-941 -E-LEARNING SATISFACTION FACTORS Oyku Isik University of North Texas College of Business Administration Department of Information Technology & Decision Sciences Denton, TX 76203 oyku.isik@unt.edu (940) 565 3668 ABSTRACT Success of e-learning systems is based on satisfaction of the user and factors affecting this satisfaction have been researched extensively. Although there are many well-studied constructs, there has been limited attention to computer self-efficacy. To address this gap, we test whether computer self-efficacy, with perceived value, perceived quality and perceived usability affect elearning satisfaction. Except perceived quality, results provide significant for the hypotheses. Keywords: e-learning, online learning, student satisfaction INTRODUCTION Web technology has been favored by educators since it made web-based instruction viable [18]. Measuring e-learning student satisfaction has been a ‘hot topic’ for the academia. At the AMCIS conference as early as 2001, e-learning was identified as one of the nine meta-tracks for information systems (IS) discipline, and multiple studies in both education and the IS literature measure student satisfaction with the online courses [18]. Research shows that perceived usability, perceived value, perceived quality are critical factors that affect user satisfaction for elearning systems [4] [17]. While computer self-efficacy is another factor that influences user’s satisfaction, few IS studies investigate this relationship [5]. Research examines computer selfefficacy as a moderating variable affecting the usability of the system [11]. However, there are insufficient studies investigating whether computer self-efficacy is a direct predictor of online course satisfaction. This study contributes to online education literature by building upon Chiu et al.’s [4] e-learning continuation model and attempts to fill the gap of deficient research on computer self-efficacy. Clearly, understanding the factors influencing user’s satisfaction with online courses is a critical issue for researchers and practitioners alike. Given the role of information and system design in online customer satisfaction, research has synthesized the IS research on user satisfaction with marketing research on customer satisfaction to gain insight on web-based system satisfaction [15]. Similarly, this study draws from both marketing and IS research to examine the factors that contribute to web-based learning systems. The following section includes a literature review of the constructs focused in this study. -942 -LITERATURE REVIEW Chiu et al. proposes an extended expectancy disconfirmation model (EDT) to examine what influences a user’s intention to continue using an e-learning service. Their objective is to measure the continuance decision, which is mainly affected by user satisfaction. The model suggests that perceived usability, perceived quality, perceived value and disconfirmations (usability disconfirmation, quality disconfirmation and value disconfirmation) together influence user’s satisfaction, which then influences user’s e-learning continuance intention. The results yield significance for all proposed hypotheses. In our study, we employ three constructs from Chiu et al.’s model; (perceived usability, perceived quality and perceived value) to measure user satisfaction. We also incorporate computer self-efficacy as an additional construct to measure elearning service satisfaction. Ours differs from Chiu et al.’s model in that our aim is to measure only e-learning user satisfaction. EDT assumes that individuals have some kind of benchmark against which to judge the success of various circumstances [4]. This assumption does not apply to e-learning systems; thus, we do not consider disconfirmation constructs in our model. The following section reviews the literature on the constructs of our model and presents hypotheses. Perceived Usability Perceived usability is the extent to which a product or service is used by a group of users in a specific context with respect to the product’s effectiveness and efficiency and user’s satisfaction [4]. Three sub-constructs (usefulness, ease of use, and compatibility) comprise perceived usability [4]. Perceived usefulness is “the degree to which a person believes using a particular system would enhance his or her job performance,” whereas perceived ease of use is “the degree to which a person believes using a particular system would be free of effort” [6, p.320]. The compatibility factor is the degree to which the system in use fits user’s current needs, values and past experiences [19]. Research shows that perceived usefulness improves user’s satisfaction with an IS [17]. Research also indicates that perceived usefulness, perceived ease of use and compatibility improves user satisfaction [4]. Therefore, the following hypothesis is proposed: H1: Perceived usability is positively related to user’s satisfaction with the online course. Perceived Value Perceived value is the “consumer’s overall assessment of the utility as a product based on perceptions of what is received and what is given” [22, p.14]. Customer satisfaction is directly related to perceived value; when perceived value increases, customer satisfaction also increases [2]. Using perceived value as a measure helps e-learning system designers to be more aware of the customer values and working toward meeting the customer goals increases the chance that customers are satisfied with the e-learning application [16]. Thus, it is important to include perceived value in models designed to measure user satisfaction. Therefore, the second hypothesis in this study is; -943 -H2: Perceived value of the e-learning tool is positively related to user satisfaction with the online course. Perceived Quality There are three sub-constructs comprising perceived quality; system quality, information quality and service quality [4]. System quality pertains to how productively information is processed within the system [7]. System quality together with information quality affects user satisfaction positively [7] [12]. Information quality refers to the relevance, timeliness and accuracy of information generated by an IS [17], and is often measured as a part of user satisfaction [7]. As information quality increases within an IS, user satisfaction also increases [12]. Research shows that there is a positive relationship between both information and system quality and web customer satisfaction [15]. Service quality refers to providing high quality service to the customer [4]. It is applicable to an e-learning system since e-learning can be considered a service providing information to students (who may be considered as the web customers). Service quality of IT-based services positively affects customer satisfaction [14]. We assume that these three dimensions constituting the perceived system quality variable positively affect user satisfaction in an online environment. As such: H3: Perceived quality of the e-learning tool is positively related to user satisfaction with the online course. Computer Self-Efficacy Self-efficacy is as a person’s self evaluation of his own capabilities regarding a specific course of action [1]. This study addresses computer self-efficacy, which can be defined as a person’s self evaluation regarding the accomplishment a task using a computer [5]. Research examines selfefficacy as both a direct [3] and as a moderating variable affecting perceived ease of use and perceived usefulness [11]. In this study, computer self-efficacy is considered a construct that directly impacts satisfaction, because we think that computer self-efficacy is necessary to understand users’ e-learning behavior. If users are not confident about successfully using a technology, then their beliefs about results may not be sufficient to manipulate their behavior [5]. Research indicates that self-efficacy and performance are positively related [8], and also that product performance and customer satisfaction are positively correlated [9]. Unfortunately, little attention is paid to the relationship between computer self-efficacy and IT usage [5]. One of this study’s primary objectives is to show that computer self-efficacy affects user satisfaction with online education systems. Research shows that computer self-efficacy is positively related to user satisfaction with ERP systems [13]. The relationship between ERP systems and e-learning tools is clear since both may be considered information technology artifacts. Thus, we suggest that a person with high computer self-efficacy uses e-learning tools more proficiently and is more satisfied when compared to users with relatively low computer self-efficacy. Thus: H4: Computer self-efficacy is positively related to user’s satisfaction with the online course. -944 -Figure 1 depicts the research model tested in this study. FIGURE 1. RESEARCH MODEL METHODOLOGY The population of interest includes college students who are enrolled in one or more 100% online courses. A convenience sampling methodology was employed; the survey was administered to online classes offered at a Texas university. There were 206 responses in total. Data was collected using an online survey. The questions for three constructs (perceived value, perceived usefulness and perceived quality) were adopted from Chui et al.’s e-learning continuance intention model. Computer self-efficacy was measured using the scale developed by Compeau and Higgins [5]. General satisfaction measure was adopted from an online course evaluation form provided by University of Arizona. All scales were in seven-point Likert format. Demographic information was also collected from students. Data collection lasted for 28 days. Content and face validity of the survey was addressed through a panel of experts formed by 15 PhD students from the College of Business Administration. Visual modifications were done to the survey based on their feedback. Construct validity was addressed through Principal Component Factor Analysis, employing the Varimax Rotation Method. Various item purifications were done and factor loadings below 0.4 were suppressed [10]. Response bias was checked via independent samples t-test and age was found to be the only significant factor. While the average age of students that answered the survey within the first 10 days is 26.1, people who answered on the last 10 days had an average age of 21.2. We believe this is not a threat for our findings because data collection took place in a short period of time. Reliability analysis was done using Cronbach’s alpha, and all of the values were above 0.9. The respondent pool consisted of 48.5 % male and 51.5% female. 68.4 % of the subjects were aged between 18 and 24 while 22.8 % of them were between 25 and 31. 49 % of them have been expecting a B in the online course that their answers referred to. A linear regression analysis was run to see if all four constructs are significantly related to online course satisfaction. To check whether any specific demographics data make a significant difference on course satisfaction, one-way ANOVA was run between dependent variable and demographics. RESULTS AND DISCUSSION Perceived Usability Perceived Quality Perceived Value Computer Self-efficacy Online Satisfaction H1 H2 H4 H3 -945 -A linear regression was run with online course satisfaction as the dependent variable and perceived usability, perceived quality, perceived value and computer self-efficacy as the independent variables. Results indicated that perceived usability, perceived value and computer self-efficacy are significant predictors of online course satisfaction, but perceived quality was not significant. This result was surprising to us since Chui et al.’s (2005) findings regarding perceived quality were significant. We believe the insignificance can be a result of our scale. There were only 6 items measuring this construct, which had 3 dimensions (system quality, information quality, service quality). More items can help better identify the effects of perceived quality. Another reason for the insignificance can be the attitudes of students; most of our participants are distant learning students, thus they did not have the chance of choosing between a traditional class and the online class. They may think that the online classes are mandatory, thus have not developed an idea of the quality of the e-learning system they use. To examine the relationship of the dependent variable with demographics, ANOVA was used. Only the expected course grade was significant. So we did a post-hoc analysis with assuming Tukey’s equal variances to closely examine the relationship; the results revealed that the students whose expected grade is high (an A or B) are more satisfied with the online course than the students whose expected grade is low (C, D or E). This finding indicates that, the e-learning tool satisfaction cannot be the only determinant for the course satisfaction; students’ performance with the course content also influences their satisfaction with the course. CONCLUSION AND FUTURE IMPLICATIONS The goal of this study is to test whether perceived value, perceived quality, perceived usability and computer self-efficacy affect a student’s online course satisfaction. Except the perceived quality construct, the results provide strong support for the hypotheses suggesting perceived value, perceived usability and computer self-efficacy affect students’ online course satisfaction. Partly consistent with our base model, our study found that computer self-efficacy is a strong predictor of online course satisfaction, rather than being a moderator affecting usability. Findings imply that the effectiveness of e-learning is related to how confident students are while using the computer and the web-based learning software. This indicates that developers need to consider the self-efficacy issues (such as ease of use, user friendliness) while developing e-learning systems. Insight into this concept could help universities offering online courses improve their student satisfaction, thus enhance the effectiveness. Our study has limitations that necessitate future research. The subjects of this study were from only one university, this is a threat to the external validity. This study may yield better results if a more diverse group of respondents can be reached. Our study also found that perceived quality of the e-learning system is not a significant predictor for the online course satisfaction, contradicting with previous findings; a more detailed study on dimensions of perceived quality and better scale development for each dimension is necessary. Overall, this paper had shown that perceived value, perceived usability and computer self-efficacy plays an important role in determining online satisfaction of students who take 100% online courses. -946 -REFERENCES [1]Bandura, A. (1986). “Social Foundations of Thought and Action”, New Jersey: Prentice Hall. [2]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, pp.5–22. [3]Chau, P. (2001). “Influence of Computer Attitude and Self-Efficacy on IT Usage Behavior,” Journal of End User Computing, 13:1, pp.26-33. [4]Chiu, C., M. Hsu, S. Sun, T. Lin, P. Sun, (2005), “Usability, quality, value and e-learning continuance decisions,” Computers & Education, 45, pp.399–416. [5]Compeau D. and C. Higgins, (1995), “Computer Self-Efficacy: Development of a Measure nd Initial Test,” MIS Quarterly, 19:2, pp.189-211. [6]Davis, F. (1989). “Perceived usefulness, perceived ease of use, and user acceptance of information technology,” MIS Quarterly, 13:3, pp.319–340. [7]DeLone, W.H., and McLean, E.R. (1992). “Information systems success: The quest for the dependent variable,” Information Systems Research, 3:1, pp.60–95. [8]Fagan M.H., S. Neill and B.R. Wooldridge (2004), “An Empirical Investigation into the Relationship between Computer Self-Efficacy, Anxiety, Experience, Support and Usage,” Journal of Computer Information Systems, 44:2, pp. 95-106. [9]Grace, D. and A. O'Cass (2005). “Examining the effects of service brand communications on brand evaluation,” The Journal of Product and Brand Management, 14:2, pp.106-117. [10]Hair, J., R. Anderson, R. Tatham and W. Black (1998), Multivariate Data Analysis 5th Edition, India: Pearson Education. [11]Hayashi, A.,Chen, C., Ryan, T. and Wu, J. (2004) “The Role of Social Presence and Moderating Role of Computer Self Efficacy in Predicting the Continuance Usage of E-Learning Systems,” Journal of Information Systems Education, 15:2; pp. 139-155. [12]Iivari, J. (2005) “An Empirical Test of the DeLone-McLean Model of Information System Success,” Database for Advances in Information Systems, 36:2, pp. 8-27. [13] Kelley, H. (2001) “Attributional Analysis of Computer Self-efficacy,” PhD Dissertation, The University of Western Ontario. [14]Lai, T.L. (2004), “Service Quality and Perceived Value’s Impact on Satisfaction, Intention and Usage of Short Message Service (SMS),” Information Systems Frontiers 6:4, pp.353–368. [15]McKinney, V., Yoon, K., & Zahedi, F.M. (2002). “The measurement of Web-customer satisfaction: An expectation and disconfirmation approach,” Information Systems Research, 13:3, pp.296–315. [16]Roffe, I. (2004) “E-learning for SMEs: Competition and Dimensions of Perceived Value,” Journal of European Industrial Training, 28:5, pp.440-455. [17]Seddon, P.B. (1997). “A respecification and extension of the DeLone and McLean model of IS success,” Information Systems Research, 8(3), 240–253. [18]Summers, J., Waigandt, A., and Whittaker, T., (2005), “A Comparison of Student Achievement and Satisfaction in an Online Versus a Traditional Face-to-Face Statistics Class,” Innovative Higher Education, 29:3, pp. 233-250. [19]Taylor, S., & Todd, P.A. (1995). “Understanding information technology usage: A test of competing models,” Information Systems Research, 6:2, pp.144–176.