Taiwan. Wisdom Master is also the ﬁrst software in Taiwan that meets the highest standard (RTE3) of the

SCORM 1.2. The synchronous mode of teaching is not always superior to the asynchronous mode (Palvia,

2000). A majority of e-learning is conducted in an asynchronous mode (Douglas & Van Der Vyver, 2004).

Therefore, we developed e-learning system with the asynchronous mode. With the high-resolution monitor

employed, subjects could clearly see the facial expressions of the people in the video clips on the e-learning

website.

Subjects received a one-hour, hand-on demonstratio n on using e-Learning website before the course actu-

ally began. Subjects can use the e-learning web system free through connecting Internet from anywhere at any

The Technology Adoption Model for Users without Prior Experience

The Technology Adoption Model for Users with Prior Experience

Compatibility

Relative

Advantage

Trialability

Ease of Use

Result

Demonstrability

Visibility Image

Actual

Adoption

Intentions

of Adoption

Compatibility

Relative

Advantage

Trialability

Ease of Use

Result

Demonstrability

Visibility Image

Intentions of

Continued Use

Fig. 1. The research model.

H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416 1409

time. After they ﬁnished the 4-week course free, the students were asked to complete a questionnaire survey.

Completion of the survey was voluntary and could be done outside class. A total of 137 surveys were com-

pleted. The age range of the sample was 20–30 years old. Of the 137 respondents, 59 were females (43%)

and 78 were males (57%). Among these respondents, 41 did not have any previous e-learning experience.

3.2. Instrument development

The survey questionnaire included a combination of items derived from earlier studi es and newly developed

items. Moore and Benbasat’s (1991) questionnaire of scales of perceptions of innovation characteristics was

used as the foundation for the development of the survey instrument. It included a total of 25 items with each

scale consists of a minimum of two items. Additionally, the intention of adoption and intention of continued

use were assessed using two items constructed following the recommendations of Davis et al. (1989). The

actual adoption and continued use were measured using two items sim ilar to those in Davis (1993). As indi-

cated by Davis, such self-reported measures are reasonable indicators of relative system use. Respondents

scored on a seven-point Likert-ty pe scale with the end points being ‘‘strongly disagree” and ‘‘strongly agree”,

except for items intended to collect demographic data.

3.3. Measures

The constructs of reliability and validity of the instrument were evaluated. Table 2 shows the number of

items comprising each scale: the reliability reported by Moore and Benbasat (1991) for the scale and Cron-

bach’s alpha for scale reliability obtained for the samples. Factor reliabilities, as represented by Cronbach’s

Table 1

Research variables and deﬁnitions

Research variables Deﬁnition

Ease of use Degree to which an innovation is considered by a learner as relatively easy to use and understand

Relative advantage Degree to which an innovation is considered as superior to its predecessor

Compatibility Degree to which an innovation is considered as compatible with existing values, beliefs, experiences and needs

of learners

Trialability Based on learners’ perceptions of the degree to which an innovation can be used on a trial basis before

conﬁrmation of the adoption must occur

Result demonstrability Degree to which the results of using an innovation are perceived to be tangible

Visibility The perception of the actual visibility of the innovation itself as opposed to the visibility of outputs

Image The perception that using an innovation helps enhance or improve the social status of users

Intention of adoption Degree to which a learner without prior experience of e-learning intends to switch over to the innovation or

increases his use in the future

Intention of continued

use

Degree to which a learner with prior experience of e-learning intends to switch over to the innovation or

increases his use in the future

Actual adoption Degree to which a learner without prior experience of e-learning uses the e-learning website actually.

Continued use Degree to which a learner with prior experience of e-learning uses the e-learning website continuously.

Prior experiences Whether learners have experiences in e-learning websites

Table 2

Scale reliabilities

Scale Number of items Moore and Benbasat reliability Reliability from our sample

Ease of use 4 0.84 0.86

Relative advantage 5 0.90 0.94

Compatibility 3 0.86 0.91

Trialability 2 0.71 0.82

Result demonstrability 4 0.79 0.87

Visibility 2 0.83 0.94

Image 3 0.79 0.96

Use intention 3 0.94

Actual use 2 0.81

1410 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416

alpha in Table 2, were between 0.81 and 0.94 for each factor. Reliability from the sample showed a reasonable

level of reliability (a > 0.70) ( Cronbach, 1970).

Factor analysis also conﬁrmed that the construct validity of the scales could be measured adequately. Using

the principal components method with varimax rotation, construct validity was examined. Table 3 reports the

factor loadings and explained varian ce for each of the factors. Bagozzi and Yi (1988) suggested that factor

loadings for each item should be over 0.6 to be valid. The factor loadings for all items exceeded 0.7 in this

study, which indicated that the individual items had discriminant validity.

4. Analysis and results

Pearson correlation coeﬃcients for all research variables are sho wn in Table 4. Although Table 4 indicates

that most user perceptions are signiﬁcantly correlated with each other, an examination of the variance inﬂa-

tion factors indicated that the multicollinearity was not signiﬁcant ( Meuter, Jo Bitner, Ostrom, & Brown,

2005). Hence, residual analysis was also conducted to verify the assumptions underlying stepwise regression

analysis. Stepwise regression enters independent variables into the regression equation one at a time, starting

with the independent variable that best predicts the dependent variable. Independent variables are entered into

the equation until adding the next best predictor does not signiﬁcantly improve the explanatory power of the

equation. All assumptions were conﬁrmed.

The next step in the analysis was to test the signiﬁcance of the individual terms in the model. These tests

correspond to H1 – H2. The null hypotheses tested, the t statistic, and signiﬁcance level are listed in Table

5 and illustrated in Fig. 2, as is whether the hypothesis was supported (alpha < 0.05). The result for use

intentions of all sub jects indicate that the innovation characteristics of relative advantage (b = 0.470,

Table 3

Summary statistics and factor loadings for all constructs

Construct Item Mean Standard deviation Factor loadings Explained variance

Ease of use V6 4.8012 1.1738 0.766 70.80%

V7 0.865

V8 0.874

V9 0.857

Compatibility V10 4.6374 1.0710 0.897 86.12%

V11 0.945

V12 0.941

Image V13 3.9737 2.0090 0.951 93.09%

V14 0.962

V15 0.962

Result demonstrability V16 4.9561 0.6844 0.917 80.16%

V17 0.897

V18 0.872

V19 0.853

Visibility V20 4.7851 1.4688 0.945 94.97%

V21 0.951

Trialability V22 4.5702 1.1259 0.924 85.37%

V23 0.924

Relative advantage V24 4.6857 1.1066 0.908 78.05%

V25 0.927

V26 0.941

V27 0.889

V28 0.794

V29 0.833

Use intention V30 4.8012 1.1738 0.949 90.78%

V31 0.944

V32 0.945

Actual use V33 4.5789 2.2895 0.919 84.50%

V34 0.919

H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416 1411

p < 0.01), and compatibility (b = 0.393, p < 0.01) are relevant in explaining intentions. The two variables

account for 53.6% of the variance in intentions to use. User intentions (b = 0.169, p < 0.1) are related to their

actual use of a web learning system. The variable explains 34.6% of the variance in actual use.

*** p<0.01 **p<0. 05 * p<0. 1

0.470***

Use Intentions

=0.536

Actual Use

=0.346

Compatibility

Relative

Advantage

0.393***

0.169*

Trialability

Result

Demonstrability

Visibility Image

Ease of Use

nificant

ath Non-si

nificant

ath

Fig. 2. Technology adoption model of e-learning website.

Table 4

Pearson correlation coeﬃcients

Ease of

use

Compatibility Image Result demonstrability Visibility Trial. Rel.

Adv.

Use

Intent

Actual

Use

Ease of use 1.000

Compatibility 0.635

***

1.000

Image 0.303

***

0.429

***

1.000

Result demonstrability 0.459

***

0.335

***

0.346

***

1.000

Visibility 0.057 0.301

***

0.198

0.239

1.000

Trialability 0.406

***

0.496

***

0.136 0.402

***

0.287

***

1.000

Relative advantage 0.525

***

0.434

***

0.268

***

0.534

***

0.241

***

0.531

***

1.000

Use intention 0.465

***

0.598

***

0.339

***

0.470

***

0.154 0.457

***

0.641

***

1.000

Actual use 0.244 0.033 0.054 0.000 0.040 0.070 0.003 0.169

1.000

p < 0.1.

p < 0.05.

***

p < 0.01.

Table 5

Regression analysis of the technology adoption model of e-learning website

Hypothesized signiﬁcant R-square Beta

(a) Intention = relative advantage 0.411 0.641

***

Intention = relative advantage + compatibility 0.536 0.470

***

relative advantage 0.393

***

+ compatibility 0.169

(b) Actual use = use intentions 0.346

p < 0.1.

p < 0.05.

***

p < 0.01.

1412 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416

Both hypotheses about relative advantage (H1b) and compatibility (H1c) were supported, while those

related to ease of use (H1a), trialability (H1d), result demonstrability (H1e), visibility (H1f) and image

(H1g) were not supported. Relative advantage has the strongest impact on use intentions, followed by per-

ceived compatibility.

Figs. 3 and 4 graphically illustrate the signiﬁcant relationships found in our study. For users with prior

experience of using e-learning, compatibility (b = 0.580, p < 0 .01) and result demonstrability (b = 0.331,

p < 0.01) signiﬁcantly and directly inﬂuence the intention of continued use. Both variables account for

Table 6

Regression analysis of the technology adoption model for users with prior experience

Hypothesized signiﬁcant R-Square Beta

(a) Intention = compatibility 0.479 0.697

***

Intention = compatibility + result demonstrability 0.571

compatibility 0.580

***

+ result demonstrability 0.331

***

p < 0.1.

p < 0.05.

***

p < 0.01.

Table 7

Regression analysis of the technology adoption model for users without prior experience

Hypothesized signiﬁcant R-square Beta

(a) Intention = relative advantage 0.674 0.828

***

Intention = relative advantage + image 0.761

relative advantage 0.852

***

+ image 0.305

***

intention = relative advantage + image + compatibility 0.794

relative advantage 0.839

***

+ image 0.265

***

+ compatibility 0.200

p < 0.1.

p < 0.05.

***

p < 0.01.

*** p<0.01 **p<0.05 * p<0.1

0.331***

Intentions of continued use

=0.571

Compatibility

Relative

Advantage

0. 580***

Trialability

Result

Demonstrability

Visibility Image

Ease of Use

nificant

ath Non-si

nificant

ath

Fig. 3. Technology adoption model for users with prior experience.

H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416 1413

57.1% of the variation in the intention of continued use. Perceived compatibility is a better predictor than

result demonstrability (Table 6).

However, for users with no previous experience using e-learning websites, compatibility (b = 0.200,

p < 0.05), image (b = 0.265, p < 0.01), and relative advantage (b = 0.839, p < 0.01) had a signiﬁcant, direct

eﬀect on intention of adoption. These three variables explain 79.4% of the variation in intention of adoption.

Relative advantage has the strongest impact on intention of adoption, followed by perceived compatibility and

image (Table 7).

The research models with only signiﬁcant paths are included, and Beta coeﬃcients are also provided in the

Figs. 3 and 4. For those with and without e-learning experience, our study shows the impact of perceived inno-

vation characteristics on intentions to use e-learning website were diﬀerent. The hypotheses about users’ prior

experience (H3) are thus supported.

5. Conclusion and discussion

The results of the investigation indicate a relationship between user perceptions of the characteristics of the

e-learning website and their intention to use the technology. More speciﬁcally, users’ perceptions of the rela-

tive advantage and compatibility of the e-learning website exhibit signiﬁcant relationships with their adoption

intention. The results are generally consistent with prior research about other technology adoption. Relative

advantage an d compatibility have received the most consistent support as factors that inﬂuence adoption and

use of an innovation in other contexts (Van Slyke et al., 2002; Van Slyke et al., 2004; Lin & Lee, 2006). Nota-

bly, examination of the relative strengths of the associations between the individual perception variables and

adoption intention clearly indicate that perceived relative advantage and compatibility can explain much of

the variation in adoption intention. In other words, for web-learning learners, the perceptions of the relative

advantage and compatibility of web learning are better predictors than other perception constructs.

The diﬀerences in the perceptions–intention relationship for learners with and without experience in e-learn-

ing website make an argument for the consideration of an experience component associated with the technology

use model. In this study, previous e-learning experiences, compatibility and result demonstrability have a sig-

niﬁcant, direct eﬀect on user’s intention of continued use. However, for users without previous experience of

using e-learning, compatibility, image and relative advantage have a signiﬁcant, direct eﬀect on user’s intention

of adoption. When learners have more experience with e-learning, the impact of perceived innovation charac-

teristics on intentions to use web learning are diﬀerent from that of inexperienced learners. Information

obtained from experience over a period of time undoubtedly has the potential to modify future intentions of

using web learning.

*** p<0. 01 **p<0.05 * p<0. 1

0.265***

0.839***

Intentions of Adoption

=0.794

Compatibility

Relative

Advantage

0.200**

Trialability

Result

Demonstrability

Visibility Image

Ease of Use

nificant

ath Non-si

nificant

ath

Fig. 4. Technology adoption model for users without prior experience.

1414 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416

The current research can lead to several further studies. First, the dependent construct here represents

behavioral intention of initial adoption and continued use. It would be valuable that studies can be conducted

to understand potential implications experience gained over time has for the technology use model. A second

concern is that the model tested here has been empirically assessed in only one conducting context. The ger-

eralizability of the results reported here is not known beyond the current sample, e-learning context and rich-

ness antecedents. However, the proposed research model provides explanations and predictions to understand

learners’ behavior. Based on this understanding, system platform manager and education institution can

determine how to improve the learners’ initial intention and continued use of e-learning websites.

Appendix. Items and scales

Ease of use

1. My interaction with the e-learning website is clear and understandable.

2. I believe it would be easy to get the e-learning website to do what I want it to do.

3. I believe the e-learning website would be easy to use.

4. Learning to use the e-learning website would be easy for me.

Relative advantage

5. Using the e-learning website would make it easier to do my work/learning.

6. Using the e-learning website would help me to accomplish tasks more quickly.

7. Using the e-learning website would improve the quality of the work/learning I do.

8. Using the e-learning would give me greater control over my work/learning.

9. Using the e-learning would enhance my eﬀectiveness in the MIS program and/or my work.

Compatibility

10. Using the e-learning website would be compatible with all aspects of my work/learning.

11. I think that using the e-learning website would ﬁt well with the way I like to work/learn.

12. Using the e-learning website would ﬁt into my work-style/learning-style.

Trialability

13. I would be permitted to use the e-learning website on a trial basis long enough to see what it could do.

14. Before deciding to use the e-learning website, I would be able to properly try it out.

Result demonstrability

15. I would have no diﬃculty telling others about the results of using the e-learning website.

16. I believe I could communicate to others the consequences of using the e-learning website.

17. The results of using the e-learning website would be apparent to me.

18. I would have diﬃculty explaining why using the e-learning website may or may not be beneﬁcial.

Visibility

19. In the school and/or my workplace one sees the e-learning website a lot.

20. The e-learning website is not very visible in the school and/or my workplace.

Image

21. People who use the e-learning website have more prestige than those who do not.

H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416 1415

22. People who use the e-learning website have a high proﬁle.

23. Using the e-learning website is a status symbol.

Intentions to use

24. I intend to increase my use of the e-learning website for work/learning in the future.

25. I intend to use the e-learning website in the future for my work/learning.

26. For future work I would use the e-learning website.

Actual use

27. I use the e-learning website a lot to learn.

28. I use the e-learning website frequently to learn.

References

Agarwal, R., & Prasad, J. (1997). The role of innovation characteristics and perceived voluntariness in the acceptance of information

technologies. Decision Sciences, 28(3), 557–582.

Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structure equations models. Academic of Marketing Science, 16, 76–94.

Cappel, J. J., & Hayen, R. L. (2004). Evaluating e-learning: A case study. Journal of Computer Information Systems, 44(4), 49–56.

Cronbach, L. J. (1970). Essentials of psychological testing. New York: Harper and Row.

Davis, F. D. (1993). User acceptance of information technology: System characteristics, user perceptions and behavioral impacts.

International Journal of Man-Machine Studies, 38(3), 475–487.

Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models.

Management Science, 35(8), 982–1003.

Douglas, D. E., & Van Der Vyver, G. (2004). Eﬀectiveness of e-learning course materials for learning database management systems: An

experimental investigation. Journal of Computer Information Systems, 44(4), 41–48.

Fazio, R. H. (1989). On the power and functionality of attitudes: the role of attitude accessibility. In A. R. Pratkanis, S. J. Breckler, & A.

G. Greenwald (Eds.), Attitude structure and function. Hillsdale: Lawrence Erlbaum Associates.

Flood, G. (2006). Make it speciﬁc. Human Resources, 64–66.

Fuller, R. M., Vician, C., & Brown, S. A. (2006). E-learning and individual characteristics: The role of computer anxiety and

communication apprehension. Journal of Computer Information Systems, 46(4), 103–115.

Huang, Z., & Cappel, J. J. (2005). Assessment of a web-based learning game in an information systems course. Journal of Computer

Information Systems, 45(4), 42–49.

Ilie, V., Van Slyke, C., Green, G., & Lou, H. (2005). Gender diﬀerences in perceptions and use of communication technologies: A diﬀusion

of innovation approach. Information Resources Management Journal, 18(3), 13–31.

Jurison, J. (2000). Perceived value and technology adoption across four end user groups. Journal of End User Computing, 12(4), 21–28.

Lee, M.-G. (2001). Proﬁling students’ adaptation styles in Web-based learning. Computers & Education, 36(2), 121–132.

Lin, H.-F., & Lee, G.-G. (2006). Eﬀects of socio-technical factors on organizational intention to encourage knowledge sharing.

Management Decision, 44(1), 74–88.

May So, W. C., Danny Wong, T. N., & Sculli, D. (2005). Factors aﬀecting intentions to purchase via the internet. Industrial Management

& Data Systems, 105(9), 1225–1234.

Meuter, M. L., Jo Bitner, M., Ostrom, A. L., & Brown, S. W. (2005). Choosing among alternative service delivery modes: An investigation

of customer trial of self-service technologies. Journal of Marketing, 69, 61–83.

Moore, G. C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology

innovation. Information Systems, 2(3), 192–222.

Palvia, S. C. (2000). Eﬀectiveness of asynchronous and synchronous modes for learning computer software for endusers: An experimental

investigation. Journal of Computer Information Systems, 41(2), 99–109.

Rogers, E. M. (1983). Diﬀusion of innovations. New York: The Free Press.

Tornatzky, L., & Fleischer, M. (1990). The processes of technological innovation. New York: Lexington Books.

Tornatzky, L. J., & Klein, K. J. (1982). Innovation characteristics and innovation adoption-implementation: A meta-analysis of ﬁndings.

IEEE Transactions on Engineering Management, 29(1), 28–45.

UDN (2006). E-learning market grow fast in global Chinese market. Available online at: http://udn.com/NEWS/NATIONAL/NAT4/

3120102.shtml (accessed 21 March 2006).

Van Slyke, C., Belanger, F., & Comunale, C. (2004). Factors inﬂuencing the adoption of web-based shopping: the impact of trust.

Database for Advances in Information Systems, 35(2), 32–46.

Van Slyke, C., Lou, H., & Day, J. (2002). The impact of perceived innovation characteristics on intention to use groupware. Information

Resources Management Journal, 15(1), 5–12.

1416 H.-L. Liao, H.-P. Lu / Computers & Education 51 (2008) 1405–1416