Computers & Education 49 (2007) 12241245

www.elsevier.com/locate/compedu

An empirical analysis of the antecedents of web-based learning continuance

Chao-Min Chiu
a

a,

, Szu-Yuan Sun b, Pei-Chen Sun c, Teresa L. Ju

Department of Information Management, National Central University, No. 300, Jungda Road, Jhongli City, Taoyuan 320, Taiwan, ROC b Department of Information Management, National Kaohsiung First University of Science and Technology, 1 University Road, Yenchao, Kaohsiung 824, Taiwan, ROC c National Kaohsiung Normal University, 116 Ho-Ping First Road, Kaohsiung 802, Taiwan, ROC d Department of Information Management, Shu-Te University, No. 59, Hun Shan Road, Hun Shan Village, Yen Chau, Kaohsiung, Taiwan, ROC Received 5 December 2005; received in revised form 21 January 2006; accepted 30 January 2006

Abstract Like any other product, service and Web-based application, the success of Web-based learning depends largely on learners satisfaction and other factors that will eventually increase learners intention to continue using it. This paper integrates the concept of subjective task value and fairness theory to construct a model for investigating the motivations behind learners intention to continue using Web-based learning. The model theorizes that four components of subjective task value (i.e., attainment, utility, intrinsic, and cost) and three dimensions of fairness (i.e., distributive, procedural, and interactional) aVect learners satisfaction. We also argue that satisfaction and four distinct components of subjective task value inXuence learners intention to continue using Web-based learning. The hypothesized model is validated empirically using data collected from 202 learners of a Web-based learning program designed for continuing education. The results showed that attainment value, utility value, intrinsic value, distributive fairness, and interactional fairness exhibited signi W cant positive

Corresponding author. Tel.: +886 3 4267251; fax: +886 3 4254604. E-mail address: cmchiu@mgt.ncu.edu.tw (C.-M. Chiu).

0360-1315/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.compedu.2006.01.010

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1225

eVects on satisfaction. Utility value and satisfaction play signiW cant roles in shaping learners intention to continue using Web-based learning. 2006 Elsevier Ltd. All rights reserved.
Keywords: Continuance intention; Fairness; Satisfaction; Subjective task value; Web-based learning

1. Introduction The proliferation of network access and advances in Internet/Web technology, in conjunction with social demands for improved access to higher education has facilitated the rapid growth of online learning or electronic learning (e-learning) (Lorenzetti, 2005). e-Learning also enables organizations to reduce the total cost of and increases the eYciency of training. According to International Data Corporation (IDC), the international e-learning market is growing by leaps and bounds. IDC estimated the international e-learning market to grow from US$ 6.6 billion in 2002 to US$ 23.7 billion in 2006, at a compound annual rate of 35.6%. Cortona Consulting said that the e-learning market could reach $50 billion in 2010. 70% of universities in the USA are now providing e-learning courses, according to the research of Market Data Retrieval. Concurrent with the organizational and universities interest in e-learning, a large number of academic papers (Arbaugh, 2002; Carswell & Venkatesh, 2002; Chiu, Hsu, Sun, Lin, & Sun, 2005) have been published on e-learning. These developments reXect the signiWcance of e-learning among scholars and practitioners. The goal of this study is to explore individuals intention to continue using Web-based learning in a voluntary setting. Web-based learning refers to learning delivered through a Web browser over the public Internet, private intranet or extranet. It is considered as a major subcomponent of the broader term e-learning. Like any Web-based application, the success of Web-based learning depends largely on user satisfaction and other factors that will eventually increase users intention to continue using it (continuance intention). The importance of continued use (continuance) is evident from the fact that customer turnover can be costly, especially given that it cost more to acquire new customers than to retain existing ones (Hart, Heskett, & Sasser, 1990; Reichheld & Schefter, 2000). In view of this, Web-based learning scholars and practitioners should look for ways to increase learners satisfaction levels and continuance intention. Technology acceptance model (TAM) (Davis, 1989) is one of the most widely used models for explaining an individuals behavioral intention and actual use of information technology (IT). Several recent studies drawing upon TAM have examined the eVects of the two salient beliefs about technological characteristics, namely perceived usefulness and ease of use, on learners attitude or behavioral intention in the context of Web-based learning or distance learning (Gong, Xu, & Yu, 2004; Lee, Cho, Gay, Davidson, & IngraVea, 2003; Stoel & Lee, 2003). However, those studies ignored the potentially important impacts of value and fairness. While the importance of technological characteristics cannot be denied, having well-designed Web-based learning sites/systems does not guarantee the success of Web-based learning. This is because value and fairness issues appear to be signiWcant in guiding a learners overall assessment of Web-based learning, thus inXuencing the learners continuance decision.

1226

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Web-based learning can be considered as an exchange of time, eVort, and money (inputs) for receiving skills, knowledge, grades, credits, or degrees (outputs) in return. According to Zeithaml (1988), it is the overall assessment of what is received and what is given that shapes learners intention to continue using Web-based learning. Researchers have conceptualized value as a function of a get component, i.e., the beneWts an individual receives, and a give component, i.e., an individuals monetary and non-monetary costs in acquiring and using a product or service (Parasuraman & Grewal, 2000; Sirdeshmukh, Singh, & Sabol, 2002). BeneWts include the extrinsic and intrinsic utility provided by the ongoing relationship with a service provider (Gwinner, Gremier, & Bitner, 1998). Literature in the marketing suggests that value (beneWts and costs) drives satisfaction, loyalty, behavioral intention to remain loyalty, and repurchase intention (Bolton & Drew, 1991; Neal, 1999; Patterson & Spreng, 1997). Value theorists argue that value is a centrally held and enduring belief and plays a central role in our everyday life decisions (Homer & Kahle, 1988; Rokeach, 1968). Equity theory theorizes that individuals seek a fair balance between input (what is given) and output (what is received). According to Adams (1965), individuals become satisWed and motivated whenever they feel their inputs are being fairly rewarded. Literature in the marketing and organization justice has aYrmed the importance of fairness considerations in the assessment of satisfaction (Maxham & Netemeyer, 2002; Smith, Bolton, & Wagner, 1999; Tax, Brown, & Chandrashekaran, 1998). Accordingly, a more complete study of the motivations underlying individuals intention to continue using Web-based learning should address issues related to value as well as fairness. To this end, two theories are applied and integrated: expectancy-value model of achievement motivation (Eccles et al., 1983) and fairness theory. Eccles et al.s (1983) expectancy-value model of achievement motivation, which is based on Atkinsons (1964) expectancy-value model, links individuals choice, persistence, and performance to expectancy for success and subjective task value. Eccles et al.s model outlines four motivational components of subjective task value: attainment value, intrinsic value, utility value, and cost. The model suggests that value related variables are likely to be more inXuential in situations of choice. Eccles and her colleagues have shown that subjective task value predicts course plans and enrollment decisions in mathematics, physics, and English courses (Eccles, 1987; Eccles, Adler, & Meece, 1984; Meece, WigWeld, & Eccles, 1990). This study follows Eccles et al. (1983) in arguing that subjective task value inXuences learners satisfaction and Web-based learning continuance intention through attainment value (the importance of doing well on Web-based learning), intrinsic value (playfulness of Web-based learning), utility value (helpfulness of Web-based learning to learners current and future career goals), and cost (negative aspects of engaging in Web-based learning). The concept of fairness or justice has long been studied in philosophy, political science, religion, organizational sciences, and economics (Yilmaz, Sezen, & Kabadayi, 2004). Fairness theory has its origins in equity theory. Equity theory (Adams, 1965) is a model of motivation that explains why people strive for fairness or justice in social exchange processes. According to Adamss (1965) equity theory, an individuals perception of the fairness of exchange relationships is determined by comparing the output/input ratio for oneself with that of referent others. When the ratios are equal, people are satisWed. Adams (1965) argued that people become demotivated, reduce input and/or seek change whenever they feel their inputs are not being fairly rewarded. Marketing and organizational justice researchers (Blodgett, Hill, & Tax, 1997; NiehoV & Moorman, 1993; Ramaswami & Singh, 2003; Tax et al., 1998) have identiWed three important dimensions of fairness: distributive, procedural, and interactional. This study follows prior research in arguing that

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1227

fairness inXuences learners satisfaction with Web-based learning through distributive fairness (fairness of the grades learners receive), procedural fairness (fairness of the procedures that produce the grades), and interactional fairness (fairness of the instructors treatment during online interaction). This paper makes two key contributions. First, although empirical evidence has demonstrated that subjective task value and fairness matter for individuals satisfaction or continuance intention, fundamental gaps remain in the understanding of value and fairness components that might explain individuals satisfaction and continuance intention in the context of Web-based learning. This study develops the measures for the components of subjective task value and fairness and tests their reliability and validity, and empirically examines their inXuences on satisfaction and continuance intention. Second, to the best of our knowledge, this is the Wrst study that examines the integrated inXuence of subjective task value and fairness on learners satisfaction and intention to continue using Web-based learning. In sum, by explicating the unique role of subjective task value and fairness, this paper aims at contributing to the continued development and success of Web-based learning in general.

2. Theoretical background 2.1. Web-based learning Web-based learning is a major subcomponent of the broader term e-learning or distance learning. Web-based learning is often called online learning. e-Learning refers to learning that content is delivered via the Internet, intranet, extranet, audio or video tape, satellite TV, and CD-ROM (Kaplan-Leiserson, 2000). e-Learning applications and processes include Web-based learning, computer-based training, virtual classrooms and digital collaboration. Distance learning refers to learning situation in which the instructor and learners are separated by time, location, or both (Perraton, 1988). Learning courses are delivered to remote locations via written correspondence, text, graphics, audio- and videotape, CD-ROM, online learning, audio- and videoconferencing, interactive TV, and FAX. The deWnition of distance education is broader than and entails the deWnition of e-learning. Numerous studies have discussed the reasons that learners choose Web-based learning or distance learning. Lorenzetti (2005) reported that time Xexibility is a major draw for students to choose to study online. Other reasons include less conXict with work (Dutton, Dutton, & Perry, 2002), the convenience of not having to travel to and attending a traditional face-to-face classroom (Motiwalla & Tello, 2000; Wegner, Holloway, & Garton, 1999), and the Xexibility of accessing course materials anytime and anywhere (wherever Internet connection is possible) (Hong, Lai, & Holton, 2003). Willis (1993) stated that adult learners pursue distance learning for a wide variety of reasons: constraints of time, distance, and Wnances (cost eVective). There have been considerable studies interested in exploring factors that inXuence learner satisfaction or outcomes in the context of Web-based learning or distance learning. For example, Carswell and Venkatesh (2002) examined the inXuence of technological characteristics (e.g., visibility, result demonstrability) on learners outcomes (i.e., intention, acceptance, and performance) in an asynchronous distance learning environment. Martins, Steil, and Todesco (2004) examined

1228

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

the eVect of innovation characteristics (e.g., observability and trialability) on adoption of the Internet as a learning tool. Other studies focused on the impact of personal cognition such as selfeYcacy (Lim, 2001; Wang & Newlin, 2002), computer conWdence (Osborn, 2000), and locus of control (Morris, Wu, & Finnegan, 2005). Arbaugh (2002) suggested examining the eVects of technological characteristics (e.g., perceived Xexibility) and behavioral characteristics (e.g., virtual immediacy behaviors) on student learning in and satisfaction with Web-based courses. Prior studies have shown that technological and personal cognition factors play signiWcant roles in Web-based learning, e-learning, or distance learning. However, those studies ignored the potentially important impact of value and fairness. According to Bandura (1986), of all the cognitive forces that guide human behavior and stand at the core of social cognitive theory is selfeYcacy. Bandura (1986) suggests that individuals who have high levels of self-eYcacy should be more likely to perform related behavior in the future. Web-based learning requires investment of time, eVort, and money in developing knowledge or receiving grades and credits. Accordingly, individuals who have high levels of self-eYcacy may have low intention to continue using Webbased learning if they think that Web-based learning has low levels of value and fairness. The overall assessment of the value and fairness of Web-based learning thus plays an important role in shaping an individuals satisfaction and intention to continue using Web-based learning. The question why do individuals spend money and their valuable time and eVort on Web-based learning, should be addressed from the perspectives of both value and fairness. Consequently, the subjective task value concept in the expectancy-value model of achievement motivation and fairness theory are applied and integrated for addressing our research question. 2.2. Subjective task value Expectancy-value model of achievement motivation (Eccles et al., 1983) posits that individuals performance, persistence, and choice are most directly predicted by their expectancies for success on the tasks and the subjective task value they attach to success on the tasks. Eccles et al. (1983) deWned diVerent components of subjective task value: attainment value, intrinsic value, utility value, and cost. Attainment value is the personal importance of doing well on the task. According to Eccles et al. (1983), it incorporates a variety of dimensions, including perceptions of the tasks ability to conWrm salient and valued characteristics of the self (e.g., masculinity, femininity, competence), to provide a challenge, and to oVer a forum for fulWlling achievement, power and social needs. Intrinsic value is the enjoyment an individual gets from performing the activity, or the subjective interest an individual has in the subject. Utility value is how well a task relates to current and future goals, such as career goals. Finally, cost is conceptualized in terms of the negative aspects of engaging in the task (e.g., performance anxiety and fear of both failure and success), as well as both the amount of eVort that is needed to succeed and the lost of opportunities resulting from making one choice rather than another. In studying how expectancies and subjective task value inXuence elementary through secondary school students performance and choice, Eccles and her colleagues (1983) showed that components of subjective task value predicted both intentions and actual decisions to keep taking mathematics and English in the context of traditional classroom education. WigWeld and Eccles (1989) found that components of subjective task value predicted junior and senior high school students intention to keep taking mathematics. Battle and WigWeld (2003) also showed that components of

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1229

subjective task value predicted intentions to attend graduate school. However, the possible impact of the components of subjective task value is still unclear in the Web-based learning context. No empirical work has been done to address this issue. This begs the question whether an individuals assessment of what is received (attainment value, intrinsic value, and utility value) and what is given (cost) is strong enough to stimulate or inhibit his or her intention to continue using Web-based learning. 2.3. Fairness theory The earliest inXuential theories of fairness or justice were the rule of distributive justice (Homans, 1961) and the equity theory (Adams, 1965). Homans (1961) simple formula for justice stressed the diVerence between the rewards people received for investments. Researchers have conceptualized three types of fairness: distributive, procedural, and interactional. Distributive fairness refers to the perceived fairness of the outcomes an individual receives (Homans, 1961). Procedural fairness refers to the perceived fairness of the procedures that produce the outcome (Thibaut & Walker, 1975). Thibaut and Walker (1975) identiWed two types of control as essential determinants of procedural fairness: control over the presentation of evidence (process control) and control over the Wnal decision (decision control). Leventhal (1980) suggested six procedural fairness rules, including consistency, bias suppression, accuracy, correctability (an appeal mechanism to appeal decisions), representativeness, and ethicality. Finally, Bies and Moag (1986) separated out interpersonal aspect of procedural justice, referred to as interactional fairness, which emphasizes the perceived fairness of the interpersonal treatment received during the implementation of a procedure (Bies & Moag, 1986). Bies and Moag (1986) identiWed truthfulness (honesty and avoiding deception), courtesy, respect for individual rights, propriety of behavior (avoiding prejudice), and justifying decisions as typifying fair treatment. Learners spend valuable time and eVorts on Web-based learning thus expect that the grades they receive, the procedures for evaluating their inputs, and the treatment by instructors during online interaction are fair. Fairness has been studied to a limited degree in marketing in conjunction with relationship quality (Kumar, Scheer, & Steenkamp, 1995), organizational citizenship (NiehoV & Moorman, 1993), work outcomes (Ramaswami & Singh, 2003), and service recovery (Blodgett et al., 1997). Research has shown the important role of fairness in explaining job satisfaction (Moorman, 1991), satisfaction with complaint handling (Tax et al., 1998), satisfaction with service encounter (Smith et al., 1999), satisfaction with recovery (Maxham & Netemeyer, 2002), and reseller satisfaction (Yilmaz et al., 2004). However, the possible impact of three dimensions of fairness is still unclear in the Web-based learning context. No empirical work has been done to address this issue. This begs the question whether an individuals perceptions of fair treatment in the Web-based learning process is strong enough to enhance his or her satisfaction with Web-based learning.

3. Research model and hypotheses The study draws on the concept of subjective task value in expectancy-value model of achievement motivation and fairness theory to investigate the inXuences of value and fairness on learners

1230

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Distributive Fairness

H1

Attainment Value H4 H5

Utility Value H6 H12 H7 Continuance Intention H10 Cost H11

Procedural Fairness

H2 Satisfaction H3

H8 Intrinsic Value

H9

Interactional Fairness

Fig. 1. Research model for Web-based learning continuance intention.

satisfaction and continuance intention in the context of Web-based learning. Following prior research (Blodgett et al., 1997; Moorman, 1991; NiehoV & Moorman, 1993; Tax et al., 1998), we deWne fairness with three distinct dimensions: distributive, procedural, and interactional Following Eccles et al. (1983), we identiWed four components of subjective task value: attainment value, utility value, intrinsic value, and cost. The proposed theoretical model is shown in Fig. 1. The dependent variable is Web-based learning continuance intention, which refers to the subjective probability that an individual will continue using Web-based learning. The basic assumption is that Web-based learning continuance intention is determined by attainment value, utility value, intrinsic value, cost, and satisfaction. Satisfaction, in turn, is jointly determined by distributive fairness, procedural fairness, interactional fairness, attainment value, utility value, intrinsic value, and cost. 3.1. Distributive fairness In this study, distributive fairness refers to learners perceived fairness of the grades they receive. Equity theory (Adams, 1965) postulates that individuals who are fairly rewarded experience satisfaction and will be motivated to engage in a certain behavior. According to Kumar et al. (1995), distributive fairness is helpful in building good relationships between instructors and learners, which in turn will lead to learners satisfaction. While the inXuence of distributive fairness on learners satisfaction has not been explicitly examined in the Web-based learning research, support for the relationships can be found in other settings. For example, some studies found that distributive fairness exerted a signiWcant inXuence on satisfaction with job (Lind & Tyler, 1988; Moorman, 1991), service encounter (Smith et al., 1999), and complaint handling (Tax et al., 1998). Therefore, the following hypothesis is proposed. H1: Distributive fairness is positively related to learners satisfaction with Web-based learning. 3.2. Procedural fairness In this study, procedural fairness refers to the perceived fairness of the procedures that produce the grade. Folger and Greenberg (1985) argued that how outcomes are determined may be more

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1231

important than the actual outcomes. Prior research indicates that if individuals believe that the procedures used to produce the outcomes are fair, they are likely to be satisWed with the outcomes, even if the outcomes are considered unfair (Lind & Tyler, 1988). Both organizational justice theorists (Folger & Konovsky, 1989) and consumer behavior researchers (Maxham & Netemeyer, 2002) suggested that perception of procedural fairness enhances the probability of maintaining a long-term overall satisfaction between parties. According to Seiders and Berry (1998), grading process is an integral part of the Web-based learning service, thus the instructors can enhance learners satisfaction with Web-based learning by engaging activities that enhance learner perceptions of procedural justice. Prior research has found relationships between procedural fairness and satisfaction with variables such as pay (Folger & Konovsky, 1989), reseller (Yilmaz et al., 2004), service encounter (Smith et al., 1999), and complaint handling (Tax et al., 1998). Accordingly, this study theorizes that if the instructors grading process is perceived by learners as fair, they are more satisWed with Web-based learning. H2: Procedural fairness is positively related to learners satisfaction with Web-based learning. 3.3. Interactional fairness In this study, interactional fairness refers to the extent to which learners feel they have been treated fairly regarding their online interaction with the instructor throughout the Web-based learning process. Since interaction between learners and instructors plays an important role in the online learning process (PalloV & Pratt, 1999), instructors can enhance learners satisfaction with the Web-based learning by engaging activities that enhance learner perceptions of interactional fairness. This notion has some support in other settings. Maxham and Netemeyer (2002) found eVects of interactional fairness on satisfaction with recovery and overall Wrm satisfaction. Tax et al. (1998) reported a strong eVect of interactional fairness on satisfaction with complaint handling, while Smith et al. (1999) demonstrated that positive perceptions of interactional fairness signiWcantly enhanced customer satisfaction with service counter. Therefore, the following hypothesis is proposed. H3: Interactional fairness is positively related to learners satisfaction with Web-based learning. 3.4. Attainment value In this study, attainment value represents the importance of doing well on Web-based learning in terms of the learners self-schema and core personal values. WigWeld and Eccles (1992) argued that tasks will have higher attainment value to the extent that they allow the individual to conWrm salient aspects of his or her self-schema. Attainment value is similar to the notions of integrated regulation of extrinsic motivation in self determination theory (Ryan & Deci, 2000). In integrated regulation, individuals assimilate identiWed importance of a behavior into their coherent sense of self, and thus fully accept it as their own. Ryan and Deci (2000) argued that when identiWcations have become fully integrated, one will behave with a true sense of volition and willingness. In general, students will be more likely to engage in a task, expend more eVort on it, and do better on it when they value it (WigWeld, 1994). Meece et al. (1990) found that the attainment value junior high school students attached to math competence predicted their intentions to continue taking math. Therefore, the following hypotheses are proposed.

1232

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

H4: Attainment value is positively related to learners satisfaction with Web-based learning. H5: Attainment value is positively related to learners intention to continue using Web-based learning. 3.5. Utility value In this study, utility value is how well taking Web-based learning courses relates to current and future career goals, such as solving problems at work, promotion, and pursuing advanced studies. Utility value of an activity can be tied to the construct of extrinsic motivation in self determination theory (Deci & Ryan, 1985), in that it taps more instrumental reasons for engaging in a task. Extrinsic motivation refers to doing something because it leads to a separable outcome. Eccles and WigWeld (2002) argued that a learning activity can have positive values to a person because it facilitates important future goals, even if he or she is not interested in the learning activity for its own sake. Ryan and Deci (2000) argued that students can perform extrinsically motivated actions with willingness that reXects an inner acceptance of the value or utility of task. Utility value has long been identiWed as a major inXuence on students achievement choices and behaviors. In a study of college womens value orientations toward family, career, and graduate school, Battle and WigWeld (2003) found that utility value predicted intentions to attend graduate school. Several researchers have found that students perceptions of the utility of mathematics are strongly related to their intentions to continue or discontinue their mathematical study (Brush, 1980; Fennema & Sherman, 1977). Sullins, Hernandez, Fuller, and Tashiro (1995) found that value factors signiWcantly predicted intention to enroll in future biology courses. In consumer behavior, perceived value has been deWned as the consumers overall assessment of the utility of a product based on perceptions of what is received and what is given (Zeithaml, 1988). A number of studies have demonstrated the eVect of personal value on satisfaction and behavioral intention. Spreng, Dixon, and Olshavsky (1993) argued that value should be a direct antecedent of satisfaction. Bojanic (1996) suggested that high levels of perceived value result in purchase and ultimately higher levels of customer satisfaction. Patterson and Spreng (1997) found that perceived value had a signiWcant eVect on customer satisfaction in the context of business-to-business professional services. Therefore, the following hypotheses are proposed. H6: Utility value is positively related to learners satisfaction with Web-based learning. H7: Utility value is positively related to learners intention to continue using Web-based learning. 3.6. Intrinsic value This study follow Davis, Bagozzi, and Warshaws (1992) deWnition of intrinsic value as the extent to which the activity of using a Web-based learning is perceived to be personally enjoyable in its own right aside from the instrumental value of the technology. According to self determination theory, learners are self-determining and intrinsically motivated in Web-based learning when they are interested in it or enjoy doing it. Triandis (1980) argues that aVect (e.g., the feelings of joy, elation, and pleasure) has an impact on behavioral intentions. Hsu and Chiu (2004) extended and empirically validated the theory of planned behavior (Ajzen, 2002) in the electronic service (e-service) context. The results showed that perceived playfulness had a signiWcant eVect on

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1233

individuals satisfaction with e-service. WigWeld and Eccles (1989) indicated that junior and senior high school students interest in math predicted their intentions to keep taking math. In a study of college students course performance and future enrollment intentions, Bong (2001) found that students who were intrinsically interested in topics covered in the present course would be more willing to take similar courses in the future. Therefore, the following hypotheses are proposed. H8:Intrinsic value is positively related to learners satisfaction with Web-based learning. H9: Intrinsic value is positively related to learners intention to continue using Web-based learning. 3.7. Cost Cost refers to negative aspects of engaging in Web-based learning. In this study, we focus on restriction on learners interactions with other learners and instructors, and feeling of isolation or not being part of the learning community. Online learning is very diVerent from face-to-face instruction. Learners are geographically separated and thus most of the time, the interaction among learners is based on asynchronous and text-based communication, which can lead to restriction on socialization or felling of isolation. Online learning students tend to suVer from isolation a signiWcant factor contributing to withdrawal from study (Bennett, Priest, & Macpherson, 1999). Kahl and Cropley (1986) suggested distance learning students feel more isolated than face-to-face students and experience lower levels of self-conWdence as a result. This has a signiWcant eVect on the distance learning students motivation and learning and even can lead them to drop out. Arbaugh (2000) found that perceived interaction diYculty was negatively correlated with students satisfaction with Web-based MBA courses. Billings, Conners, and Skiba (2001) found that students felt somewhat isolated from other students and instructors in the Web-based learning environment. Isolation was negatively related to satisfaction with Webbased courses. Battle and WigWeld (2003) found that cost was a signiWcant and negative predictor of college womens intentions to attend graduate school. Therefore, the following hypotheses are proposed. H10: Cost is negatively related to learners satisfaction with Web-based learning. H11: Cost is negatively related to learners intention to continue using Web-based learning. 3.8. Satisfaction Satisfaction is an individuals feelings of pleasure or disappointment resulting from comparing the perceived performance (or outcome) of Web-based learning in relation to his or her expectations. Oliver (1980) theorized that satisfaction is positively correlated with future intention, both directly and indirectly via its impact on attitude. In the Wnal step of satisfaction formation processes, satisfaction determines intentions to patronize or not to patronize the store in the future (Swan & Trawick, 1981). McGorry (2003) argued that student satisfaction with online learning courses is likely to determine whether the student takes subsequent courses. Chiu et al. (2005) reported that learners satisfaction with e-learning was signiWcantly associated with their continuance intention. Therefore, the following hypothesis is proposed. H12: Satisfaction is positively related to learners intentions to continue using Web-based learning.

1234

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

4. Research methodology 4.1. Measurement development Measurement items were adapted from the literature wherever possible. A pretest of the questionnaire was performed using six experts in the IS area to assess its logical consistency, ease of understanding, sequence of items, and contextual relevance. The comments collected from these experts led to several minor modiW cations of the wording and the item sequence. Furthermore, a pilot study was conducted involving 20 master students who ever took Webbased learning course. Comments and suggestions on the item contents and structure of the instrument were solicited. The questionnaire was further modi Wed based on the comments and suggestions. Items for measuring attainment value were adapted from Eccles et al. (1983) and Battle and WigWeld (2003). The items measured the importance of doing well on Web-based learning. Items for measuring utility value were also adapted from Eccles et al. (1983) and Battle and WigWeld (2003). These items focused on how well taking Web-based learning courses relates to learners current and future career goals, such as getting a job, solving problems at work, promotion, and pursuing advanced studies. Items for intrinsic value reXect whether Web-based learning is interesting, enjoyable and fun, similar to those applied by Battle and WigWeld (2003) and Moon and Kim (2001). Items for measuring cost were based on Eccles et al.s (1983) concept. These items measured four negative eVects of Web-based learning. Distributive fairness was measured with items adapted from Folger and Konovsky (1989). The measure focused on the extent to which the grade received by a learner was perceived to be related to his or her eVort, performance, and the amount of time spent. Procedural fairness was assessed with items adapted to reXect the degree to which fair procedures were used by the instructor in arriving at the grading decisions, following Folger and Konovsky (1989) and Moorman (1991). Interactional fairness was measured with items based on Folger and Konovsky (1989) and Moorman (1991), focusing on the instructors interaction with learners. SpeciWc items asked whether the instructor showed concern to the learners questions, whether the instructor treated the learner with dignity and respect, and whether the instructor dealt with the learner in a truthful manner. Items related to satisfaction were adapted from Oliver (1980) and Oliver and Swan (1989). Items for measuring continuance intention were adapted from prior work by Mathieson (1991) and Bhattacherjee (2001). For all the measures, a seven-point Likert scale was adopted with anchors ranging from strongly disagree (1) to strongly agree (7). 4.2. Survey administration Data were gathered from students of a Web-based learning program provided by National Kaohsiung Normal University (NKNU) in Taiwan. The program provides credits for individuals interested in special education and computer skills. The Web-based learning service is a Unixbased system and its learning is delivered via blend of asynchronous and synchronous technologies. Synchronous Web-based learning service consists of real-time interaction between learners and instructors, facilitated by tools such as Web Meeting (videoconferencing) and chat rooms. Asynchronous Web-based learning service is a self-study that may be supplemented by non-real

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1235

time interaction with the instructor, for instance through email, message boards, and forums. The survey began on October 4, 2004 and ended on November 25, 2004. The Wrst page of the questionnaire explained the purpose of this study and ensured the conWdentiality. A total of 600 surveys were distributed to individuals who at least took one course oVered by the NKNU Web-based learning service. A total of 221 surveys were returned. The exclusion of responses from incomplete questionnaires resulted in a total of 202 usable questionnaires (a net response rate of 33.7%). Among the respondents, 69% were female and 31% were male. Fifty-six percent have earned a bachelors degree and 25% have received a masters degree. Among the respondents, 80% ever took one Web-based course, 15% ever took two or three Web-based courses, and 5% ever took at least four Web-based courses. On average, the respondents were 28 years of age and spent 17 h per week in using the Internet. 4.3. Data analysis Data analysis utilized a two-step approach as recommended by Anderson and Gerbing (1988). The Wrst step involves the analysis of the measurement model, while the second step tests the structural relationships among latent constructs. The aim of the two-step approach is to assess the reliability and validity of the constructs before their use in the full model. ConWrmatory factor analysis (CFA) was applied to assess the construct validity of the nine scales (attainment value, utility value, intrinsic value, cost, distributive fairness, procedural fairness, interactional fairness, satisfaction, and continuance intention) with LISREL. Each item was modeled as a reXective indicator of its latent construct. The nine constructs were allowed to co-vary freely in the CFA model. Model estimation was done using the maximum likelihood approach, with the item correlation matrix as input. Table 1 presents the results of the CFA analysis. For measurement models to have suYciently good model Wt, the 2 value normalized by degrees of freedom ( 2/df) should not exceed 3, goodness-of-Wt index (GFI), Adjusted GFI (AGFI) should exceed 0.8 (Doll, Xia, & Torkzadeh, 1994), non-normed Wt index (NNFI) and comparative Wt index (CFI) should exceed 0.9, and root mean square error of approximation (RMSEA) should not exceed 0.08. As shown in Table 2, the Wt indices are within accepted thresholds, except for AGFI, which is slightly lower than the commonly cited threshold: 2 to degrees of freedom ratio of 1.45 ( 2 D 797; df D 551), GFI D 0.82, NNFI D 0.95, CFI D 0.95, and RMSEA D 0.047. According to Boudreau, Gefen, and Straub (2001) and Gefen, Karahanna, and Straub (2003), it is common that not all Wt indices are beyond acceptable thresholds in LISREL. Hence, this model Wtted reasonably well with the data. Reliability was examined using the composite alpha values. As shown in Table 1, all the values were above 0.7; this indicates a commonly acceptable level for explanatory research. Additionally, convergent validity of the resulting scales was veriWed by using two criteria suggested by Fornell and Larcker (1981): (1) all indicator loadings should be signiWcant and exceed 0.7 and (2) average variance extracted (AVE) by each construct should exceed the variance due to measurement error for that construct (i.e., AVE should exceed 0.50). For the current CFA model, only two of the 36 loadings were slightly below the 0.7 threshold (see Table 1). AVE ranged from 0.54 to 0.81 (see Table 3), greater than variance due to measurement error. Hence, the convergent validity was strongly supported.

1236

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Table 1 Summary of measurement scales Construct Attainment value (AV) AV1 AV2 AV3 AV4 Utility value (UV) UV1 UV2 UV3 UV4 Intrinsic value (IV) IV1 IV2 IV3 Cost (C) C1 C2 C3 C4 Distributive fairness (DF) DF1 DF2 DF3 DF4 Procedural fairness (PF) PF1 PF2 PF3 Measure Composite alpha D 0.89 The amount of eVort it took through Web-based learning was worthwhile to me Web-based learning made me a more knowledgeable person Web-based learning broadened my view Being successful at Web-based learning conWrmed my competence Composite alpha D 0.83 What I learned through Web-based learning was helpful for me to get a job What I learned through Web-based learning was useful for resolving my problems at work What I learned through Web-based learning was useful for my promotion The credits I got through Web-based learning were useful for pursuing advanced studies Composite alpha D 0.92 I think Web-based learning is interesting I think Web-based learning is enjoyable I think Web-based learning is fun Composite alpha D 0.90 Web-based learning reduced opportunities for interaction among learners Web-based learning reduced opportunities of studentteacher dialogue Web-based learning reduced the sense of being part of the learning community Web-based learning reduced opportunities for discussing with and learning from other learners Composite alpha D 0.91 The grade I received was fair considering my eVort The grade I received was fair considering my performance The grade I received was fair considering the amount of time I spent Overall, the grade I received was fair Composite alpha D 0.91 The instructor collected accurate information necessary for making the grading decisions The instructor provided opportunities to appeal or challenge the grading decisions The instructor generated standard so the grading decisions could be made with consistency Mean 5.77 6.00 6.02 5.77 SD 0.85 0.78 0.77 0.95 Loading 0.79 0.93 0.89 0.63

5.28 5.55 4.86 5.70

1.11 0.94 1.15 0.95

0.78 0.74 0.62 0.79

5.40 5.29 5.27 4.80 4.83 5.21 4.85

0.98 1.00 1.00 1.35 1.35 1.26 1.39

0.78 0.95 0.92 0.78 0.81 0.87 0.86

4.88 4.84 4.94 5.02 4.73 4.76 4.74

1.10 1.10 1.04 1.00 1.13 1.21 1.09

0.80 0.87 0.93 0.81 0.84 0.80 0.86

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245 Table 1 (continued) Construct IF4 IF5 Satisfaction (S) S1 S2 S3 S4 Continuance intention (CI) CI1 CI2 CI3 Measure The instructor provided me with timely feedback to my questions The instructor provided appropriate answers or suggestions to my questions or problems Composite alpha D 0.91 My decision to use Web-based learning is a wise one I think Web-based learning is a good idea I am pleased with the experience of using Web-based learning I am satisWed with the experience of using Webbased learning Composite alpha D 0.93 I intend to continue using the Web-based learning in the future I will continue using the Web-based learning as much as possible in the future I will continue using the Web-based learning in the future Mean 5.05 5.08 SD 1.12 1.05

1237

Loading 0.86 0.89

5.69 5.57 5.49 5.42

0.91 0.96 0.92 0.98

0.70 0.77 0.93 0.90

5.50 5.20 5.41

0.98 1.14 0.99

0.93 0.83 0.93

Table 2 Model Wt indices for the measurement model Model Wt indices statistic GFI AGFI CFI NNFI RMSEA
2 2

Results 1.45 0.82 0.78 0.95 0.95 0.047

Recommended value 63 70.8 70.8 70.9 70.9 60.08

/df

Finally, discriminant validity of the resulting scales was assessed using the guideline suggested by Fornell and Larcker (1981): the square root of the AVE from the construct should be greater than the correlation shared between the construct and other constructs in the model. Table 3 lists the correlations among the constructs, with the square root of the AVE on the diagonal. All the diagonal values exceed the inter-construct correlations; hence the test of discriminant validity was acceptable. Therefore, we conclude that the measures have demonstrated suYcient construct validity. The structural model reXecting the assumed linear, causal relationships among latent constructs was tested with the data collected from the validated measures. The structural model yielded an acceptable Wt: 2/df D 1.45 ( 2 D 802; df D 554), GFI D 0.82, AGFI D 0.78, CFI D 0.95, NNFI D 0.94, and RMSEA D 0.047. The test results of the structural model are summarized in Fig. 2. Seven out of the 12 paths exhibited a P-value less than 0.05, while the remaining Wve were not signiWcant at the 0.05 level of signiWcance. The paths from distributive fairness, interactional fairness, attainment value, utility

1238 Table 3 Correlations and AVE Construct AV UV IV C DF PF IF S CI AVE 0.67 0.54 0.79 0.69 0.73 0.68 0.79 0.69 0.81

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Construct AV 0.82 0.64 0.63 0.02 0.38 0.35 0.35 0.62 0.60 UV 0.73 0.62 0.14 0.34 0.34 0.38 0.53 0.59 IV C DF PF IF S CI

0.89 0.10 0.49 0.44 0.44 0.63 0.63

0.83 0.03 0.05 0.09 0.00 0.02

0.85 0.71 0.52 0.62 0.62

0.82 0.64 0.54 0.53

0.89 0.50 0.48

0.83 0.82

0.90

Diagonal elements (in bold) are the square root of the average variance extracted (AVE). OV-diagonal elements are the correlations among constructs. For discriminant validity, diagonal elements should be larger than oV-diagonal elements. AV, attainment value; UV, utility value; IV, intrinsic value; C, cost; DF, distributive fairness; PF, procedural fairness; IF, interactional fairness; S, satisfaction; CI, continuance intention.

value, and intrinsic value to satisfaction were signiWcant, while procedural fairness and cost showed no signiWcant inXuence on satisfaction Consequently, hypotheses 1, 3, 4, 6 and 8 were supported empirically while hypotheses 2 and 10 were not supported. Utility value signiWcantly and positively aVected continuance intention, while attainment value, intrinsic value, and cost did not exhibit signiWcant positive eVects on continuance intention. Hypothesis 7 was supported while hypotheses 5, 9, and 11 were not supported. Satisfaction exhibited a strong positive eVect on continuance intention, supporting hypothesis 12. The explanatory power of the research model is also shown in Fig. 2. The R2 value for continuance intention shows that attainment value, utility value, intrinsic value, cost, and satisfaction account for 66% of variance of continuance intention. The R2 value for satisfaction shows that

Distributive Fairness

0.48*

Attainment Value 0.20* 0.01

Utility Value 0.19* 0.81* 0.15* R2 = 0.66 Continuance Intention 0.05 Cost

Procedural Fairness

-0.10

Satisfaction R2 = 0.72 0.15*

0.14* Interactional Fairness

0.00

0.01

Intrinsic Value

*p-value < 0.05

Fig. 2. SEM analysis of research model.

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1239

distributive fairness, procedural fairness, interactional fairness, attainment value, utility value, intrinsic value, and cost account for 72% of variance of satisfaction.

5. Discussion and conclusions Creating perceptions of value is a primary means of enhancing learners feeling of satisfaction and developing long-term relationship with them, which is as important as learners perceptions of fairness. This study provides an initial step toward understanding the integrated inXuence of subjective task value and fairness on learners satisfaction and continuance intention in the context of Web-based learning. The Wndings indicated that learners who experienced higher levels of distributive fairness and interactional fairness were more satisWed with Web-based learning. Contrary to our expectations, we found that procedural fairness did not have a signiWcant impact on satisfaction This Wnding is inconsistent with Folger and Greenbergs (1985) argument that how outcomes are determined may be more important than the actual outcomes. One possible explanation for the insigniWcant relationship between procedural fairness and satisfaction is that higher levels of distributive fairness and interactional fairness compensate for lower levels of procedural fairness. In other words, learners may be satisWed with the Web-based learning when a low quality of procedures were used by the instructor to arrive at grading decisions, provided that the instructor dealt with them in a truthful manner, the instructor showed concern to their questions or problems, and the grades they received were fair. The grading of the National Kaohsiung Normal Universitys (NKNU) Web-based learning service is based on learners midterm and Wnal exams, term reports, and participation in online interaction (oYce hours). Learners would receive the Wnal grades through mail two or three weeks latter after the semester ended. To enhance learners satisfaction, this study suggests that the NKNU Web-based learning system should provide a real-time feedback mechanism that allows learners to access the details about grading process. For example, the mechanism could show learners grades of each part (e.g., exams, reports, and online interaction) and their ranking in the whole class for each part. Instructors of the NKNU Web-based learning service were scheduled to oVer six times of online interaction during the semester and each interaction should last three hours. The Wndings indicated that even though each instructor had about 50 learners, he or she could still oVer fair treatment to learners during online interaction, which in turn increased learners satisfaction with Web-based learning. The Wndings suggest that a learner should collect information about whether or not instructors are fair in grading and treating learners during online interaction before selecting a course in order to enhance his or her satisfaction with the Web-based course. The Wndings also suggest that Web-based learning providers (academics) could provide some reward mechanisms to stimulate instructors to deal with learners in a truthful manner and show concern and respect to learners during online interaction. Additionally, the Wndings indicated that the inXuences of attainment value, utility value, and intrinsic value on satisfaction were signiWcant, whereas the inXuence of cost was not signiWcant. One possible explanation for the Wnding is that learners participating in Web-based learning do not really expect extensive interaction. The Wndings suggest that learners may be satisWed with the Web-based learning even when negative aspects of it existed, provided that doing well on it is important to them, taking Web-based learning courses relates to their current and future career

1240

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

goals, and it is interesting, enjoyable and fun. Prensky (2001) argued that computer games give fun in the learning process and thus motivate individuals to learn. Klaila (2001) argued that elearning consumers should expect programs that incorporate the same innovative tools and techniques used in the computer gaming industry such as graphics, interaction, and skill-building challenges to deliver an educational experience thats compelling, informative, and fun. Neal and Normore (2005) suggested that surprising and challenging situations, collaboration, interactivity, and Xexibility of the program make e-learning fun. Accordingly, developers and designers of Web-based learning sites can employ aforementioned elements to reduce monotony, exploit learners playful characteristics, and increase their satisfaction. To enhance satisfaction, learners could take courses from Web-based learning academics working with corporations to oVer courses that are practical and helpful to their careers. The key Wnding of this study is that only utility value and satisfaction make signiWcant contributions to learners intention to continue using Web-based learning. The Wnding suggests that satisfaction plays the major role in shaping learners intention to continue using Web-based learning. Utility value explained a signiWcantly greater percentage of variance of learners continuance intention than did attainment value, intrinsic value, and cost. The Wnding suggests that learners in our study put more emphasis on the utility of learning activities than on the importance, interest, and negative aspects as they evaluate the value of Web-based learning. In other words, learners are goal-oriented and concern extrinsic reasons for engaging in Web-based learning. To increase learners intention to continue using Web-based learning, the providers should oVer courses that could help learners develop their skills and knowledge and degrees that meet learners career goals. As shown in Table 4, attainment value and intrinsic value had positive and indirect eVects on continuance intention through satisfaction. The results indicate a strong mediation of satisfaction in the eVects of attainment value and intrinsic value. It implies that while attainment value and intrinsic value may increase learners satisfaction with Web-based learning, they may not be strong enough to increase learners continuance intention directly. Web-based learning providers should develop a long-term strategy to enhance learners satisfaction and maintain or enhance users loyalties to Web-based learning services. Additionally, distributive fairness was found to be the second factor inXuencing continuance intention in terms of total eVect (direct plus indirect eVect), while utility value was found to be the third factor. The results suggest that learners are more concerned with the fairness of the outcomes than the utility of Web-based learning. The results also suggest that individuals who are fairly rewarded not only experience satisfaction but also will be motivated
Table 4 EVects of fairness and value on continuance intention Construct Distributive fairness Procedural fairness Interactional fairness Attainment value Utility value Intrinsic value Cost Satisfaction Direct eVect 0.00 0.00 0.00 0.01 0.15 0.00 0.01 0.81 Indirect eVect 0.389 (0.480.81) 0.081 (0.100.81) 0.113 (0.140.81) 0.162 (0.200.81) 0.154 (0.190.81) 0.122 (0.150.81) 0.041 (0.050.81) 0.00 Total eVect 0.389 0.081 0.113 0.172 0.304 0.122 0.051 0.810

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1241

to continue using Web-based learning. Accordingly, an interesting area for future research is to examine the direct eVects of the three dimensions of fairness on Web-based learning continuance intention. Finally, the results suggest that failure to include the inXuence of distributive fairness, interactional fairness, attainment value, utility value, and intrinsic value may lead to inappropriate conclusions and limit the explanatory power of learners satisfaction with Web-based learning. Although the Wndings are encouraging and useful, the present study has certain limitations. First, whether our Wndings could be generalized to all types of Web-based learning is unclear. This study focuses on Web-based learning for continuing education. Factors inXuencing satisfaction and continuance intention of part-time students might be diVerent from those of full-time students. Further research is necessary to verify the generalizability of our Wndings. Second, the results may have been impacted by self-selection bias. Our sample comprised only active users of Web-based learning. Individuals who had already ceased to participate in Web-based learning might have diVerent perceptions about the inXuence of components of subjective task value and dimensions of fairness, and so may have been diVerently aVected by them. Therefore, the results should be interpreted as only explaining satisfaction and continuance intention of active users of Web-based learning. Whether the results can be generalized to non-participants or to disaVected users will require additional research. Third, the eVects of fairness and value components on satisfaction and continuance intention may change with increasing user experience over time. For example, attainment value might exert its inXuence on learners continuance intention at the early adoption stage. Its inXuence on continuance intention might decrease as learners have more experience in using Web-based learning (take more Web-based courses). Yet, the learners responses in our research were cross-sectional data and did not present an opportunity to examine the longterm trend of these hypothesized relationships. Further longitudinal studies are recommended to validate our research model in this regard. Finally, the program oVered by NKNU is for any individual who has a need for continuing education Learners participated in the NKNU Web-based learning in a voluntary setting (i.e., under their full volitional control). Some organizations or companies oVered online learning in which individuals participated in the learning programs in a mandatory setting (i.e., based on the organizations regulation or other normative considerations). Our Wndings may not be generalized to the mandatory setting. Further research is necessary to verify the diVerences between the voluntary and mandatory settings. Finally, DeLone and McLeans (2003) IS success model posits that information quality, system quality and service quality are major determinants of users satisfaction with and intention to use information systems. Later studies could explore the eVect of information quality, system quality and service quality on learners satisfaction with and intention to continue using Web-based learning. Theory of planned behavior (TPB) (Ajzen, 2002) theorizes that attitude, subjective norm, perceived controllability, and self-eYcacy are important predictors of behavioral intention. Therefore, future research could examine the inXuences of attitudinal beliefs, subjective norm, perceived controllability, and self-eYcacy on Web-based learning continuance intention.

References
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 267299). New York: Academic Press.

1242

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Ajzen, I. (2002). Perceived behavioral control, self-eYcacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology, 32, 665683. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: a review and recommended two-step approach. Psychological Bulletin, 103(3), 411423. Arbaugh, J. B. (2000). Virtual classroom characteristics and student satisfaction in Internet-based MBA courses. Journal of Management Education, 24(1), 3254. Arbaugh, J. B. (2002). Managing the on-line classroom: a study of technological and behavioral characteristics of webbased MBA courses. The Journal of High Technology Management Research, 13(2), 203223. Atkinson, J. W. (1964). An introduction to motivation. Princeton, NJ: Van Nostrand. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood CliVs, NJ: Prentice Hall. Battle, A., & WigWeld, A. (2003). College womens orientations toward family, career, and graduate school. Journal of Vocational Behavior, 62(1), 5675. Bennett, S., Priest, A. M., & Macpherson, C. (1999). Learning about online learning: an approach to staV development for university teachers. Australian Journal of Educational Technology, 15(3), 207221. Bhattacherjee, A. (2001). Understanding information systems continuance: an expectation-conWrmation model. MIS Quarterly, 25(3), 351370. Bies, R. J., & Moag, J. S. (1986). Interactional justice: communication criteria of fairness. In R. J. Lewicki, B. H. Sheppard, & M. H. Bazerman (Eds.), Research on negotiation in organizations (pp. 4355). Greenwich, CT: JAI Press. Billings, D., Conners, H., & Skiba, D. (2001). Benchmarking best practices in Web-based nursing courses. Advances in Nursing Science, 23, 4152. Blodgett, J. G., Hill, D. J., & Tax, S. S. (1997). The eVects of distributive, procedural, and interactional justice on post complaint behavior. Journal of Retailing, 73(2), 185210. Bojanic, D. C. (1996). Consumer perceptions of price, value and satisfaction in the hotel industry: an exploratory study. Journal of Hospitality and Leisure Marketing, 4(1), 522. Bolton, R. N., & Drew, J. H. (1991). A multistage model of customers assessments of service quality and value. Journal of Consumer Research, 17, 375384. Bong, M. (2001). Role of self-eYcacy and task-value in predicting college students course performance and future enrollment intentions. Contemporary Educational Psychology, 26, 553570. Boudreau, M., Gefen, D., & Straub, D. W. (2001). Validation in information systems research: a state-of-the-art assessment. MIS Quarterly, 25(1), 116. Brush, L. R. (1980). Encouraging girls in mathematics: The problem and the solution. Cambridge, MA: Abt. Books. Carswell, A. D., & Venkatesh, V. (2002). Learner outcomes in an asynchronous distance education environment. International Journal of Human-Computer Studies, 56, 475494. Chiu, C. M., Hsu, M. H., Sun, S. Y., Lin, T. C., & Sun, P. C. (2005). Usability, quality, value and e-learning continuance decisions. Computers & Education, 45, 399416. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319340. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22(14), 11111132. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York, NY: Plenum Press. DeLone, W. H., & McLean, E. R. (2003). The DeLone and McLean model of information systems success: a ten-year update. Journal of Management Information Systems, 19(4), 930. Doll, W. J., Xia, W., & Torkzadeh, G. A. (1994). A conWrmatory factor analysis of the end-user computing satisfaction instrument. MIS Quarterly, 18, 453461. Dutton, J., Dutton, M., & Perry, J. (2002). How do online students diVer from lecture students? Journal of Asynchronous Learning Networks, 1(6), 120. Eccles, J. S. (1987). Gender roles and womens achievement-related decisions. Psychology of Women Quarterly, 11(2), 135172.

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1243

Eccles, J. S., Adler, T. F., Futterman, R., GoV, S. B., Kaczala, C. M., Meece, J. L., et al. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and Achievement Motivation (pp. 75146). San Francisco, CA: W.H. Freeman. Eccles, J. S., Adler, T., & Meece, J. L. (1984). Sex diVerences in achievement: a test of alternate theories. Journal of Personality and Social Psychology, 46(1), 2643. Eccles, J. S., & WigWeld, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53(1), 109132. Fennema, E., & Sherman, J. (1977). Sex-related diVerences in mathematics achievement, spatial visualisation and aVective factors. American Educational Research Journal, 14(1), 5171. Folger, R., & Greenberg, J. (1985). Procedural justice: an interpretative analysis of personnel systems. In K. Rowland & G. Ferris (Eds.), Research in Personnel and Human Resources Management (pp. 141183). Greenwich, CT: JAI Press. Folger, R., & Konovsky, M. A. (1989). EVects of procedural and distributive justice on reactions to pay raise decisions. Academy of Management Journal, 32(1), 115130. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable and measurement error. Journal of Marketing Research, 18(1), 3950. Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: an integrated model. MIS Quarterly, 27(1), 5190. Gong, M., Xu, Y., & Yu, Y. (2004). An enhanced technology acceptance model for Web-Based learning. Journal of Information Systems Education, 15(4), 365373. Gwinner, K. P., Gremier, D. D., & Bitner, M. J. (1998). Relational beneWts in services industries: the customers perspective. Journal of the Academy of Marketing Science, 26, 101114. Hart, C. W., Heskett, J. L., & Sasser, W. E., Jr. (1990). The proWtable art of service recovery. Harvard Business Review, 68(4), 148156. Homans, G. G. (1961). Social behavior: Its elementary forms. New York: Harcourt Brace. Homer, P. M., & Kahle, L. R. (1988). A structural equation test of the valueattitudebehavior hierarchy. Journal of Personality and Social Psychology, 54(4), 638646. Hong, K. S., Lai, K. W., & Holton, D. (2003). Students satisfaction and perceived learning with Web-based course. Educational Technology and Society, 6(1), 116124. Hsu, M. H., & Chiu, C. M. (2004). Internet self-eYcacy and electronic service acceptance. Decision Support Systems, 38(3), 369381. Kahl, T. N., & Cropley, A. J. (1986). Face-to-face versus distance learning: psychological consequences and practical implications. Distance Education, 7(1), 3848. Kaplan-Leiserson, E., (2000). e-Learning glossary, Available from http://www.learningcircuits.org/glossary.html. Klaila, D., (2001). Game-based e-learning gets real, Learning Circuits. ASTDs online magazine. Available from http:// www.learningcircuits.org/2001/jan2001/klaila.html. Kumar, N., Scheer, L. K., & Steenkamp, J. E. (1995). The eVects of supplier fairness on vulnerable resellers. Journal of Marketing Research, 32(1), 5465. Lee, J. S., Cho, H., Gay, G., Davidson, B., & IngraVea, A. (2003). Technology acceptance and social networking in distance learning. Educational Technology & Society, 6(2), 5061. Leventhal, G. S. (1980). What should be done with equity theory? In K. J. Gergen, M. S. Greenberg, & R. W. Willis (Eds.), Social exchange: Advances in equity and research (pp. 2755). New York: Plenum. Lim, C. K. (2001). Computer self-eYcacy, academic self-concept, and other predictors of satisfaction and future participation of adult distance learners. American Journal of Distance Education, 15(2), 4151. Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice. New York: Plenum Press. Lorenzetti, J. P. (2005). How e-Learning is changing higher education: A new look. Distance education report. pp. 47. Martins, B. M. J., Steil, A. V., & Todesco, J. L. (2004). Factors inXuencing the adoption of the Internet as a teaching tool at foreign language schools. Computers & Education, 42(4), 353374. Mathieson, K. (1991). Predicting user intentions: comparing the technology acceptance model with the theory of planned behavior. Information Systems Research, 2(3), 173191. Maxham, J. G., & Netemeyer, R. G. (2002). Modeling customer perceptions of complaint handling over time: The eVects of perceived justice on satisfaction and intent. Journal of Retailing, 78(4), 239252. McGorry, S. Y. (2003). Measuring quality in online learning programs. The Internet and Higher Education, 6, 159177.

1244

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

Meece, J. L., WigWeld, A., & Eccles, J. S. (1990). Predictors of math anxiety and its consequences for young adolescents course enrollment intentions and performances in mathematics. Journal of Educational Psychology, 82(1), 6070. Moon, J. W., & Kim, Y. G. (2001). Extending the TAM for a World Wide Web context. Information & Management, 38(4), 217230. Moorman, R. H. (1991). Relationship between organizational justice and organizational citizenship behaviors: do fairness perceptions inXuence employee citizenship? Journal of Applied Psychology, 76(6), 845855. Morris, L., Wu, S., & Finnegan, C. (2005). Predicting retention in online general education courses. The American Journal of Distance Education, 19, 2336. Motiwalla, L., & Tello, S. (2000). Distance learning on the Internet: an exploratory study. The Internet and Higher Education, 2(4), 253264. Neal, W. D. (1999). Satisfaction is nice, but value drives loyalty. Marketing Research, 11, 2123. Neal, L., & Normore, L. (2005). eLearning and fun: A report from the CHI 2005 special interest. eLearn, 2005(7), 2. NiehoV, B. P., & Moorman, R. H. (1993). Justice as a mediator of the relationship between methods of monitoring and organizational citizenship behavior. Academy of Management Journal, 36(3), 527556. Oliver, R. L. (1980). A cognitive model for the antecedents and consequences of satisfaction. Journal of Marketing Research, 17(4), 460469. Oliver, R. L., & Swan, J. E. (1989). Consumer perceptions of interpersonal equity and satisfaction in transactions: a Weld survey approach. Journal of Marketing, 53(2), 2135. Osborn, V. (2000). The distributed learning survey. Denton, TX: University of North Texas. PalloV, R. M., & Pratt, K. (1999). Building learning communities in cyberspace: EVective strategies for the online classroom. San Francisco, CA: Jossey-Bass. Parasuraman, A., & Grewal, D. (2000). The Impact of technology on the qualityvalueloyalty chain: a research agenda. Journal of the Academy of Marketing Science, 28(1), 168174. Patterson, P. G., & Spreng, R. A. (1997). Modeling the relationship between perceived value, satisfaction and repurchase intentions in a business-to-business, services context: an empirical examination. International Journal of Service Industry Management, 8(5), 414434. Perraton, H. (1988). A theory for distance education. In D. Sewart, D. Keegan, & B. Holmberg (Eds.), Distance education: International perspectives (pp. 95113). New York: Routledge. Prensky, M. (2001). Digital game-based learning. New York: McGraw Hill. Ramaswami, S. N., & Singh, J. (2003). Antecedents and consequences of merit pay fairness for industrial salespeople. Journal of Marketing, 67(4), 4666. Reichheld, F., & Schefter, P. (2000). E-loyalty: your secret weapon on the Web. Harvard Business Review, 78(4), 105113. Rokeach, M. J. (1968). Beliefs, attitudes, and values. San Francisco, CA: Jossey Bass. Ryan, R. M., & Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic deWnitions and new directions. Contemporary Educational Psychology, 25(1), 5467. Seiders, K., & Berry, L. L. (1998). Service fairness: What it is and why it matters. The Academy of Management Executive, 12(2), 821. Sirdeshmukh, D., Singh, J., & Sabol, B. (2002). Consumer trust, value, and loyalty in relational exchange. Journal of Marketing, 66, 1537. Smith, A. K., Bolton, R. N., & Wagner, J. (1999). A model of customer satisfaction with service encounters involving failure and recovery. Journal of Marketing Research, 36(3), 356372. Spreng, R. A., Dixon, A. L., & Olshavsky, R. W. (1993). The impact of perceived value on consumer satisfaction. Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior, 6, 5055. Stoel, L., & Lee, K. H. (2003). Modeling the eVect of experience on student acceptance of Web-based courseware. Internet Research, 13(5), 364374. Sullins, E. S., Hernandez, D., Fuller, C., & Tashiro, J. (1995). Predicting who will major in a science discipline: expectancy-value theory as part of an ecological model for studying academic communities. Journal of Research in Science Teaching, 32, 99119. Swan, J. E., & Trawick, I. F. (1981). DisconWrmation of expectations and satisfaction with a retail service. Journal of Retailing, 57, 4067.

C.-M. Chiu et al. / Computers & Education 49 (2007) 12241245

1245

Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences: implications for relationship marketing. Journal of Marketing, 62(2), 6076. Thibaut, J., & Walker, L. (1975). Procedural justice: A psychological analysis. Hillsdale, NJ: Lawrence Erlbaum. Triandis, H. C. (1980). Values, attitudes, and interpersonal behavior. In H. E. Howe & M. M. Page (Eds.), Nebraska symposium on motivation (pp. 195260). Lincoln, NE: University of Nebraska Press. Wang, A. Y., & Newlin, M. H. (2002). Predictors of web-student performance: the role of self-eYcacy and reasons for taking an on-line class. Computers in Human Behavior, 18, 151163. Wegner, S., Holloway, K., & Garton, E. (1999). The eVects of internet-based instruction on student learning. Journal of Asynchronous Learning Networks, 3(2). Available from http://www.sloan-c.org/publications/jaln/v3n2/v3n2_ wegner.asp. WigWeld, A. (1994). The role of childrens 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 (pp. 101124). Mahwah, NJ: Erlbaum. WigWeld, A., Eccles, J., (1989). Relations of expectancies and values to students math grades and intentions. Paper presented at the meeting of the American educational research association San Francisco, CA. WigWeld, A., & Eccles, J. (1992). The development of achievement task values: a theoretical analysis. Developmental Review, 12, 265310. Willis, B., (1993). Strategies for teaching at a distance. ERIC Document Reproduction Service, No. ED 35 1008. Yilmaz, C., Sezen, B., & Kabadayi, E. T. (2004). Supplier fairness as a mediating factor in the supplier performance reseller satisfaction relationship. Journal of Business Research, 57(8), 854863. Zeithaml, V. A. (1988). Consumer perceptions of price, quality and value: a means-end model and synthesis of evidence. Journal of Marketing, 52(3), 222.