- Email: [email protected]
Understanding Web-based learning continuance intention: The role of subjective task value
Information & Management 45 (2008) 194–201
Contents lists available at ScienceDirect
Information & Management journal homepage: www.elsevier.com/locate/im
Understanding Web-based learning continuance intention: The role of subjective task value§ Chao-Min Chiu *, Eric T.G. Wang Department of Information Management, National Central University, No. 300, Jungda Road, Jhongli City, Taoyuan 320, Taiwan, ROC
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 February 2005 Received in revised form 16 November 2007 Accepted 14 February 2008 Available online 3 April 2008
The success of Web-based learning depends on learner loyalty, i.e., subsequent continued usage (continuance). We extended the Unified Theory of Acceptance and Use of Technology (UTAUT) by introducing components of subjective task value into a model for studying learners’ continuance intentions in Web-based learning. Based on survey data from 286 respondents, SEM was employed to assess the model. The results indicated that performance expectancy, effort expectancy, computer selfefficacy, attainment value, utility value, and intrinsic value were significant predictors of individuals’ intentions to continue using Web-based learning, while anxiety had a significant negative effect. The results suggested the beneficial effect of positive subjective task value on stimulating learners’ intentions to continue using Web-based learning, which is as important as performance expectancy and effort expectancy. Implications and limitations of our study are discussed. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Benefits Continuance Costs Subjective task value Unified Theory of Acceptance and Use of Technology Web-based learning
1. Introduction The proliferation of network access and advances in Internet/ Web technology, in conjunction with the social demand for improved access to higher education, have stimulated the rapid growth of e-learning. It also helps organizations by reducing the cost of and increasing availability of training. Cortona consulting estimated that the e-learning market will reach $50 billion in 2010. Seventy percent of the universities in the US are now providing elearning courses. Web-based learning is based on material delivered through a Web browser over the public Internet, private intranet, or extranet. Its success depends mainly on learners’ loyalty, i.e., continued use. The importance of continuance is obvious: customer turnover can be costly—the cost of acquiring new customers is higher than that of retaining existing ones. We desired to explore individuals’ intentions to continue using Web-based learning in a voluntary setting. Two models were used to assess the technological and value issues and thus obtain an
§ This study is funded by the National Science Council under project number NSC93-2416-H-008-032. * Corresponding author. Tel.: +886 3 426 7251; fax: +886 3 425 4604. E-mail address: [email protected] (C.-M. Chiu).
0378-7206/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.im.2008.02.003
understanding of individuals’ actions: Unified Theory of Acceptance and Use of Technology and expectancy-value model of achievement motivation. As with any IS usage, the trade-off between benefits and costs has an important impact on continuance intention [15]. Researchers have conceptualized value as a function of a ‘‘get’’ component, i.e., benefits an individual receives, and a ‘‘give’’ component, i.e., an individual’s financial and nonmonetary costs in acquiring and using a product or service [31,35]. Value theorists argue that value is a centrally held and enduring belief that plays a central role in everyday decisions [23]. Eccles et al.’s [17] expectancy-value model of achievement motivation links individuals’ choice, persistence, and performance to expectancy for success and subjective task value. This model outlines four motivational components of subjective task value: attainment value (importance), intrinsic value (interest), utility value (usefulness), and cost. They showed that subjective task value predicted course plans and enrollment decisions in mathematics, physics, and English courses [16,18,27]. Consequently, we also argued that subjective task value influenced Web-based learning continuance intention through these variables. Web-based learning acceptance and usage can be partially explained by the Unified Theory of Acceptance and Use of Technology (UTAUT) [40], which is a parsimonious and robust model of individual acceptance of new IT. While it initially focused
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
on user acceptance and usage of IT in the workplace, it has recently been used in understanding the acceptance and use of mobile services and Web-based applications [8,11,43]. Therefore, we considered the major UTAUT constructs in determining usage intention and behavior: performance expectancy, effort expectancy, social influence, and facilitating conditions. 2. Theoretical background 2.1. Web-based learning There are many approaches to Web-based learning. On one end of the spectrum, individuals access information resources on the Web to learn and solve daily tasks by themselves however, in our study, we focused on online classes offered at institutions of higher education. It usually provides certificate, diploma, or degree programs, normally adopting a course management system that integrates a set of tools to support online content delivery, learning communities, and assessment. Some learning environments also facilitate cooperative learning, etc. Time flexibility has been reported as a major reason for students to choose to study online. Other reasons include less conflict with work, the convenience of not having to travel to attend a traditional face-to-face class, and the flexibility of accessing course materials anytime and anywhere. While Web-based learning has advantages, several weaknesses have also been described, including delay in response [32], feelings of isolation, and danger of arbitrariness [21]. There have been many studies of factors that influence learner satisfaction or outcomes. Carswell and Venkatesh [9] examined the influence of technological characteristics on learners’ outcomes (intention, acceptance, and performance). Other studies have examined the effect of learning patterns [25] on performance and also focused on the impact of personal cognition such as selfefficacy [42], computer confidence [30], and locus of control [28]. 2.2. Unified Theory of Acceptance and Use TAM posited that perceived usefulness and perceived ease of use were salient beliefs in determining user acceptance and use of IT. Venkatesh and Davis [39] extended the model to explain perceived usefulness and usage intentions in terms of social influence and cognitive instrumental processes. The Unified Theory of Acceptance and Use of Technology (UTAUT) was based on studies of eight competing models in IS adoption research. It posited that performance expectancy, effort expectancy and social influence were direct determinants of intention to use, and that facilitating conditions and intention were direct determinants of usage behavior. It seemed reasonable to assume that UTAUT could be used to study the acceptance and use of Web-based learning; however it ignored the impact of value. Bandura [1] suggested that an individual with high self-efficacy and outcome expectations may have low intention to continue using Web-based learning if he or she thinks that the process has low benefit and high cost. Consequently, we introduced subjective task value to UTAUT in addressing our research question.
well on a task), intrinsic value (enjoyment from performing the activity), and utility value (how well the task relates to current and future goals). Cost is conceptualized in terms of the negative aspects of engaging in the task (e.g., anxiety and fear of failure), as well as the amount of effort needed to succeed and any lost opportunities in other work. Eccles and her colleagues showed that components of subjective task value predicted both intentions and actual decisions to keep taking courses in the context of traditional classroom education. Also, components of subjective task value predicted high school students’ intention to keep taking mathematics [46] or to attend graduate school [2]. 3. Research model and hypotheses Web-based learning is an emerging application of the WWW and is different from IS used in the workplace. Existing variables of UTAUT do not reflect learners’ motives. Homer and Kahle argued that values are a social factor that guides an individual’s behavior. Values play such a central role, providing a basis for understanding human behavior in and across culture. Therefore we extended UTAUT by adding subjective task value. Although there are several classifications of values, our model adopted the specification of Eccles et al., since Web-based learning is an achievement-related activity. This expectancy-value model of achievement motivation focuses on subjective task value as a key factor influencing intention and choice. In our study, the dependent variable was Web-based learning continuance intention, which referred to the subjective probability that an individual would continue using Web-based learning. Venkatesh et al. indicated that seven constructs were significant direct determinants of intention or use. They theorized that attitudes toward using technology, self-efficacy, and anxiety were not direct determinants of intention. The intrinsic value component of our subjective task value was measured in a manner analogous to Venkatesh et al.’s measure of attitude but we did not include attitude per se. Self-efficacy was included in our model, while the cost component of subjective task value was conceptualized in terms of the negative aspects of engaging in the task. In our study, we identified four negative aspects of Web-based learning: social isolation, anxiety, delay in responses, and risk of arbitrary learning. Fig. 1 shows our model; in addition to the four core constructs of UTAUT, computer self-efficacy and the components of subjective task value were assumed to affect individuals’ intentions to continue using Web-based learning. 3.1. Performance expectancy Performance expectancy is the extent to which a person believes that a system enhances his or her performance. It pertains to perceived usefulness in TAM. Chau [10] defined two types of
2.3. Subjective task value The expectancy-value model of achievement motivation posits that individuals’ performance, persistence, and choice are directly predicted by their expectancies of success on the tasks and the subjective task value that represents success. Expectancy for success is thus analogous to self-efficacy. Subjective task value also can involve attainment value (the personal importance of doing
195
Fig. 1. Research model for Web-based learning continuance.
196
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
perceived usefulness: long- and near-term. Similarly, Thompson et al. [37] considered it to include near-term job fit and long-term consequences. Performance expectancy is analogous to near-term usefulness. Venkatesh et al. found that performance expectancy was a strong predictor of an individual’s intention to use a new technology in the workplace. Ong et al. [29], Mahmod et al. [26], and Sadde and Bahli [34] provided empirical support for the relationship between perceived usefulness and behavioral intention in the context of e-learning and Web-based learning. Accordingly, the following hypothesis was proposed. H1. Performance expectancy is positively related to Web-based learning continuance intention. 3.2. Effort expectancy Effort expectancy is to the extent to which a learner believes that using a system is free of effort. Effort expectancy pertains to perceived ease of use in TAM, which assumed that a system perceived to be easier to use was more likely to induce perception of usefulness and behavioral intention. To the extent that increased effort expectancy leads to improved performance, effort expectancy should have a direct effect on performance expectancy and continuance intention. Sadde and Bahli and Ong et al. indicated that perceived ease of use was positively associated with perceived usefulness and behavioral intention in the context of Web-based learning. Therefore, we proposed.
3.5. Facilitating conditions Factors and resources that an individual believes exist to support his or her activities are termed facilitating conditions. In our study, they included both technical and non-technical support. Triandis [38] stated that behavior could not occur if objective, facilitating conditions prevented it. Accordingly, the following hypothesis was proposed. H7. Facilitating conditions is positively related to Web-based learning continuance intention. 3.6. Attainment value Attainment value is the importance of doing well in terms of self-image and core personal values (e.g., achievement and competence needs). In general, students will be more likely to engage in a task, expend more effort, and do better when they value it [45]. Some studies on traditional classroom education have provided support for this notion. For example, Meece et al. found that the attainment value of junior high school students predicted their intentions to continue taking mathematics. Therefore, we made the hypothesis: H8. Attainment value is positively related to Web-based learning continuance intention. 3.7. Utility value
H2. Effort expectancy is positively related to Web-based learning continuance intention. H3. Effort expectancy is positively related to performance expectancy. 3.3. Computer self-efficacy Computer self-efficacy is the self-assessment of a person’s ability to use a system to complete important tasks. People who have low computer self-efficacy are less likely to continue using Web-based learning. Gong et al. [20] and Ong et al. showed that computer self-efficacy had a positive effect on perceived ease of use in the Web-based and e-learning contexts. Gong et al. also found that computer self-efficacy had a direct positive effect on intention to use Web-based learning. Therefore, we proposed. H4. Computer self-efficacy is positively related to effort expectancy. H5. Computer self-efficacy is positively related to Web-based learning continuance intention. 3.4. Social influence Social influence is to the degree to which an individual perceives that important others believe he or she should use a technology. The concept is similar to subjective norm in theory of planned behavior (TPB) which argued that the more favorable the social influence of a behavior, the stronger would be an individual’s intention to perform it. According to innovation diffusion theory (IDT) [33], users tend to increase communication with others to interpret their IT adoption. Such increased interactions can influence adoption decision. Studies have showed that subjective norm is a significant predictor of intention to use a system. Accordingly, the following hypothesis was proposed. H6. Social influence is positively related to Web-based learning continuance intention.
Utility value measures how well using a system relates to current and future career goals, such as promotion, and salary. It is analogous to long-term usefulness or consequences and it can be tied to the construct of extrinsic motivation in self-determination theory [14]. It has long been identified as a major influence on students’ achievement choices and behavior intention. Several researchers found that students’ perceptions of the utility of mathematics were strongly related to their intentions to continue or discontinue their mathematical study. Therefore, the following hypothesis was formulated. H9. Utility value is positively related to Web-based learning continuance intention. 3.8. Intrinsic value Intrinsic value is the extent to which an activity is perceived to be personally enjoyable. According to self-determination theory, learners are self-determining and intrinsically motivated in Webbased learning when they are interested in or enjoying doing it. Triandis argued that affect (e.g., feelings of joy, elation, and pleasure) had an impact on an individual’s behavior. Bong [5] found that students who were intrinsically interested in topics covered in their course were more likely to take similar courses in the future. Therefore, the following hypothesis was postulated. H10. Intrinsic value is positively related to Web-based learning continuance intention. 3.9. Social isolation Social isolation occurs when there are fewer opportunities for a learner to interact with other learners and instructors. If they are geographically and temporally separated, interaction is generally asynchronous and text-based; this can lead to restricted socialization or feeling of isolation. Bennett et al. [3] indicated that online learning students tended to suffer from social isolation—resulting
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
in students withdrawing from the course. Daugherty and Funke [13] found that a feeling of isolation was an important criterion for student dissatisfaction with a Web-based course. Kahl and Cropley [24] suggested that distance learning students felt more isolated than face-to-face students and experienced lower levels of selfconfidence. Therefore, we made the following hypothesis. H11. Social isolation is negatively related to Web-based learning continuance intention. 3.10. Anxiety Anxiety is apprehension or discomfort experienced by an individual with technology. It is characterized as an affective response, an emotional fear of potential negative outcomes, such as low performance or exposure to an unknown audience. It has been argued that such negative feelings detract from task performance and have a significant impact on intention to adopt e-learning through perceived behavioral control. Therefore, the following hypothesis was proposed. H12. Anxiety is negatively related to Web-based learning continuance intention. 3.11. Delay in responses Delay in responses is a lack of immediacy in receiving responses from a system. Vonderwell [41] indicated that one disadvantage of online courses was their lack of immediate feedback from the instructor. Participants in Hara and Kling’s [22] study reported frustration due to lack of immediacy. Therefore, the following hypothesis was proposed. H13. Delay in responses is negatively related to Web-based learning continuance intention. 3.12. Risk of arbitrary learning Risk of arbitrary learning occurs when students have difficulty in self-motivation. A key feature of Web-based learning is control of the learning pace. ‘‘Self-paced instruction places a substantial burden on the student to maintain interest, focus, and pace’’ [7]. In a survey of learners’ perception of online learning, Song et al. [36] reported that most learners agreed that motivation and time management impacted the success of their online learning experience. Accordingly, we assumed that learners with low motivation or bad study habits tended to fall behind, leading to low learning performance, etc. H14. Risk of arbitrary learning is negatively related to Web-based learning continuance intention. 4. Research methodology 4.1. Sample and procedures Data were gathered from part-time students who took Webbased courses offered by a university in Taiwan. The Web-based learning service is a credit and voluntary program designed primarily for continuing education. The tuition fee was US$ 90 per credit. Face-to-face classes were only open to full-time students; thus part-time students can only take Web-based courses. Three thousand e-mails providing a hyperlink to our Web survey were sent to part-time students who had registered for at least one Web-based course. A total of 286 surveys were returned. Among the respondents, 51% were female and 49% male; 77% had
197
taken one Web-based course, 15% had taken two or three, and 8% had taken at least four. On average, the respondents were 34 years old and spent 20 h per week using the Internet. Measurement items were adapted from the literature wherever possible. New items were developed from information in the literature. A pretest of the questionnaire was conducted using four experts to assess logical consistency, ease of understanding, sequence of items, and task relevance. Overall, they stated that the questionnaire was clear and easy to complete. A number of suggestions were made about the wording of several items and the overall structure of the questionnaire; the suggestions were discussed and changes were made to the instrument. A pilot study with 20 part-time Master’s degree students who had taken Webbased course was also conducted. The purpose of this pilot study was to gain additional comments on the questionnaire content and structure. For the main study, the participants were directed to the homepage of our Web survey; this indicated the purpose and importance of the study. All respondents were guaranteed confidentiality of their separate responses. In order to control for the fact that respondents had different frames of reference, we asked them to answer the questionnaire items according to their experience with the specified Web-based learning system. Followup e-mails were sent to individuals not having responded within 3 weeks. To maximize the response rate, 30 randomly selected respondents were given a US$ 10 incentive. 4.2. Instrument Table 1 shows the instrument. For all the measures, a sevenpoint Likert type scale was adopted with anchors ranging from strongly disagree (1) to strongly agree (7). 4.3. Measurement model Confirmatory factor analysis (CFA) was applied to test the adequacy of the measurement model using LISREL 8.5. The adequacy of the measurement models was evaluated on the criteria of model fit, reliability, convergent validity, and discriminant validity. For a good model fit, the Chi-square value normalized by degrees of freedom (x2/d.f.) should not exceed 3, adjusted goodness of fit index (AGFI) should exceed 0.8, non-normed fit index (NNFI) and comparative fit index (CFI) should exceed 0.9, and root mean square error of approximation (RMSEA) should not exceed 0.08. For our CFA model, x2/d.f. was 1.43 (x2 = 1117; d.f. = 782), AGFI was 0.81, NNFI was 0.96, CFI was 0.96, and RMSEA was 0.039, suggesting adequate model fit. Reliability was examined using the composite reliability values. As shown in Table 1, all of them were above 0.7, indicating a commonly acceptable level for confirmatory research. Convergent validity was evaluated for measurement scales using two criteria suggested by Fornell and Larcker [19]: all indicator factor loadings should be significant and exceed 0.70 and average variance extracted (AVE) for each construct should exceed the variance due to measurement error for that construct (i.e., should exceed 0.50). Most items exhibited loading higher than 0.7 on their respective construct, providing evidence of acceptable item convergence on the intended constructs. Two exceptions were the third item of computer self-efficacy and the third item of the social influence with loadings slightly below 0.7. AVE ranged from 0.59 to 0.89 (see Table 2), greater than the variances due to measurement errors. Thus conditions for convergent validity were essentially met. Fornell and Larcker suggested that for satisfactory discriminant validity, the square root of the AVE from a construct should be greater than the correlation shared between the construct and
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
198 Table 1 Summary of measurement scales Construct
Measure
Performance expectancy (PE) (composite reliability = 0.85) PE1 Using the Web-based learning improves my performance in my learning activities PE2 Using the Web-based learning enhances my effectiveness on learning activities PE3 Using the Web-based learning in my learning activities increases my productivity PE4 I find the Web-based learning useful for my learning activities Effort expectancy (EE) (composite reliability = 0.89) EE1 It is easy for me to become skillful at using the Web-based learning system EE2 My interaction with the Web-based learning system is clear and understandable EE3 I find it easy to get the Web-based learning system to do what I want it to do EE4 I find the Web-based learning system to be easy to use
Mean (S.D.)
Loading
Source
5.30 (1.16)
0.75
Venkatesh et al.
5.33 (1.19)
0.73
5.31 (1.21)
0.78
5.76 (0.98)
0.83
5.99 (0.94)
0.89
5.45 (1.17)
0.76
5.83 (1.02)
0.83
5.97 (0.87)
0.86
Computer self-efficacy (CSE) (composite reliability = 0.84) (I could complete my learning activities using the Web-based system . . .) CSE1 If I had never used a system like it before 5.15 (1.31) 0.70 CSE2 If I had only the system manuals for reference 5.67 (1.00) 0.83 CSE3 If I had seen someone else using it before 5.86 (0.86) 0.69 trying it myself CSE4 If I had just the built-in help facility for 5.52 (1.06) 0.78 assistance Social influence (SI) (composite reliability = 0.89) SI1 People who influence my behavior think that I should participate in Web-based leaning activities SI2 People who are important to me think that I should participate in Web-based leaning activities SI3 The senior management of the organization has supported my participation in Web-based leaning activities Facilitating conditions (FC) (composite reliability = 0.82) FC1 I have the resources necessary to use the Web-based leaning system FC2 I have the knowledge necessary to use the Web-based leaning system FC3 The Web-based leaning system is compatible with other systems I use Attainment value (AV) (composite reliability = 0.91) AV1 I think Web-based learning makes me a more knowledgeable person AV2 I think Web-based learning offers a forum for fulfilling achievement AV3 I think being successful at Web-based learning confirms my competence AV4 I think being successful at Web-based learning give me a sense of confidence Utility value (UV) (composite reliability = 0.89) UV1 I think what I learn through Web-based learning is useful for my promotion UV2 I think what I learn through Web-based learning is useful for getting salary raise UV3 I think what I learn through Web-based learning is helpful for me to get a job Intrinsic value (IV) (composite reliability = 0.93) IV1 I think Web-based learning is interesting IV2 I think Web-based learning is enjoyable IV3 I think Web-based learning is fun Social isolation (SIL) (composite reliability = 0.93) SIL1 I think Web-based learning restricts socialization, due to the reduction of opportunities of face-to-face interactions among students SIL2 I think Web-based learning reduces opportunities of face-to-face interactions between students and the instructor Anxiety (AN) (composite reliability = 0.96)
Venkatesh et al.
Compeau and Higgins [12]
5.13 (1.29)
0.94
Venkatesh et al.
5.08 (1.30)
0.91
5.34 (1.19)
0.68
6.28 (0.75)
0.73
6.00 (0.87)
0.83
5.88 (0.79)
0.78
6.03 (0.79)
0.74
5.55 (1.17)
0.89
5.76 (1.06)
0.84
5.58 (1.13)
0.92
4.44 (1.52)
0.88
4.01 (1.57)
0.92
4.56 (1.47)
0.74
5.16 (1.19) 5.37 (1.14) 5.39 (1.15)
0.86 0.93 0.93
Battle and Wigfield
5.52 (1.11)
0.88
Eccles et al and Song et al.
5.53 (1.14)
0.99
Venkatesh et al.
Eccles et al. and Battle and Wigfield
Eccles et al. and Battle and Wigfield
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
199
Table 1 (Continued ) Construct AN1
AN2 AN3
Measure
Mean (S.D.)
Loading
Source
I feel apprehensive about using Web-based leaning to accomplish my learning tasks I feel uneasy about using Web-based leaning to accomplish my learning tasks I feel uncomfortable about using Web-based leaning to accomplish my learning tasks
2.98 (1.37)
0.94
Venkatesh et al.
2.87 (1.38)
0.99
2.73 (1.31)
0.90
4.37 (1.35)
0.95
4.39 (1.34)
0.95
4.30 (1.38)
0.93
3.29 (1.54)
0.88
3.27 (1.55)
0.92
3.28 (1.51)
0.95
3.20 (1.50)
0.85
5.88 (0.94)
0.92
5.87 (0.90)
0.97
5.71 (1.03)
0.86
Delay in response (DR) (composite reliability = 0.96) DR1 I felt a lack of immediacy in getting responses back from the instructor DR2 I felt a lack of immediacy in getting responses back from other students DR3 I felt a lack of immediacy in getting responses back from the teaching assistant Risk of arbitrary learning (RAL) (composite reliability = 0.94) RAL1 I did not consistently participate in the Web-based leaning activities RAL2 I was not able to maintain my focus in the Web-based leaning activities RAL3 I was not able to maintain my interest in the Web-based leaning activities RAL4 I was not able to accomplish my learning goals in the Web-based leaning activities Continuance intention (CI) (composite reliability = 0.94) CI1 If I could, I would like to continue using Web-based learning in my learning activities in the future CI2 It is likely that I will continue using Web-based learning in the future CI3 I expect to continue using Web-based learning in the future
Song et al.
Learning and Study Strategies Inventory [44] and Self-regulation Questionnaire [6]
Bhattacherjee [4]
less than 0.05. The explanatory power of the research model is also shown. The R2 value shows that performance expectancy, effort expectancy, computer self-efficacy, social influence, facilitating conditions, attainment value, utility value, intrinsic value, social isolation, anxiety, delay in response, and risk of arbitrary learning together accounted for 60% of the variance of Web-based learning continuance intention.
other constructs in the model. The diagonal values exceeded the inter-construct correlations; hence the test of discriminant validity was acceptable. Therefore we concluded that the construct validity of the measurement scales was sufficiently high. 4.4. Model testing results The overall fit and the strengths of the hypothesized paths were examined. As shown in Table 3, the fit indices were within accepted thresholds: A x2 to degrees of freedom ratio of 1.60 (x2 = 1284; d.f. = 801), AGFI = 0.80, NNFI = 0.94, CFI = 0.95, and RMSEA = 0.046. Hence, this model fitted the data reasonably well. The significance of individual paths is shown in Fig. 2 and summarized in Table 4. Nine out of 14 paths exhibited a P-value
5. Discussion and implications 5.1. Summary of results An interesting finding of our study was that performance expectancy (near-term usefulness) and utility value (long-term
Table 2 Squared correlations of latent variables Construct
AVE
PE EE CSE SI FC AV UV IV SIL AN DR RAL CI
0.59 0.66 0.57 0.73 0.60 0.72 0.73 0.82 0.88 0.89 0.89 0.81 0.84
PE 0.77 0.48 0.36 0.40 0.35 0.61 0.23 0.52 0.18 0.26 0.41 0.16 0.54
EE
0.81 0.50 0.30 0.53 0.48 0.22 0.52 0.13 0.38 0.44 0.19 0.57
CSE
0.75 0.29 0.49 0.40 0.17 0.37 0.07 0.27 0.27 0.05 0.47
SI
0.85 0.22 0.50 0.37 0.49 0.01 0.10 0.19 0.08 0.40
FC
0.78 0.42 0.16 0.42 0.01 0.43 0.29 0.06 0.44
AV
0.85 0.37 0.71 0.15 0.28 0.40 0.14 0.65
UV
0.85 0.47 0.01 0.07 0.14 0.03 0.38
IV
0.91 0.09 0.19 0.38 0.15 0.66
SIL
0.94 0.15 0.22 0.30 0.08
AN
0.95 0.52 0.24 0.35
DR
0.95 0.34 0.40
RAL
0.90 0.11
CI
0.92
Diagonal elements (in bold) are the square root of the average variance extracted (AVE). Off-diagonal elements are the correlations among constructs. For discriminant validity, diagonal elements should be larger than off-diagonal elements. PE = performance expectancy; EE = effort expectancy; CSE = computer self-efficacy; SI = social influence; FC = facilitating conditions; AV = attainment value; UV = utility value; IV = intrinsic value; SIL = social isolation; AN = anxiety; DR = delay in response; RAL = risk of arbitrary learning; CI = continuance intention.
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
200 Table 3 Overall model fit indices for the research model Model fit indices
Results
Chi-square statistic x2/d.f. AGFI CFI NNFI RMSEA
1.60 (1283/801) 0.80 0.95 0.94 0.046
Table 4 Results of hypothesis testing Recommended value 3 0.80 0.90 0.90 <0.08
usefulness) had almost the same effects on continuance intention for working professionals with limited time for continuing education. Contrary to our hypotheses, social influence and facilitating conditions were not significant predictors of continuance intention. Additionally, social isolation, delay in response, and risk of arbitrary learning were not significant negative predictors of continuance. Apparently learners feel that social isolation and delay in response are negative aspects of Web-based learning, but they still have strong intentions to continue using it. The R2 value showed that performance expectancy, effort expectancy, computer self-efficacy, social influence and facilitating conditions together accounted for 46.6% of the variance of continuance intention. It increased when the three dimensions of positive subjective task value and four dimensions of cost were added as independent variables. 5.2. Implications for theory and practice Our research contributed to an overall conceptual understanding of the nature and the importance of components of subjective task value as continuance intention factors in continuing education. Apparently the perceived technological characteristics of Web-based learning themselves are insufficient to increase continuance intention. According to social exchange theory, individuals behave by rational self-interest. Therefore, continuance intention will be stimulated when rewards of Webbased learning outweigh its cost. From a practical perspective, our study implied that an understanding of what performance expectancy, effort expectancy, and positive subjective task value mean to learners is likely to establish longer-term relationships with learners. In addition, our findings also suggested that intrinsic value was the strongest predictor of an individual’s intention to continue using Web-based learning. Developers and designers of Web-based learning sites should presumably employ ways to reduce monotony and exploit learners’ playful characteristics. Social isolation, risk of arbitrary learning, and delay in response did not have significant negative effects on continuance intention. Thus, our findings implied that positive values or benefits out-
Hypotheses
Results
H1: Performance expectancy is positively related to Web-based learning continuance intention H2: Effort expectancy is positively related to Web-based learning continuance intention H3: Effort expectancy is positively related to performance expectancy H4: Computer self-efficacy is positively related to effort expectancy H5: Computer self-efficacy is positively related to Web-based learning continuance intention H6: Social influence is positively related to Web-based learning continuance intention H7: Facilitating conditions is positively related to Web-based learning continuance intention H8: Attainment value is positively related to Web-based learning continuance intention H9: Utility value is positively related to Web-based learning continuance intention H10: Intrinsic value is positively related to Web-based learning continuance intention H11: Social isolation is negatively related to Web-based learning continuance H12: Anxiety is negatively related to Web-based learning continuance intention H13: Delay in responses is negatively related to Web-based learning continuance intention H14: Risk of arbitrary learning is negatively related to Web-based learning
Supported Supported Supported Supported Supported Not supported Not supported Supported Supported Supported Not supported Supported Not supported Not supported
weighed the costs of Web-based learning. However, Web-based learning systems should provide well-designed tools or mechanisms for online interaction and instructors should always be available to provide quick responses, thus reducing the negative impact of social isolation and delay in response. 5.3. Limitations Although our findings were encouraging and useful, the study had several limitations. First, our results are not generalizable; we examined only one Web-based learning system and the subjects were adult learners (part-time students in continuing education). Factors influencing continuance intention of adult learners might be different from those of full-time students. Second, the results could be affected by self-selection bias. Our sample consisted of active participants of Web-based learning courses. Individuals who had already ceased to participate might have different perceptions about the influence of technological attributes of Web-based learning and subjective task value. Therefore, the results should be interpreted as only explaining continuance intention of current users of Webbased learning. Third, the data are cross-sectional. Individuals’ intention to use Web-based learning is an ongoing process. Our constructs were measured at a single point. Fourth, the usage of the Web-based learning service is voluntary (under users’ full volitional control). The findings might not be generalizable to a mandatory setting. Finally, our study focused on empirically validating the hypotheses derived from the research model.
References
Fig. 2. SEM analysis of the research model.
[1] A. Bandura, Social Foundations of Thought and Action: A Social Cognitive Theory, Prentice Hall, Englewood Cliffs, NJ, 1986. [2] A. Battle, A. Wigfield, College women’s orientations toward family, career, and graduate school, Journal of Vocational Behavior 62 (1) (2003) 56–75. [3] S. Bennett, A.M. Priest, C. Macpherson, Learning about online learning: an approach to staff development for university teachers, Australian Journal of Educational Technology 15 (3) (1999) 207–221. [4] A. Bhattacherjee, Understanding information systems continuance: an expectation-confirmation model, MIS Quarterly 25 (3) (2001) 351–370.
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201 [5] M. Bong, Role of self-efficacy and task-value in predicting college students’ course performance and future enrollment intentions, Contemporary Educational Psychology 26 (2001) 553–570. [6] J.M. Brown, W.R. Miller, L.A. Lawendowski, The self-regulation questionnaire, in: L. VandeCreek, T.L. Jackson (Eds.), Innovations in Clinical Practice: A Source Book, Professional Resource Press, Sarasota, FL, 1999, pp. 281–289. [7] California Distance Learning Project, What is Distance Learning? Available at http://www.cdlponline.org/index.cfm?fuseaction=whatis&pg=2, 2005. [8] C. Carlsson, J. Carlsson, K. Hyvo¨nen, J. Puhakainen, P. Walden, Adoption of mobile devices/services—searching for answers with the UTAUT, in: Proceedings of the 39th Hawaii International Conference on System Sciences, 2006. [9] A.D. Carswell, V. Venkatesh, Learner outcomes in an asynchronous distance education environment, International Journal of Human–Computer Studies 56 (2002) 475–494. [10] P.Y.K. Chau, An empirical assessment of a modified technology acceptance model, Journal of Management Information Systems 13 (2) (1996) 185–204. [11] E. Cody-Allen, R. Kishore, An extension of the UTAUT model with e-quality, trust, and satisfaction constructs, in: Proceedings of the 2006 ACM SIGMIS CPR Conference on Computer Personnel Research, 2006, pp. 82–89. [12] D.R. Compeau, C.A. Higgins, Computer self-efficacy: development of a measure and initial test, MIS Quarterly 19 (2) (1995) 189–211. [13] M. Daugherty, B.L. Funke, University faculty and student perceptions of Webbased instruction, Journal of Distance Education 13 (1) (1998) 21–39. [14] E.L. Deci, R.M. Ryan, Intrinsic Motivation and Self-determination in Human Behavior, Plenum Press, New York, NY, 1985. [15] W.H. DeLone, E.R. McLean, The DeLone and McLean model of information systems success: a ten-year update, Journal of Management Information Systems 19 (4) (2003) 9–30. [16] J.S. Eccles, Gender roles and women’s achievement-related decisions, Psychology of Women Quarterly 11 (2) (1987) 135–172. [17] J.S. Eccles, T.F. Adler, R. Futterman, S.B. Goff, C.M. Kaczala, J.L. Meece, C. Midgley, Expectancies, values, and academic behaviors, in: J.T. Spence (Ed.), Achievement and Achievement Motivation, W.H. Freeman, San Francisco, CA, 1983, pp. 75–146. [18] J.S. Eccles, T. Adler, J.L. Meece, Sex differences in achievement: a test of alternate theories, Journal of Personality and Social Psychology 46 (1) (1984) 26–43. [19] C. Fornell, D.F. Larcker, Evaluating structural equation models with unobservable and measurement error, Journal of Marketing Research 18 (1) (1981) 39–50. [20] M. Gong, Y. Xu, Y. Yu, An enhanced technology acceptance model for Web-based learning, Journal of Information Systems Education 15 (4) (2004) 365–373. [21] I. Hamburg, C. Lindecke, Social aspects of e-learning and blending learning methods, in: Proceedings of the 4th European Conference E-Comm-Line, Bucharest, Romania, 2003. [22] N. Hara, R. Kling, Students’ frustrations with a web-based distance education course, First Monday 4(12), available at http://www.firstmonday.dk/issues/ issue4_12/index.html, 1999. [23] P.M. Homer, L.R. Kahle, A structural equation test of the value-attitude-behavior hierarchy, Journal of Personality and Social Psychology 54 (4) (1988) 638–646. [24] T.N. Kahl, A.J. Cropley, Face-to-face versus distance learning: psychological consequences and practical implications, Distance Education 7 (1) (1986) 38–48. [25] J. Lu, C.S. Yu, C. Liu, Learning style, learning patterns, and learning performance in a WebCT-based MIS course, Information & Management 40 (2003) 497–507. [26] R. Mahmod, N. Dahlan, T. Ramayah, N. Karia, M.H.A.H. Asaari, Attitudinal belief on adoption of e-MBA program in Malaysia, Turkish Online Journal of Distance Education 6 (2) (2005) 115–124. [27] J.L. Meece, A. Wigfield, J.S. Eccles, Predictors of math anxiety and its consequences for young adolescents’ course enrollment intentions and performances in mathematics, Journal of Educational Psychology 82 (1) (1990) 60–70. [28] L. Morris, S. Wu, C. Finnegan, Predicting retention in online general education courses, The American Journal of Distance Education 19 (2005) 23–36. [29] C.S. Ong, J.Y. Laia, Y.S. Wang, Factors affecting engineers’ acceptance of asynchronous e-learning systems in high-tech companies, Information & Management 41 (2004) 795–804. [30] V. Osborn, The Distributed Learning Survey, University of North Texas, Denton, TX, 2000. [31] A. Parasuraman, D. Grewal, The impact of technology on the quality-value-loyalty chain: a research agenda, Journal of the Academy of Marketing Science 28 (1) (2000) 168–174. [32] L.A. Petrides, Web-based technologies for distributed (or distance) learning: creating learning-centered educational experiences in the higher education classroom, International Journal of Instructional Media 29 (1) (2002) 69–77.
201
[33] E.M. Rogers, Diffusion of Innovations, The Free Press, New York, NY, 1995. [34] R.G. Saade´, B. Bahli, The impact of cognitive absorption on perceived usefulness and perceived ease of use in on-line learning: an extension of the technology acceptance model, Information & Management 42 (2005) 317–327. [35] D. Sirdeshmukh, J. Singh, B. Sabol, Consumer trust, value, and loyalty in relational exchange, Journal of Marketing 66 (2002) 15–37. [36] L. Song, E.S. Singleton, J.R. Hill, M. Koh, Improving online learning: student perceptions of useful and challenging characteristics, Internet and Higher Education 7 (1) (2004) 59–70. [37] R.L. Thompson, C.A. Higgins, J.M. Howell, Personal computing: toward a conceptual model of utilization, MIS Quarterly 15 (1) (1991) 125–143. [38] H.C. Triandis, Values, attitudes, and interpersonal behavior, in: H.E. Howe, M.M. Page (Eds.), Nebraska Symposium on Motivation, 1979, University of Nebraska Press, Lincoln, NE, 1980, pp. 195–260. [39] V. Venkatesh, F.D. Davis, A theoretical extension of the technology acceptance model: four longitudinal field studies, Management Science 46 (2) (2000) 186– 204. [40] V. Venkatesh, M. Morris, G. Davis, F. Davis, User acceptance of information technology: toward a unified view, MIS Quarterly 27 (3) (2003) 425–478. [41] S. Vonderwell, An examination of asynchronous communication experiences and perspectives of students in an online course: a case study, Internet and Higher Education 6 (1) (2003) 77–90. [42] A.Y. Wang, M.H. Newlin, Predictors of web-student performance: the role f selfefficacy and reasons for taking an on-line class, Computers in Human Behavior 18 (2002) 151–163. [43] H.I. Wang, H.L. Yang, The role of personality traits in UTAUT model under online stocking, Contemporary Management Research 1 (1) (2005) 69–82. [44] C.E. Weinstein, D.R. Palmer, A.C. Schulte, Learning and Study Strategies Inventory, H & H Publishing, Clearwater, FL, 1987. [45] A. Wigfield, The role of children’s achievement values in the self-regulation of their learning outcomes, in: D.H. Schunk, B.J. Zimmerman (Eds.), Self-regulation of Learning and Performance: Issues and Educational Applications, Erlbaum, Mahwah, NJ, 1994, pp. 101–124. [46] A. Wigfield, J. Eccles, Relations of expectancies and values to students’ math grades and intentions, in: Paper Presented at the Meeting of the American Educational Research Association, San Francisco, CA, 1989. Chao-Min Chiu is a professor in the Department of Information Management at the National Central University, Taiwan (ROC). He holds a Ph.D. in management from the Rutgers University. His research interests include electronic commerce, virtual communities, and knowledge management. His research has appeared in Information & Management, Decision Support Systems, Information Systems Journal, International Journal of Human–Computer Studies, Behavior & Information Technology, Computers & Education, Information and Software Technology, Information Systems Management, Information Technology and Management, Journal of Information Science, and others. Eric T.G. Wang is information management chair professor at National Central University, Taiwan (ROC). He received the Ph.D. degree in business administration, specialized in computer & information systems, from the William E. Simon Graduate School of Business Administration, University of Rochester. His research interests include electronic commerce, outsourcing, organizational economics, and organizational impact of information technology. His research has appeared in Information & Management, Management Science, Information Systems Research, Journal of Management Information Systems, Decision Sciences, Decision Support Systems, Information Systems Journal, Omega, European Journal of Information Systems, European Journal of Operational Research, International Journal of Information Management, and others.
Contents lists available at ScienceDirect
Information & Management journal homepage: www.elsevier.com/locate/im
Understanding Web-based learning continuance intention: The role of subjective task value§ Chao-Min Chiu *, Eric T.G. Wang Department of Information Management, National Central University, No. 300, Jungda Road, Jhongli City, Taoyuan 320, Taiwan, ROC
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 February 2005 Received in revised form 16 November 2007 Accepted 14 February 2008 Available online 3 April 2008
The success of Web-based learning depends on learner loyalty, i.e., subsequent continued usage (continuance). We extended the Unified Theory of Acceptance and Use of Technology (UTAUT) by introducing components of subjective task value into a model for studying learners’ continuance intentions in Web-based learning. Based on survey data from 286 respondents, SEM was employed to assess the model. The results indicated that performance expectancy, effort expectancy, computer selfefficacy, attainment value, utility value, and intrinsic value were significant predictors of individuals’ intentions to continue using Web-based learning, while anxiety had a significant negative effect. The results suggested the beneficial effect of positive subjective task value on stimulating learners’ intentions to continue using Web-based learning, which is as important as performance expectancy and effort expectancy. Implications and limitations of our study are discussed. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Benefits Continuance Costs Subjective task value Unified Theory of Acceptance and Use of Technology Web-based learning
1. Introduction The proliferation of network access and advances in Internet/ Web technology, in conjunction with the social demand for improved access to higher education, have stimulated the rapid growth of e-learning. It also helps organizations by reducing the cost of and increasing availability of training. Cortona consulting estimated that the e-learning market will reach $50 billion in 2010. Seventy percent of the universities in the US are now providing elearning courses. Web-based learning is based on material delivered through a Web browser over the public Internet, private intranet, or extranet. Its success depends mainly on learners’ loyalty, i.e., continued use. The importance of continuance is obvious: customer turnover can be costly—the cost of acquiring new customers is higher than that of retaining existing ones. We desired to explore individuals’ intentions to continue using Web-based learning in a voluntary setting. Two models were used to assess the technological and value issues and thus obtain an
§ This study is funded by the National Science Council under project number NSC93-2416-H-008-032. * Corresponding author. Tel.: +886 3 426 7251; fax: +886 3 425 4604. E-mail address: [email protected] (C.-M. Chiu).
0378-7206/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.im.2008.02.003
understanding of individuals’ actions: Unified Theory of Acceptance and Use of Technology and expectancy-value model of achievement motivation. As with any IS usage, the trade-off between benefits and costs has an important impact on continuance intention [15]. Researchers have conceptualized value as a function of a ‘‘get’’ component, i.e., benefits an individual receives, and a ‘‘give’’ component, i.e., an individual’s financial and nonmonetary costs in acquiring and using a product or service [31,35]. Value theorists argue that value is a centrally held and enduring belief that plays a central role in everyday decisions [23]. Eccles et al.’s [17] expectancy-value model of achievement motivation links individuals’ choice, persistence, and performance to expectancy for success and subjective task value. This model outlines four motivational components of subjective task value: attainment value (importance), intrinsic value (interest), utility value (usefulness), and cost. They showed that subjective task value predicted course plans and enrollment decisions in mathematics, physics, and English courses [16,18,27]. Consequently, we also argued that subjective task value influenced Web-based learning continuance intention through these variables. Web-based learning acceptance and usage can be partially explained by the Unified Theory of Acceptance and Use of Technology (UTAUT) [40], which is a parsimonious and robust model of individual acceptance of new IT. While it initially focused
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
on user acceptance and usage of IT in the workplace, it has recently been used in understanding the acceptance and use of mobile services and Web-based applications [8,11,43]. Therefore, we considered the major UTAUT constructs in determining usage intention and behavior: performance expectancy, effort expectancy, social influence, and facilitating conditions. 2. Theoretical background 2.1. Web-based learning There are many approaches to Web-based learning. On one end of the spectrum, individuals access information resources on the Web to learn and solve daily tasks by themselves however, in our study, we focused on online classes offered at institutions of higher education. It usually provides certificate, diploma, or degree programs, normally adopting a course management system that integrates a set of tools to support online content delivery, learning communities, and assessment. Some learning environments also facilitate cooperative learning, etc. Time flexibility has been reported as a major reason for students to choose to study online. Other reasons include less conflict with work, the convenience of not having to travel to attend a traditional face-to-face class, and the flexibility of accessing course materials anytime and anywhere. While Web-based learning has advantages, several weaknesses have also been described, including delay in response [32], feelings of isolation, and danger of arbitrariness [21]. There have been many studies of factors that influence learner satisfaction or outcomes. Carswell and Venkatesh [9] examined the influence of technological characteristics on learners’ outcomes (intention, acceptance, and performance). Other studies have examined the effect of learning patterns [25] on performance and also focused on the impact of personal cognition such as selfefficacy [42], computer confidence [30], and locus of control [28]. 2.2. Unified Theory of Acceptance and Use TAM posited that perceived usefulness and perceived ease of use were salient beliefs in determining user acceptance and use of IT. Venkatesh and Davis [39] extended the model to explain perceived usefulness and usage intentions in terms of social influence and cognitive instrumental processes. The Unified Theory of Acceptance and Use of Technology (UTAUT) was based on studies of eight competing models in IS adoption research. It posited that performance expectancy, effort expectancy and social influence were direct determinants of intention to use, and that facilitating conditions and intention were direct determinants of usage behavior. It seemed reasonable to assume that UTAUT could be used to study the acceptance and use of Web-based learning; however it ignored the impact of value. Bandura [1] suggested that an individual with high self-efficacy and outcome expectations may have low intention to continue using Web-based learning if he or she thinks that the process has low benefit and high cost. Consequently, we introduced subjective task value to UTAUT in addressing our research question.
well on a task), intrinsic value (enjoyment from performing the activity), and utility value (how well the task relates to current and future goals). Cost is conceptualized in terms of the negative aspects of engaging in the task (e.g., anxiety and fear of failure), as well as the amount of effort needed to succeed and any lost opportunities in other work. Eccles and her colleagues showed that components of subjective task value predicted both intentions and actual decisions to keep taking courses in the context of traditional classroom education. Also, components of subjective task value predicted high school students’ intention to keep taking mathematics [46] or to attend graduate school [2]. 3. Research model and hypotheses Web-based learning is an emerging application of the WWW and is different from IS used in the workplace. Existing variables of UTAUT do not reflect learners’ motives. Homer and Kahle argued that values are a social factor that guides an individual’s behavior. Values play such a central role, providing a basis for understanding human behavior in and across culture. Therefore we extended UTAUT by adding subjective task value. Although there are several classifications of values, our model adopted the specification of Eccles et al., since Web-based learning is an achievement-related activity. This expectancy-value model of achievement motivation focuses on subjective task value as a key factor influencing intention and choice. In our study, the dependent variable was Web-based learning continuance intention, which referred to the subjective probability that an individual would continue using Web-based learning. Venkatesh et al. indicated that seven constructs were significant direct determinants of intention or use. They theorized that attitudes toward using technology, self-efficacy, and anxiety were not direct determinants of intention. The intrinsic value component of our subjective task value was measured in a manner analogous to Venkatesh et al.’s measure of attitude but we did not include attitude per se. Self-efficacy was included in our model, while the cost component of subjective task value was conceptualized in terms of the negative aspects of engaging in the task. In our study, we identified four negative aspects of Web-based learning: social isolation, anxiety, delay in responses, and risk of arbitrary learning. Fig. 1 shows our model; in addition to the four core constructs of UTAUT, computer self-efficacy and the components of subjective task value were assumed to affect individuals’ intentions to continue using Web-based learning. 3.1. Performance expectancy Performance expectancy is the extent to which a person believes that a system enhances his or her performance. It pertains to perceived usefulness in TAM. Chau [10] defined two types of
2.3. Subjective task value The expectancy-value model of achievement motivation posits that individuals’ performance, persistence, and choice are directly predicted by their expectancies of success on the tasks and the subjective task value that represents success. Expectancy for success is thus analogous to self-efficacy. Subjective task value also can involve attainment value (the personal importance of doing
195
Fig. 1. Research model for Web-based learning continuance.
196
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
perceived usefulness: long- and near-term. Similarly, Thompson et al. [37] considered it to include near-term job fit and long-term consequences. Performance expectancy is analogous to near-term usefulness. Venkatesh et al. found that performance expectancy was a strong predictor of an individual’s intention to use a new technology in the workplace. Ong et al. [29], Mahmod et al. [26], and Sadde and Bahli [34] provided empirical support for the relationship between perceived usefulness and behavioral intention in the context of e-learning and Web-based learning. Accordingly, the following hypothesis was proposed. H1. Performance expectancy is positively related to Web-based learning continuance intention. 3.2. Effort expectancy Effort expectancy is to the extent to which a learner believes that using a system is free of effort. Effort expectancy pertains to perceived ease of use in TAM, which assumed that a system perceived to be easier to use was more likely to induce perception of usefulness and behavioral intention. To the extent that increased effort expectancy leads to improved performance, effort expectancy should have a direct effect on performance expectancy and continuance intention. Sadde and Bahli and Ong et al. indicated that perceived ease of use was positively associated with perceived usefulness and behavioral intention in the context of Web-based learning. Therefore, we proposed.
3.5. Facilitating conditions Factors and resources that an individual believes exist to support his or her activities are termed facilitating conditions. In our study, they included both technical and non-technical support. Triandis [38] stated that behavior could not occur if objective, facilitating conditions prevented it. Accordingly, the following hypothesis was proposed. H7. Facilitating conditions is positively related to Web-based learning continuance intention. 3.6. Attainment value Attainment value is the importance of doing well in terms of self-image and core personal values (e.g., achievement and competence needs). In general, students will be more likely to engage in a task, expend more effort, and do better when they value it [45]. Some studies on traditional classroom education have provided support for this notion. For example, Meece et al. found that the attainment value of junior high school students predicted their intentions to continue taking mathematics. Therefore, we made the hypothesis: H8. Attainment value is positively related to Web-based learning continuance intention. 3.7. Utility value
H2. Effort expectancy is positively related to Web-based learning continuance intention. H3. Effort expectancy is positively related to performance expectancy. 3.3. Computer self-efficacy Computer self-efficacy is the self-assessment of a person’s ability to use a system to complete important tasks. People who have low computer self-efficacy are less likely to continue using Web-based learning. Gong et al. [20] and Ong et al. showed that computer self-efficacy had a positive effect on perceived ease of use in the Web-based and e-learning contexts. Gong et al. also found that computer self-efficacy had a direct positive effect on intention to use Web-based learning. Therefore, we proposed. H4. Computer self-efficacy is positively related to effort expectancy. H5. Computer self-efficacy is positively related to Web-based learning continuance intention. 3.4. Social influence Social influence is to the degree to which an individual perceives that important others believe he or she should use a technology. The concept is similar to subjective norm in theory of planned behavior (TPB) which argued that the more favorable the social influence of a behavior, the stronger would be an individual’s intention to perform it. According to innovation diffusion theory (IDT) [33], users tend to increase communication with others to interpret their IT adoption. Such increased interactions can influence adoption decision. Studies have showed that subjective norm is a significant predictor of intention to use a system. Accordingly, the following hypothesis was proposed. H6. Social influence is positively related to Web-based learning continuance intention.
Utility value measures how well using a system relates to current and future career goals, such as promotion, and salary. It is analogous to long-term usefulness or consequences and it can be tied to the construct of extrinsic motivation in self-determination theory [14]. It has long been identified as a major influence on students’ achievement choices and behavior intention. Several researchers found that students’ perceptions of the utility of mathematics were strongly related to their intentions to continue or discontinue their mathematical study. Therefore, the following hypothesis was formulated. H9. Utility value is positively related to Web-based learning continuance intention. 3.8. Intrinsic value Intrinsic value is the extent to which an activity is perceived to be personally enjoyable. According to self-determination theory, learners are self-determining and intrinsically motivated in Webbased learning when they are interested in or enjoying doing it. Triandis argued that affect (e.g., feelings of joy, elation, and pleasure) had an impact on an individual’s behavior. Bong [5] found that students who were intrinsically interested in topics covered in their course were more likely to take similar courses in the future. Therefore, the following hypothesis was postulated. H10. Intrinsic value is positively related to Web-based learning continuance intention. 3.9. Social isolation Social isolation occurs when there are fewer opportunities for a learner to interact with other learners and instructors. If they are geographically and temporally separated, interaction is generally asynchronous and text-based; this can lead to restricted socialization or feeling of isolation. Bennett et al. [3] indicated that online learning students tended to suffer from social isolation—resulting
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
in students withdrawing from the course. Daugherty and Funke [13] found that a feeling of isolation was an important criterion for student dissatisfaction with a Web-based course. Kahl and Cropley [24] suggested that distance learning students felt more isolated than face-to-face students and experienced lower levels of selfconfidence. Therefore, we made the following hypothesis. H11. Social isolation is negatively related to Web-based learning continuance intention. 3.10. Anxiety Anxiety is apprehension or discomfort experienced by an individual with technology. It is characterized as an affective response, an emotional fear of potential negative outcomes, such as low performance or exposure to an unknown audience. It has been argued that such negative feelings detract from task performance and have a significant impact on intention to adopt e-learning through perceived behavioral control. Therefore, the following hypothesis was proposed. H12. Anxiety is negatively related to Web-based learning continuance intention. 3.11. Delay in responses Delay in responses is a lack of immediacy in receiving responses from a system. Vonderwell [41] indicated that one disadvantage of online courses was their lack of immediate feedback from the instructor. Participants in Hara and Kling’s [22] study reported frustration due to lack of immediacy. Therefore, the following hypothesis was proposed. H13. Delay in responses is negatively related to Web-based learning continuance intention. 3.12. Risk of arbitrary learning Risk of arbitrary learning occurs when students have difficulty in self-motivation. A key feature of Web-based learning is control of the learning pace. ‘‘Self-paced instruction places a substantial burden on the student to maintain interest, focus, and pace’’ [7]. In a survey of learners’ perception of online learning, Song et al. [36] reported that most learners agreed that motivation and time management impacted the success of their online learning experience. Accordingly, we assumed that learners with low motivation or bad study habits tended to fall behind, leading to low learning performance, etc. H14. Risk of arbitrary learning is negatively related to Web-based learning continuance intention. 4. Research methodology 4.1. Sample and procedures Data were gathered from part-time students who took Webbased courses offered by a university in Taiwan. The Web-based learning service is a credit and voluntary program designed primarily for continuing education. The tuition fee was US$ 90 per credit. Face-to-face classes were only open to full-time students; thus part-time students can only take Web-based courses. Three thousand e-mails providing a hyperlink to our Web survey were sent to part-time students who had registered for at least one Web-based course. A total of 286 surveys were returned. Among the respondents, 51% were female and 49% male; 77% had
197
taken one Web-based course, 15% had taken two or three, and 8% had taken at least four. On average, the respondents were 34 years old and spent 20 h per week using the Internet. Measurement items were adapted from the literature wherever possible. New items were developed from information in the literature. A pretest of the questionnaire was conducted using four experts to assess logical consistency, ease of understanding, sequence of items, and task relevance. Overall, they stated that the questionnaire was clear and easy to complete. A number of suggestions were made about the wording of several items and the overall structure of the questionnaire; the suggestions were discussed and changes were made to the instrument. A pilot study with 20 part-time Master’s degree students who had taken Webbased course was also conducted. The purpose of this pilot study was to gain additional comments on the questionnaire content and structure. For the main study, the participants were directed to the homepage of our Web survey; this indicated the purpose and importance of the study. All respondents were guaranteed confidentiality of their separate responses. In order to control for the fact that respondents had different frames of reference, we asked them to answer the questionnaire items according to their experience with the specified Web-based learning system. Followup e-mails were sent to individuals not having responded within 3 weeks. To maximize the response rate, 30 randomly selected respondents were given a US$ 10 incentive. 4.2. Instrument Table 1 shows the instrument. For all the measures, a sevenpoint Likert type scale was adopted with anchors ranging from strongly disagree (1) to strongly agree (7). 4.3. Measurement model Confirmatory factor analysis (CFA) was applied to test the adequacy of the measurement model using LISREL 8.5. The adequacy of the measurement models was evaluated on the criteria of model fit, reliability, convergent validity, and discriminant validity. For a good model fit, the Chi-square value normalized by degrees of freedom (x2/d.f.) should not exceed 3, adjusted goodness of fit index (AGFI) should exceed 0.8, non-normed fit index (NNFI) and comparative fit index (CFI) should exceed 0.9, and root mean square error of approximation (RMSEA) should not exceed 0.08. For our CFA model, x2/d.f. was 1.43 (x2 = 1117; d.f. = 782), AGFI was 0.81, NNFI was 0.96, CFI was 0.96, and RMSEA was 0.039, suggesting adequate model fit. Reliability was examined using the composite reliability values. As shown in Table 1, all of them were above 0.7, indicating a commonly acceptable level for confirmatory research. Convergent validity was evaluated for measurement scales using two criteria suggested by Fornell and Larcker [19]: all indicator factor loadings should be significant and exceed 0.70 and average variance extracted (AVE) for each construct should exceed the variance due to measurement error for that construct (i.e., should exceed 0.50). Most items exhibited loading higher than 0.7 on their respective construct, providing evidence of acceptable item convergence on the intended constructs. Two exceptions were the third item of computer self-efficacy and the third item of the social influence with loadings slightly below 0.7. AVE ranged from 0.59 to 0.89 (see Table 2), greater than the variances due to measurement errors. Thus conditions for convergent validity were essentially met. Fornell and Larcker suggested that for satisfactory discriminant validity, the square root of the AVE from a construct should be greater than the correlation shared between the construct and
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
198 Table 1 Summary of measurement scales Construct
Measure
Performance expectancy (PE) (composite reliability = 0.85) PE1 Using the Web-based learning improves my performance in my learning activities PE2 Using the Web-based learning enhances my effectiveness on learning activities PE3 Using the Web-based learning in my learning activities increases my productivity PE4 I find the Web-based learning useful for my learning activities Effort expectancy (EE) (composite reliability = 0.89) EE1 It is easy for me to become skillful at using the Web-based learning system EE2 My interaction with the Web-based learning system is clear and understandable EE3 I find it easy to get the Web-based learning system to do what I want it to do EE4 I find the Web-based learning system to be easy to use
Mean (S.D.)
Loading
Source
5.30 (1.16)
0.75
Venkatesh et al.
5.33 (1.19)
0.73
5.31 (1.21)
0.78
5.76 (0.98)
0.83
5.99 (0.94)
0.89
5.45 (1.17)
0.76
5.83 (1.02)
0.83
5.97 (0.87)
0.86
Computer self-efficacy (CSE) (composite reliability = 0.84) (I could complete my learning activities using the Web-based system . . .) CSE1 If I had never used a system like it before 5.15 (1.31) 0.70 CSE2 If I had only the system manuals for reference 5.67 (1.00) 0.83 CSE3 If I had seen someone else using it before 5.86 (0.86) 0.69 trying it myself CSE4 If I had just the built-in help facility for 5.52 (1.06) 0.78 assistance Social influence (SI) (composite reliability = 0.89) SI1 People who influence my behavior think that I should participate in Web-based leaning activities SI2 People who are important to me think that I should participate in Web-based leaning activities SI3 The senior management of the organization has supported my participation in Web-based leaning activities Facilitating conditions (FC) (composite reliability = 0.82) FC1 I have the resources necessary to use the Web-based leaning system FC2 I have the knowledge necessary to use the Web-based leaning system FC3 The Web-based leaning system is compatible with other systems I use Attainment value (AV) (composite reliability = 0.91) AV1 I think Web-based learning makes me a more knowledgeable person AV2 I think Web-based learning offers a forum for fulfilling achievement AV3 I think being successful at Web-based learning confirms my competence AV4 I think being successful at Web-based learning give me a sense of confidence Utility value (UV) (composite reliability = 0.89) UV1 I think what I learn through Web-based learning is useful for my promotion UV2 I think what I learn through Web-based learning is useful for getting salary raise UV3 I think what I learn through Web-based learning is helpful for me to get a job Intrinsic value (IV) (composite reliability = 0.93) IV1 I think Web-based learning is interesting IV2 I think Web-based learning is enjoyable IV3 I think Web-based learning is fun Social isolation (SIL) (composite reliability = 0.93) SIL1 I think Web-based learning restricts socialization, due to the reduction of opportunities of face-to-face interactions among students SIL2 I think Web-based learning reduces opportunities of face-to-face interactions between students and the instructor Anxiety (AN) (composite reliability = 0.96)
Venkatesh et al.
Compeau and Higgins [12]
5.13 (1.29)
0.94
Venkatesh et al.
5.08 (1.30)
0.91
5.34 (1.19)
0.68
6.28 (0.75)
0.73
6.00 (0.87)
0.83
5.88 (0.79)
0.78
6.03 (0.79)
0.74
5.55 (1.17)
0.89
5.76 (1.06)
0.84
5.58 (1.13)
0.92
4.44 (1.52)
0.88
4.01 (1.57)
0.92
4.56 (1.47)
0.74
5.16 (1.19) 5.37 (1.14) 5.39 (1.15)
0.86 0.93 0.93
Battle and Wigfield
5.52 (1.11)
0.88
Eccles et al and Song et al.
5.53 (1.14)
0.99
Venkatesh et al.
Eccles et al. and Battle and Wigfield
Eccles et al. and Battle and Wigfield
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
199
Table 1 (Continued ) Construct AN1
AN2 AN3
Measure
Mean (S.D.)
Loading
Source
I feel apprehensive about using Web-based leaning to accomplish my learning tasks I feel uneasy about using Web-based leaning to accomplish my learning tasks I feel uncomfortable about using Web-based leaning to accomplish my learning tasks
2.98 (1.37)
0.94
Venkatesh et al.
2.87 (1.38)
0.99
2.73 (1.31)
0.90
4.37 (1.35)
0.95
4.39 (1.34)
0.95
4.30 (1.38)
0.93
3.29 (1.54)
0.88
3.27 (1.55)
0.92
3.28 (1.51)
0.95
3.20 (1.50)
0.85
5.88 (0.94)
0.92
5.87 (0.90)
0.97
5.71 (1.03)
0.86
Delay in response (DR) (composite reliability = 0.96) DR1 I felt a lack of immediacy in getting responses back from the instructor DR2 I felt a lack of immediacy in getting responses back from other students DR3 I felt a lack of immediacy in getting responses back from the teaching assistant Risk of arbitrary learning (RAL) (composite reliability = 0.94) RAL1 I did not consistently participate in the Web-based leaning activities RAL2 I was not able to maintain my focus in the Web-based leaning activities RAL3 I was not able to maintain my interest in the Web-based leaning activities RAL4 I was not able to accomplish my learning goals in the Web-based leaning activities Continuance intention (CI) (composite reliability = 0.94) CI1 If I could, I would like to continue using Web-based learning in my learning activities in the future CI2 It is likely that I will continue using Web-based learning in the future CI3 I expect to continue using Web-based learning in the future
Song et al.
Learning and Study Strategies Inventory [44] and Self-regulation Questionnaire [6]
Bhattacherjee [4]
less than 0.05. The explanatory power of the research model is also shown. The R2 value shows that performance expectancy, effort expectancy, computer self-efficacy, social influence, facilitating conditions, attainment value, utility value, intrinsic value, social isolation, anxiety, delay in response, and risk of arbitrary learning together accounted for 60% of the variance of Web-based learning continuance intention.
other constructs in the model. The diagonal values exceeded the inter-construct correlations; hence the test of discriminant validity was acceptable. Therefore we concluded that the construct validity of the measurement scales was sufficiently high. 4.4. Model testing results The overall fit and the strengths of the hypothesized paths were examined. As shown in Table 3, the fit indices were within accepted thresholds: A x2 to degrees of freedom ratio of 1.60 (x2 = 1284; d.f. = 801), AGFI = 0.80, NNFI = 0.94, CFI = 0.95, and RMSEA = 0.046. Hence, this model fitted the data reasonably well. The significance of individual paths is shown in Fig. 2 and summarized in Table 4. Nine out of 14 paths exhibited a P-value
5. Discussion and implications 5.1. Summary of results An interesting finding of our study was that performance expectancy (near-term usefulness) and utility value (long-term
Table 2 Squared correlations of latent variables Construct
AVE
PE EE CSE SI FC AV UV IV SIL AN DR RAL CI
0.59 0.66 0.57 0.73 0.60 0.72 0.73 0.82 0.88 0.89 0.89 0.81 0.84
PE 0.77 0.48 0.36 0.40 0.35 0.61 0.23 0.52 0.18 0.26 0.41 0.16 0.54
EE
0.81 0.50 0.30 0.53 0.48 0.22 0.52 0.13 0.38 0.44 0.19 0.57
CSE
0.75 0.29 0.49 0.40 0.17 0.37 0.07 0.27 0.27 0.05 0.47
SI
0.85 0.22 0.50 0.37 0.49 0.01 0.10 0.19 0.08 0.40
FC
0.78 0.42 0.16 0.42 0.01 0.43 0.29 0.06 0.44
AV
0.85 0.37 0.71 0.15 0.28 0.40 0.14 0.65
UV
0.85 0.47 0.01 0.07 0.14 0.03 0.38
IV
0.91 0.09 0.19 0.38 0.15 0.66
SIL
0.94 0.15 0.22 0.30 0.08
AN
0.95 0.52 0.24 0.35
DR
0.95 0.34 0.40
RAL
0.90 0.11
CI
0.92
Diagonal elements (in bold) are the square root of the average variance extracted (AVE). Off-diagonal elements are the correlations among constructs. For discriminant validity, diagonal elements should be larger than off-diagonal elements. PE = performance expectancy; EE = effort expectancy; CSE = computer self-efficacy; SI = social influence; FC = facilitating conditions; AV = attainment value; UV = utility value; IV = intrinsic value; SIL = social isolation; AN = anxiety; DR = delay in response; RAL = risk of arbitrary learning; CI = continuance intention.
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201
200 Table 3 Overall model fit indices for the research model Model fit indices
Results
Chi-square statistic x2/d.f. AGFI CFI NNFI RMSEA
1.60 (1283/801) 0.80 0.95 0.94 0.046
Table 4 Results of hypothesis testing Recommended value 3 0.80 0.90 0.90 <0.08
usefulness) had almost the same effects on continuance intention for working professionals with limited time for continuing education. Contrary to our hypotheses, social influence and facilitating conditions were not significant predictors of continuance intention. Additionally, social isolation, delay in response, and risk of arbitrary learning were not significant negative predictors of continuance. Apparently learners feel that social isolation and delay in response are negative aspects of Web-based learning, but they still have strong intentions to continue using it. The R2 value showed that performance expectancy, effort expectancy, computer self-efficacy, social influence and facilitating conditions together accounted for 46.6% of the variance of continuance intention. It increased when the three dimensions of positive subjective task value and four dimensions of cost were added as independent variables. 5.2. Implications for theory and practice Our research contributed to an overall conceptual understanding of the nature and the importance of components of subjective task value as continuance intention factors in continuing education. Apparently the perceived technological characteristics of Web-based learning themselves are insufficient to increase continuance intention. According to social exchange theory, individuals behave by rational self-interest. Therefore, continuance intention will be stimulated when rewards of Webbased learning outweigh its cost. From a practical perspective, our study implied that an understanding of what performance expectancy, effort expectancy, and positive subjective task value mean to learners is likely to establish longer-term relationships with learners. In addition, our findings also suggested that intrinsic value was the strongest predictor of an individual’s intention to continue using Web-based learning. Developers and designers of Web-based learning sites should presumably employ ways to reduce monotony and exploit learners’ playful characteristics. Social isolation, risk of arbitrary learning, and delay in response did not have significant negative effects on continuance intention. Thus, our findings implied that positive values or benefits out-
Hypotheses
Results
H1: Performance expectancy is positively related to Web-based learning continuance intention H2: Effort expectancy is positively related to Web-based learning continuance intention H3: Effort expectancy is positively related to performance expectancy H4: Computer self-efficacy is positively related to effort expectancy H5: Computer self-efficacy is positively related to Web-based learning continuance intention H6: Social influence is positively related to Web-based learning continuance intention H7: Facilitating conditions is positively related to Web-based learning continuance intention H8: Attainment value is positively related to Web-based learning continuance intention H9: Utility value is positively related to Web-based learning continuance intention H10: Intrinsic value is positively related to Web-based learning continuance intention H11: Social isolation is negatively related to Web-based learning continuance H12: Anxiety is negatively related to Web-based learning continuance intention H13: Delay in responses is negatively related to Web-based learning continuance intention H14: Risk of arbitrary learning is negatively related to Web-based learning
Supported Supported Supported Supported Supported Not supported Not supported Supported Supported Supported Not supported Supported Not supported Not supported
weighed the costs of Web-based learning. However, Web-based learning systems should provide well-designed tools or mechanisms for online interaction and instructors should always be available to provide quick responses, thus reducing the negative impact of social isolation and delay in response. 5.3. Limitations Although our findings were encouraging and useful, the study had several limitations. First, our results are not generalizable; we examined only one Web-based learning system and the subjects were adult learners (part-time students in continuing education). Factors influencing continuance intention of adult learners might be different from those of full-time students. Second, the results could be affected by self-selection bias. Our sample consisted of active participants of Web-based learning courses. Individuals who had already ceased to participate might have different perceptions about the influence of technological attributes of Web-based learning and subjective task value. Therefore, the results should be interpreted as only explaining continuance intention of current users of Webbased learning. Third, the data are cross-sectional. Individuals’ intention to use Web-based learning is an ongoing process. Our constructs were measured at a single point. Fourth, the usage of the Web-based learning service is voluntary (under users’ full volitional control). The findings might not be generalizable to a mandatory setting. Finally, our study focused on empirically validating the hypotheses derived from the research model.
References
Fig. 2. SEM analysis of the research model.
[1] A. Bandura, Social Foundations of Thought and Action: A Social Cognitive Theory, Prentice Hall, Englewood Cliffs, NJ, 1986. [2] A. Battle, A. Wigfield, College women’s orientations toward family, career, and graduate school, Journal of Vocational Behavior 62 (1) (2003) 56–75. [3] S. Bennett, A.M. Priest, C. Macpherson, Learning about online learning: an approach to staff development for university teachers, Australian Journal of Educational Technology 15 (3) (1999) 207–221. [4] A. Bhattacherjee, Understanding information systems continuance: an expectation-confirmation model, MIS Quarterly 25 (3) (2001) 351–370.
C.-M. Chiu, E.T.G. Wang / Information & Management 45 (2008) 194–201 [5] M. Bong, Role of self-efficacy and task-value in predicting college students’ course performance and future enrollment intentions, Contemporary Educational Psychology 26 (2001) 553–570. [6] J.M. Brown, W.R. Miller, L.A. Lawendowski, The self-regulation questionnaire, in: L. VandeCreek, T.L. Jackson (Eds.), Innovations in Clinical Practice: A Source Book, Professional Resource Press, Sarasota, FL, 1999, pp. 281–289. [7] California Distance Learning Project, What is Distance Learning? Available at http://www.cdlponline.org/index.cfm?fuseaction=whatis&pg=2, 2005. [8] C. Carlsson, J. Carlsson, K. Hyvo¨nen, J. Puhakainen, P. Walden, Adoption of mobile devices/services—searching for answers with the UTAUT, in: Proceedings of the 39th Hawaii International Conference on System Sciences, 2006. [9] A.D. Carswell, V. Venkatesh, Learner outcomes in an asynchronous distance education environment, International Journal of Human–Computer Studies 56 (2002) 475–494. [10] P.Y.K. Chau, An empirical assessment of a modified technology acceptance model, Journal of Management Information Systems 13 (2) (1996) 185–204. [11] E. Cody-Allen, R. Kishore, An extension of the UTAUT model with e-quality, trust, and satisfaction constructs, in: Proceedings of the 2006 ACM SIGMIS CPR Conference on Computer Personnel Research, 2006, pp. 82–89. [12] D.R. Compeau, C.A. Higgins, Computer self-efficacy: development of a measure and initial test, MIS Quarterly 19 (2) (1995) 189–211. [13] M. Daugherty, B.L. Funke, University faculty and student perceptions of Webbased instruction, Journal of Distance Education 13 (1) (1998) 21–39. [14] E.L. Deci, R.M. Ryan, Intrinsic Motivation and Self-determination in Human Behavior, Plenum Press, New York, NY, 1985. [15] W.H. DeLone, E.R. McLean, The DeLone and McLean model of information systems success: a ten-year update, Journal of Management Information Systems 19 (4) (2003) 9–30. [16] J.S. Eccles, Gender roles and women’s achievement-related decisions, Psychology of Women Quarterly 11 (2) (1987) 135–172. [17] J.S. Eccles, T.F. Adler, R. Futterman, S.B. Goff, C.M. Kaczala, J.L. Meece, C. Midgley, Expectancies, values, and academic behaviors, in: J.T. Spence (Ed.), Achievement and Achievement Motivation, W.H. Freeman, San Francisco, CA, 1983, pp. 75–146. [18] J.S. Eccles, T. Adler, J.L. Meece, Sex differences in achievement: a test of alternate theories, Journal of Personality and Social Psychology 46 (1) (1984) 26–43. [19] C. Fornell, D.F. Larcker, Evaluating structural equation models with unobservable and measurement error, Journal of Marketing Research 18 (1) (1981) 39–50. [20] M. Gong, Y. Xu, Y. Yu, An enhanced technology acceptance model for Web-based learning, Journal of Information Systems Education 15 (4) (2004) 365–373. [21] I. Hamburg, C. Lindecke, Social aspects of e-learning and blending learning methods, in: Proceedings of the 4th European Conference E-Comm-Line, Bucharest, Romania, 2003. [22] N. Hara, R. Kling, Students’ frustrations with a web-based distance education course, First Monday 4(12), available at http://www.firstmonday.dk/issues/ issue4_12/index.html, 1999. [23] P.M. Homer, L.R. Kahle, A structural equation test of the value-attitude-behavior hierarchy, Journal of Personality and Social Psychology 54 (4) (1988) 638–646. [24] T.N. Kahl, A.J. Cropley, Face-to-face versus distance learning: psychological consequences and practical implications, Distance Education 7 (1) (1986) 38–48. [25] J. Lu, C.S. Yu, C. Liu, Learning style, learning patterns, and learning performance in a WebCT-based MIS course, Information & Management 40 (2003) 497–507. [26] R. Mahmod, N. Dahlan, T. Ramayah, N. Karia, M.H.A.H. Asaari, Attitudinal belief on adoption of e-MBA program in Malaysia, Turkish Online Journal of Distance Education 6 (2) (2005) 115–124. [27] J.L. Meece, A. Wigfield, J.S. Eccles, Predictors of math anxiety and its consequences for young adolescents’ course enrollment intentions and performances in mathematics, Journal of Educational Psychology 82 (1) (1990) 60–70. [28] L. Morris, S. Wu, C. Finnegan, Predicting retention in online general education courses, The American Journal of Distance Education 19 (2005) 23–36. [29] C.S. Ong, J.Y. Laia, Y.S. Wang, Factors affecting engineers’ acceptance of asynchronous e-learning systems in high-tech companies, Information & Management 41 (2004) 795–804. [30] V. Osborn, The Distributed Learning Survey, University of North Texas, Denton, TX, 2000. [31] A. Parasuraman, D. Grewal, The impact of technology on the quality-value-loyalty chain: a research agenda, Journal of the Academy of Marketing Science 28 (1) (2000) 168–174. [32] L.A. Petrides, Web-based technologies for distributed (or distance) learning: creating learning-centered educational experiences in the higher education classroom, International Journal of Instructional Media 29 (1) (2002) 69–77.
201
[33] E.M. Rogers, Diffusion of Innovations, The Free Press, New York, NY, 1995. [34] R.G. Saade´, B. Bahli, The impact of cognitive absorption on perceived usefulness and perceived ease of use in on-line learning: an extension of the technology acceptance model, Information & Management 42 (2005) 317–327. [35] D. Sirdeshmukh, J. Singh, B. Sabol, Consumer trust, value, and loyalty in relational exchange, Journal of Marketing 66 (2002) 15–37. [36] L. Song, E.S. Singleton, J.R. Hill, M. Koh, Improving online learning: student perceptions of useful and challenging characteristics, Internet and Higher Education 7 (1) (2004) 59–70. [37] R.L. Thompson, C.A. Higgins, J.M. Howell, Personal computing: toward a conceptual model of utilization, MIS Quarterly 15 (1) (1991) 125–143. [38] H.C. Triandis, Values, attitudes, and interpersonal behavior, in: H.E. Howe, M.M. Page (Eds.), Nebraska Symposium on Motivation, 1979, University of Nebraska Press, Lincoln, NE, 1980, pp. 195–260. [39] V. Venkatesh, F.D. Davis, A theoretical extension of the technology acceptance model: four longitudinal field studies, Management Science 46 (2) (2000) 186– 204. [40] V. Venkatesh, M. Morris, G. Davis, F. Davis, User acceptance of information technology: toward a unified view, MIS Quarterly 27 (3) (2003) 425–478. [41] S. Vonderwell, An examination of asynchronous communication experiences and perspectives of students in an online course: a case study, Internet and Higher Education 6 (1) (2003) 77–90. [42] A.Y. Wang, M.H. Newlin, Predictors of web-student performance: the role f selfefficacy and reasons for taking an on-line class, Computers in Human Behavior 18 (2002) 151–163. [43] H.I. Wang, H.L. Yang, The role of personality traits in UTAUT model under online stocking, Contemporary Management Research 1 (1) (2005) 69–82. [44] C.E. Weinstein, D.R. Palmer, A.C. Schulte, Learning and Study Strategies Inventory, H & H Publishing, Clearwater, FL, 1987. [45] A. Wigfield, The role of children’s achievement values in the self-regulation of their learning outcomes, in: D.H. Schunk, B.J. Zimmerman (Eds.), Self-regulation of Learning and Performance: Issues and Educational Applications, Erlbaum, Mahwah, NJ, 1994, pp. 101–124. [46] A. Wigfield, J. Eccles, Relations of expectancies and values to students’ math grades and intentions, in: Paper Presented at the Meeting of the American Educational Research Association, San Francisco, CA, 1989. Chao-Min Chiu is a professor in the Department of Information Management at the National Central University, Taiwan (ROC). He holds a Ph.D. in management from the Rutgers University. His research interests include electronic commerce, virtual communities, and knowledge management. His research has appeared in Information & Management, Decision Support Systems, Information Systems Journal, International Journal of Human–Computer Studies, Behavior & Information Technology, Computers & Education, Information and Software Technology, Information Systems Management, Information Technology and Management, Journal of Information Science, and others. Eric T.G. Wang is information management chair professor at National Central University, Taiwan (ROC). He received the Ph.D. degree in business administration, specialized in computer & information systems, from the William E. Simon Graduate School of Business Administration, University of Rochester. His research interests include electronic commerce, outsourcing, organizational economics, and organizational impact of information technology. His research has appeared in Information & Management, Management Science, Information Systems Research, Journal of Management Information Systems, Decision Sciences, Decision Support Systems, Information Systems Journal, Omega, European Journal of Information Systems, European Journal of Operational Research, International Journal of Information Management, and others.