Investigating users perspectives on e-learning-An integration of TAM and IS success model

Computers in Human Behavior 45 (2015) 359–374 Contents lists available at ScienceDirect Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh Investigating users’ perspectives on e-learning: An integration of TAM and IS success model Hossein Mohammadi ⇑ Department of Public Administration, Allameh Tabataba’i University, Tehran, Iran a r t i c l e i n f o Keywords: E-learning Quality Satisfaction Intention to use Actual use a b s t r a c t The purpose of this paper is to examine an integrated model of TAM and D&M to explore the effects of quality features, perceived ease of use, perceived usefulness on users’ intentions and satisfaction, alongside the mediating effect of usability towards use of e-learning in Iran Based on the e-learning user data collected through a survey, structural equations modeling (SEM) and path analysis were employed to test the research model The results revealed that ‘‘intention’’ and ‘‘user satisfaction’’ both had positive effects on actual use of e-learning ‘‘System quality’’ and ‘‘information quality’’ were found to be the primary factors driving users’ intentions and satisfaction towards use of e-learning At last, ‘‘perceived usefulness’’ mediated the relationship between ease of use and users’ intentions The sample consisted of e-learning users of four public universities in Iran Past studies have seldom examined an integrated model in the context of e-learning in developing countries Moreover, this paper tries to provide a literature review of recent published studies in the field of e-learning Ó 2014 Elsevier Ltd All rights reserved Introduction To meet educational purposes and students’ demands, e-learning development emerges to be a catalyst for today educational institutions (Alsabswy, Cater-Steel, & Soar, 2013; Docimini & Palumbo, 2013) E-learning can be defined as a dynamic and immediate learning environment through the use of internet to improve the quality of learning by providing students with access to resources and services, together with distant exchange and collaboration (Docimini & Palumbo, 2013; Jeong & Hong, 2013) E-learning supports learners with some special capabilities such as interactivity, strong search, immediacy, physical mobility and situating of educational activities, self-organized and self-directed learning, corporate training, personalized learning, and effective technique of delivering lesson and gaining knowledge (Bidin & Ziden, 2013; Docimini & Palumbo, 2013; Jeong & Hong, 2013; Martin & Ertzberger, 2013; Viberg & Gronlung, 2013) E-learning has a positive impact on both teachers and students in that it positively affects the duration of their attention, learning and training tenacity, and their attitudes towards collaboration and interaction (Chen & Tseng, 2012; Ozdamli & Uzunboylu, 2014) Past studies have indicated that anywhere and anytime learning and access to ⇑ Address: Pars Pamchal Alley, block 17, No 2, Naghshe Iran St Ansar Alhossein St Second Square, Kosar, Qazvin, Iran Tel.: +98 9192864512, +98 9214563704 E-mail addresses: H.mohammadi901@st.atu.ac.ir, Hossein662@gmail.com http://dx.doi.org/10.1016/j.chb.2014.07.044 0747-5632/Ó 2014 Elsevier Ltd All rights reserved information and communication are facilitated through using e-learning (Chen & Tseng, 2012; Ho & Dzeng, 2010; Islam, 2013; Pena-Ayala, Sossa, & Mendez, 2014) Kratochvíl (2013) and Abachi and Muhammad (2013) note that all individuals involved in e-learning are fond of using it towards learning because of flexible access in terms of time, space, and pace and online collaborative learning However, demand for the development of e-learning is increasingly growing; still the need for research on potential factors affecting e-learning adoption like quality which is the heart of education and training in all countries (Ehlers & Hilera, 2012), is felt especially in the context of developing countries (Masoumi & Lindstrom, 2012), a fact that warrants investigation into it Past studies have used information technology adoption theories such as Technology Acceptance Model (TAM), Innovation Diffusion Theory (IDT) and the Unified Theory of Acceptance and Use of Technology (UTAUT) and the DeLone & McLean’s model to explore e-learning users’ behavioral patterns Some of these studies have taken the barriers and the drivers of e-learning adoption into consideration (e.g., Chen & Tseng, 2012; Islam, 2012, 2013, 2014; Sumak, Hericko, & Punik, 2011) In this paper it is attempted to introduce an integrated model of TAM and DeLone & McLean’s model for predicting individual’s actual use of e-learning system in Iran As Li, Duan, Fu, and Alford (2012) note, it is essential to examine the relationship between e-learners’ experiences, perceptions, and their behavioral intentions to use, because system use is an important indicator of the system’s success 360 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Hassanzadeh, Kanaani, and Elahi (2012), in their attempts to assess e-learning systems success in Iranian universities, identified technical system quality, educational system quality, content and information quality, service quality, user satisfaction, and intention to use, influential towards use of system, system loyalty, and goal achievement Motaghian, Hassanzadeh, and Karimzadegan Moghadam (2013), in their attempts to assess the influence of IS-oriented, psychological and behavioral factors on instructors’ adoption of web-based learning systems in Iran, identified that perceived usefulness, perceived ease of use, and system quality improve instructors’ intentions to use web-based learning systems However, only a limited number of published works have applied an integrated model of IS success model and TAM to explore e-learning usage drivers in the context of developing countries This research, compared to Hassanzadeh et al (2012), tries to step forward to investigate the students’ perceptions of e-learning services based on an integrated model of TAM and IS success model and provides a literature review of recent published works in the context of e-learning which appear to be the main contributions of the paper This paper is focused on Iran as a developing country in the Middle East, which possesses a large population of over 75 million individuals, 37 million of which according to Internetworldstats.com (2012) are internet users, ranking Iran first in the Middle East and fourth in Asia This study attempts to fill a research gap by addressing the effects of quality features of e-learning systems including educational quality, service quality, technical system quality, and content and information quality, accompanied with perceived ease of use and perceived usefulness on students’ satisfactions and intentions towards use of e-learning, besides investigating mediating effect of perceived ease of use on intention through perceived usefulness The remainder of the paper is structured as follows: we address literature review in the next section This is followed by the presentation of the research hypotheses, discussion of findings, conclusions, and finally recommendations for future studies Literature review Owing to complicated, interrelated, and multi-faceted nature of IS success, early attempts fell short in defining information system success To address this problem, a success model was presented by DeLone and McLean (1992) which was later modified to compensate for changing in IS over time IS success model (DeLone & McLean, 2003) identified six components of IS success as follows: system quality, information quality, and service quality, intention to use/use, user satisfaction, and net benefits In IS success model, system use precedes user satisfaction and positive experience with use contributes to the enhancement of satisfaction which sequentially leads to a higher intention to use (Petter, DeLone, & McLean, 2008) The revised IS success model, as one of the most widely used model for IS success, has so far been frequently adopted to examine e-learning system success The Technology Acceptance Model proposed by Davis and Bagozzi (Bagozzi, Davis, & Warshaw, 1992) appears to be the most widely used innovation adoption model This model has been used in a variety of studies to explore the factors affecting individual’s use of new technology (Venkatesh & Davis, 2000) Davis (1989) suggests that the sequential relationship of belief–attitude–intention–behavior in TAM, enables us to predict the use of new technologies by users In fact, TAM is an adaptation of TRA in regard to IS which notes that perceived usefulness and perceived ease of use determine an individual’s attitudes towards their intention to use an innovation with the intention serving as a mediator to the actual use of the system Perceived usefulness is also considered to be affected directly by perceived ease of use Cheng (2012) in his study to examine whether quality factors can affect learners’ intention to use e-learning system, incorporated instructor quality to other components of IS success model and concluded that information, service, system, and instructor quality play the antecedent role and come to be as the key drivers of employees’ perceptions with regard to e-learning acceptance Saba (2013), who carried out a study on implications of e-learning systems and self-efficacy on students’ outcomes, concluded that system quality, information quality, and computer self-efficacy all affected system use, user satisfaction, and self-managed learning behaviors of student Kim, Trimi, Park, and Rhee (2012) on their study on investigating the impact of quality on the outcomes of elearning based on IS success model, found that system quality, information quality, and instructional quality positively influence user satisfaction Li et al (2012) identified that e-learning service quality, course quality, perceived usefulness, perceived ease of use, and self-efficacy directly affect, system functionality and system response indirectly affect, while system interactivity insignificantly affects on users’ intentions towards use Chang (2013) showed that web quality significantly and positively influences user value and user satisfaction; furthermore, he concluded that perceived value and satisfaction play the antecedent role in user’s intention towards use of e-learning Wang and Chiu (2011) who incorporated communication quality, information quality, and service quality in his model showed that all had significant positive effects on user satisfaction and loyalty intention to use the e-learning system for interacting experience, collaborating with others, and getting feedback Owing to the rarity of research in examining the students’ learning satisfaction with system quality of a system, Tajuddin, Baharudin, and Hoon (2013) carried out a study to examine the relationship between learning satisfaction and system quality which revealed a positive relationship According to Tseng, Lin, and Chen (2011), the most significant determinants of e-learning effectiveness were the quality of the e-learning system and learner attractiveness In his study, increased usage of multimedia features was figured out to attract learner’s attention and eventually improve his attractiveness and reduction in the response time and waiting time for materials to load was found to improve the quality of the system; accompanied with the responsiveness of instructors to learners’ questions which need to be maintained and improved Islam (2012) who included perceived system quality in the his expectation–confirmation based IS model revealed that perceived usefulness, confirmation of initial expectation, and system quality significantly influenced students’ satisfaction, sequentially satisfaction in addition to perceived usefulness significantly determined continuance intention towards e-learning usage Udo, Bagchi, and Kirs (2011) indicated an instrument for assessing e-learning quality comprises five components including assurance, empathy, responsiveness, reliability, and website content that four of which (except reliability) are valid and reliable constructs to measure e-learning quality and influence learners’ satisfactions and intentions to attend in online courses 2.1 Other related theories and studies On the other hand, there are other related theories that deserve to be mentioned These are theories such as Theory of Planned Behavior (TPB) which discusses that adoption behavior is preceded by behavioral intention which in itself is a function of the individual’s attitude, their beliefs about the extent to which they can control a particular behavior and other external factors; Social Cognitive Theory (SCT) is a framework for understanding, predicting, and changing behavior which introduces human behavior as a result of the interaction between personal factors, behavior, and the environment; Diffusion of Innovation Theory (IDT) which considers adoption of IS as a social construct that gradually develops H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 through the population over time; the Decomposed Theory of Planned Behavior (DTPB), an extended version of TAM, which models perceived ease of use and perceived usefulness as mediators of behavioral intention in which compatibility serves as an antecedent for both of them, and the Unified Theory of User Acceptance of Technology (UTAUT) which notes that four key constructs (performance expectancy, effort expectancy, social influence, and facilitating conditions) are the main determinants of consumers’ usage intention and behavior (Hanafizadeh, Byron, & Khedmatgozar, 2014) The empirical study conducted by Alsabswy et al (2013) confirmed that the role of IT infrastructure services is vital in e-learning service success through positively influencing perceived usefulness, user satisfaction, customer value, and organizational value According to Ossiannilsson (2012), technology and digital study accelerates the change of academic learning, but more emphasis should be put on cultural and structural changes without which technology fall short in changing it; in fact, technology should play the supporting role in this regard Sloan, Porter, Robins, and McCourt (2014) in his study on in elearning challenges on how to support international postgraduate students identified that providing student with support and feedback in understanding the content is of immense importance Sawang, Newton, and Jamieson (2013) figured out that high levels of support can compensate for low technological efficacy in elearning adoption and that e-learning characteristics mediate the relationship between learner characteristics and intention towards its further adoption Lambropoulos, Faulkner, and Culwin (2012) found that e-teacher play a critical role to facilitate and support active participation and engagement towards collaborative learning In accordance with Lara, Lizcano, Martinez, Pazos, and Riera (2014) study, temporal and spatial gap between the teacher and student appears to be an impediment to student follow-up by teacher, and student supervision is essential for the distinction of student behaviors that bring about course dropout Steet and Goh (2012) who conducted a study on exploring the acceptance of an e-reader device as a collaborative learning system, identified that provision of five major determinants including mobility, support, connectivity, immediacy, collaborative support significantly affect users’ acceptance of proposed system; while sustainability affordance, was found to have limited influence on the acceptance of proposed system Jordan (2013) identified that students’ effectiveness mostly relies on the teachers’ way of designing and orienting the online learning experience In other words, the succession of an online learning environment emanated from a strong pedagogical method that put emphasis on a constructivist approach in practice, without which the technology tool fall short to guarantee the succession online learning by its own Leeds (2014) in their study on how technology changes temporal culture in e-learning, found that starting to study online for the first time, e-learners may experience temporal culture shock which needs to be addressed in elearning program Therefore, their time preferences need to be included in designing an e-learning environment to make certain that it’s equipped with enough temporal flexibility, and it should be explicit so that learner expectation can be managed Chang (2013) figured out that perceived value determines users’ intentions towards use of system He also added that perceived support had a significant effect on perceived usefulness of the e-learning system Islam (2013) identified that there exist to be three main constructs significantly affect students’ perceptions including perceived learning assistant, perceived community building assistant, and perceived academic performance which are influenced by perceived usefulness and perceived ease of use and how an e-learning system is used Xu, Huang, Wang, and Heales (2014), on the other hand, concluded that personalized e-learning facilities improve 361 online learning effectiveness in terms of examination, satisfaction, and self-efficacy criteria Learning in future will be reoriented along concepts of collaboration and networking (Ossiannilsson & Landgren, 2012) Alverez, Martin, Fernandez-Castro, and Urretavizcaya (2013) found that e-learning provides the possibility of offering hybrid courses which is a blend of face-to face classroom instruction with web-based learning Barker et al (2013) concluded that e-learning, although welcomed by students, needs to be supplementary to face-to-face learning In his study, four main themes were identified, moving with the times, global networking, inequity as a barrier, and transfer of internet learning into practice Corti-Novo, VerelaCandamino, and Ramil-Diaz (2013) identified that combined e-learning and face-to-face learning clearly improves the participation of students, increase their motivation, competencies and so, their performance in terms of qualifications The study carried out by Rolstadas (2013) suggested that training based on a combination of on-campus and web-based learning is an effective approach and this approach is more beneficial than those traditional with only plenary session or virtual content Hajili, Bugshan, Lin, and Featherman (2013) in their study on the impact of Web 2.0 emergence in learning context and its benefits and values in education notes that the future of e-learning is social learning, in which learning online is facilitated due to the prevalence of social media In fact, social relationships and interactions of individuals in the internet and online communities clearly improve their learning competencies and qualities Diamond and Irwin (2013) identified that e-learning facilities were mostly adopted to provide flexible access to information, followed by support for communication and collaboration, and were scarcely used for the development of specific skills, personal identity and confidence Gupna, Chauhan, and Dutta (2013) indicated that e-learning system radically changes the concept of education whether it is full time, part time, or a distant education program In their study, classroom teaching based on e-learning was well welcomed and paid attention by student as a precious experience and student were figured out to be very comfortable with the courses presented through e-learning and this virtual environment believed to strengthen the face to face classroom They further declared that quality of e-learning system influence the quality of teaching in educational sector Manca and Pozzi (2014) introduced a three dimensional model for evaluation of e-learning system based on which students, teachers, and e-learning managers should be involved in the evaluation He declared that platform, resources and approach are the e-learning system components to be assessed, and design, running and evaluation are the phases of the course lifecycle to be analyzed Troussas, Virou, and Alepis (2013) who put collaboration among learners into consideration in a computer assisted learning environment, found that effective collaboration learning results when students appropriately perceive the significance of working actively with others in order to learn and act in ways which improve the educational procedure and emphasize the value of cooperation Zhang, Fang, Wei, and Wang (2012) in their study on investigating the intention of students to continue their participation in e-learning system, found that psychological safety communication environment evokes the intention to continue participation in e-learning Chen and Tseng (2012) found that motivation to use and internet self-efficacy both had significant positive effects while computer anxiety had a significant negative effect on intention towards web-based e-learning Perceived usefulness and motivation to use were ultimately found key reasons for the acceptance of e-learning system in their study Kim, Lee, and Ryu (2013) showed that taking learners’ personality characteristics and its effects on learning preferences into 362 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 consideration empowers us to improve both the primary learning experience and the information derived Fryer, Bovee, and Nakao (2014) identified two primary reasons impeding users from participating and involving in e-learning studies incorporating poor ability belief and low task value Chien’s (2012) study concluded that learner’s computer self-efficacy is a primary reason affecting e-learning self-effectiveness Chien (2012) further discussed that both system factors involving functionality and instructor factors involving self-efficacy contributes to greater elearning effectiveness, alongside the fact that learner’s computer self-efficacy moderates the relationship between system functionality and training effectiveness Poulova and Simonova (2014) found that learners experienced higher inner satisfaction mostly with an instruction method which reflects their preferences Gonzalez-Gonzalez, Gallardo-Gallardo, and Jimenez-Zarco (2014) introduced critical thinking as one of the students’ prerequisite competencies to improve the efficiency and efficacy of their activities Wang (2014) in his study provided evidences indicating that personalized dynamic assessment automatically developed by system for each learner strengthened students learning effectiveness and facilitated their learning achievements and disappeared misconceptions Le (2012) concluded that e-portfolio results in the improvement of educational qualities since teaching and learning focus is transferred from supervisor-centered to student-centered learning and research, as well as from technological control to technological empowerment According to his study, e-portfolio enables students to completely overcome to their own learning and research practices Bhuasiri, Xaymoungkhoun, Zo, Rho, and Ciganek (2012) informed that technology awareness, motivation, and changing learners’ behavior are three major requirements for successful implementation of e-learning Al-Samarraie, Teo, and Abbas (2013) discussed that structured content contributes to the enhancement of e-learning motivation, attention, and interactivity which results in students’ better thinking skills, and influences their meta-cognitive activities and facilitates understanding Yoo, Han, and Huang (2012) declared that intrinsic motivators including effort expectancy, attitudes, and anxiety more rigorously orient intention towards use of e-learning than did the extrinsic motivators including performance expectancy, social influence, and facilitating conditions Castillo-Merino and Serradell-Lopez (2014) informed that motivation is the most important variable influencing performance of online students which is positively influenced by students’ perception of their own efficiency He further declared that their perception about their ability to use digital technologies leads to better achievements Islam (2014) who adopted his theoretical assumptions from Oliver’s expectation–confirmation theory, Herzberg’s two-factor theory and Kano’s satisfaction model concluded that individuals’ satisfactions towards e-learning were mostly resulted from both environmental and job-specific factors while their dissatisfactions were mostly resulted from only environmental factors Moreno, Moreno, and Molina (2013) who studied how satisfaction is generated towards e-learning system, identified that disconfirmation in the case of measuring expectation before using the service, and expectation in the case of measuring expectation after using the service, occurs as the most important in the model He further added that perceived usefulness and effort expectancy positively affect satisfaction Fiorella and Mayer (2014) indicated that those who actually taught the material by explaining the material to others, outperformed those who did not teach; though, this effect was strongest for those who were prepared to teach In his attempt, preparing to teach resulted in short-term learning gains, whereas the act of teaching coupled with preparing to teach was important for long-term learning Hopp (2013) identified that stan- dardized blended-learning system of mobile and e-learning can be a powerful tool to increase e-learning benefits to the maximum In the context of m-learning as a form of e-learning also some investigations are carried out as follows: Viberg and Gronlung (2013) showed that individuals’ attitudes towards m-learning were most positive with personalization, followed by collaboration and authenticity They concluded that Hofsted’s factors failed to explain the differences in students’ attitudes; gender in their study was identified as influential Mahat, Ayub, and Wong (2012) discovered the significant effects of three main factors including personal innovativeness, readiness to use, and self efficacy on students’ intentions towards m-learning adoption Cheon, Lee, Crooks, and Song (2012) informed that TPB appropriately explained college students’ adoption of m-learning, and attitude, subjective norm, and behavioral control positively affected their intention to adopt mobile learning Liu, Li, and Carlsson (2010) indicated that perceived near-term/long-term usefulness and personal innovativeness significantly affect intention towards m-learning adoption, while perceived long-term usefulness had a significant influence on near-term usefulness Hwang and Tsai (2011) who investigated the trends in mobile and ubiquitous learning from 2001 to 2010, found that the number of articles had significantly increased during the past 10 years and researchers had focused on the related fields in recent years Wu et al (2012) who performed a meta-analysis and reviewed the trends of mobile learning studies from 2003 to 2010, identified that most studies in the context of mobile learning had investigated effectiveness, followed by mobile learning system design They further figured out that surveys and experimental approaches were the most preferred research methods, research outcomes were significantly positive in mobile learning studies, mobile phones and PDAs were the most frequently used devices for mobile learning, and mobile learning was mostly adopted by higher education institutions 2.2 E-learning Networked devices are growingly used for educational purposes and have applied a radical change in the scope of education (Ehlers & Hilera, 2012; Hsu, Hwang, & Chang, 2013) E-learning can be defined as making use of technology as a mediating tool for learning through electronic devices which enable users to readily access information and interact with others online (Wu et al., 2012) The learning styles falls into four categories comprises the computer-aided learning, e-learning, remote learning, and on-line learning (Ho & Dzeng, 2010) The former three ones are the learning ways conducted through electronic media, such as CD, auxiliary software, and interactive TV The online learning is conducted through internet or intranet to generate the interaction among learners, course, and teacher E-learning indeed is a form of online learning; therefore, online learning is called elearning at present (Ho & Dzeng, 2010) E-learning seeks to improve the culture of equal participation among students and teachers for them to share their efforts to gain greater success (Shipee & Keengwee, 2014) and better achieve the key educational objective which is the enhancement of learning effectiveness and efficiency Thus, the students’ perceptions of e-learning technologies are of great importance and precede the successful integration of these technologies in education (Ozdamli & Uzunboylu, 2014) Therefore, exploring the learners’ perceptions concerning e-learning are of immense importance to researchers, because it helps educational institutions such as schools, colleges and universities, and even organizations to get a real advantage by enabling enhanced understanding of key factors that affect intention to use e-learning H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Research model and Hypotheses In this section, the research variables and hypotheses are presented 363 Technical system quality, thus, is assumed to have a positive effect on both individuals’ satisfaction and their intentions towards use of system H5 Technical system quality positively affects user satisfaction 3.1 Educational quality Educational quality, as a new component to IS success model incorporated by Hassanzadeh et al (2012), is seen as system quality in terms of characteristics and features it can render to facilitate users learning and training (Hassanzadeh et al., 2012) Educational quality can be defined as the extent to which an IS system managed to provide a conductive learning environment for learners in terms of collaborative learning (Hassanzadeh et al., 2012; Kim et al., 2012) As Hassanzadeh et al (2012) concluded in their study, educational quality positively affects individuals’ satisfactions which is also confirmed by Kim et al (2012) who found that instructional quality have a significant positive effect on user satisfaction Educational quality, therefore, is assumed to have a positive effect on individuals’ satisfaction; however, it is assumed to have a positive effect on intention to use as well H1 Educational quality positively affects user satisfaction H2 Educational quality positively affects intention to use H6 Technical system quality positively affects intention to use 3.4 Content and information quality The success dimension content and information quality represents the desirable characteristics of an IS’s output (Petter & McLean, 2009) An example would be the information the system and student can generate using the e-learning system Thus, it includes measures focusing on the quality of the information that the system generates and its usefulness for the user Information quality is often seen as a key antecedent for user satisfaction (Hassanzadeh et al., 2012; Kim et al., 2012; Roca et al., 2006; Saba, 2013; Wang & Chiu, 2011), and for intention to use e-learning system (Cheng, 2012; Ramayah et al., 2010; Wang & Chiu, 2011) In this study, therefore, content and information quality is assumed to have a positive impact on both individuals’ satisfaction and their intentions to use H7 Content and information quality positively affects user satisfaction 3.2 Service quality Service quality constitutes the quality of the support that users receive from the IS (Wang & Wang, 2009) such as training (Petter & McLean, 2009) and helpdesk The inclusion of this success dimension is not undoubted, since it normally seen as subordinate to system quality in the model, but some researchers claim that it could stand as an independent variable owing to the great change in IS role in recent years (Wang & Liao, 2008) Service quality has been found to have a significant positive effect on satisfaction in e-learning context (Poulova and Simonova, 2014; Roca, Chiu, & Martinez, 2006; Tajuddin et al., 2013; Wang & Chiu, 2011; Xu et al., 2014), and on intention to use e-learning system in some studies (Cheng, 2012; Hassanzadeh et al., 2012; Li et al., 2012; Ramayah, Ahmad, & Lo, 2010; Wang & Chiu, 2011) In this study, service quality is assumed to have a positive impact on both individuals’ satisfaction and their intentions to use H8 Content and information quality positively affects intention to use 3.5 Perceived ease of use H4 Service quality positively affects intention to use Perceived ease of use is defined as the degree to which a person believes that using a particular system would be free of effort (Davis, 1989), which is an imminent acceptance driver of new technology-based applications (Venkatesh, 2000) The effect of perceived ease of use on intention towards use of e-learning is revealed in some past studies (e.g., Chen and Tseng, 2012; Chow, Herold, Choo, & Chan, 2012; Islam, 2013; Li et al., 2012; Liu et al., 2010; Sumak et al., 2011) As a result, the greater the perceived ease of use of e-learning system, the more positive is the intention towards its usage; thus greater the likelihood that it will be used Moreover, perceived ease of use is assumed to have an indirect effect on intention to use through perceived usefulness in e-learning context as well (Chen and Tseng, 2012) Therefore, perceived ease of use is further expected to have an indirect effect on users’ intentions via perceived usefulness 3.3 Technical system quality H9 Perceived ease of use positively affects intention to use H3 Service quality positively affects user satisfaction In IS success model proposed by DeLone and McLean (2003), technical system quality refers to technical success and the accuracy and efficiency of the communication system that produces information; in fact, it constitutes the desirable characteristics and measures of an IS and relates to the presence and absence of a bug in system (Rabaa’i, 2009) Technical system quality has been found to have a significant positive effect on satisfaction in e-learning context (Alsabawy et al., 2013; Hassanzadeh et al., 2012; Islam, 2012; Kim et al., 2012; Motaghian et al., 2013; Rai, Acton, Golden, & Conboy, 2009; Saba, 2013; Tajuddin et al., 2013; Wang & Chiu, 2011; Wu, Hsia, Liao, & Tennyson, 2008), and on intention to use e-learning system (Cheng, 2012; Islam, 2012; Li et al., 2012; Ramayah et al., 2010; Wang & Chiu, 2011) H10 Perceived ease of use positively affects perceived usefulness 3.6 Perceived usefulness Perceived usefulness is a key determinant of intention, which encourages 21st century IS users to adopt more innovative and user-friendly technologies that give them greater freedom (Pikkarainen, Pikkarainen, and Karjaluoto, 2004) In fact, an individual’s willingness to use a specific IS for their activities depends on their perception of its use (Hanafizadeh, Behboudi, Khoshksaray, & Shirkhani Tabar, 2014) Perceived usefulness has been found to have a significant positive effect on usage intention 364 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Table Definitions of dimensions Construct Definition Source Educational quality A conductive learning environment in terms of collaborative learning Service quality Technical system quality Information quality Perceived ease usefulness Perceived ease of use Satisfaction Intention to use The quality of the support that users receive from IS system The desirable characteristics and features of IS system Kim et al (2012), Hassanzadeh et al (2012) Petter et al (2008) Petter et al (2008) The desirable characteristics and features of the output The degree to which a person believes that using a particular system would enhance his or her job performance The degree to which a person believes that using a particular system would be free of effort The extent to which user believe that their needs, goals, and desires have been fully met Key likelihood that an individual will use a technology Petter et al (2008) Davis (1989) Davis (1989) Sanchez-Franco (2009) Schierz, Schilke, and Wirtz (2010) H12 Satisfaction positively affects intention to use Educational Quality Service quality Technical system quality Content and information quality H13 Satisfaction positively affects actual use H2 H1 H4 3.8 Intention to use Satisfaction H5 H6 H13 Actual use H12 H7 H8 Perceived ease of use H3 H14 Intention to use H9 H10 Perceived usefulness H11 Fig Research model towards use of e-learning services (Chen and Tseng, 2012; Cheng, Wang, Moormann, Olaniran, & Cheng, 2012; Chow et al., 2012; Islam, 2012, 2013; Li et al., 2012; Liu et al., 2010; Sumak et al., 2011) As a consequence, the greater the perceived usefulness of e-learning system, the more positive is the intention towards its usage; thus greater the likelihood that it will be used Intention, which is the main dependent variable identified in the studies conducted based on the TAM, is defined as the likelihood that an individual will use an IS Intention plays a critical role in the actual use of a new technology (Davis, 1989) Intention to use can also be considered as an attitude (DeLone & McLean, 2003) In the acceptance domain, some researchers have studied the relationship between intention and actual use in e-learning context (e g., Alkhalaf, Drew, AlGhamdi, & Alfarraj, 2012; Chow et al., 2012; Hassanzadeh et al., 2012) Petter et al (2008) note that to refrain more complexity, IS success model did not distinct between intention to use and system use in their updated model, but intention to use is generally an individual level construct Venkatesh, Morris, Davis, and Davis (2003) confirms the positive relationship between intention to use and actual use Thus, in the context of this study, intention to use assumed to have a positive impact on actual use Table lists the dimensions’ and definitions; Fig shows the conceptual model H14 Intention to use positively affects actual use H11 Perceived usefulness positively affects intention to use Instrument development 3.7 Satisfaction Rather than to sell, to supply, or to serve, the main objective of every business is to satisfy the needs and meet the satisfaction of its users (Docimini and Palumbo, 2013) Satisfaction is defined as the individuals’ perceptions of the extent to which their needs, goals, and desires have been fully met (Sanchez-Franco, 2009) and refers to their overall view of IS (Wang & Wang, 2009) It sounds better to note that user satisfaction refers to the extent to which users are pleased with IS and support services (Petter et al., 2008) The updated IS success model assumes that system use precedes user satisfaction which leads to an increased satisfaction which sequentially results in a higher intention to use (Petter et al., 2008) Satisfaction has been found to have a significant positive effect on intention towards use of e-learning services in some studies (Chang, 2013; Hassanzadeh et al., 2012; Islam, 2012; Petter et al., 2008; Roca et al., 2006; Udo et al., 2011) Satisfaction has been found to have a significant positive effect on actual use as well Hassanzadeh et al (2012) in their study uncovered the positive effect of satisfaction on actual use of e-learning system Therefore, in the context of this study, satisfaction assumed to have a positive impact on both intention to use and actual use The final structured instrument was used to collect data using a seven-point Likert scale: perceived usefulness and perceived ease of use were adopted from Kim and Mirusmonov (2010), intention to use from Lin (2011), system, service, and information quality, and satisfaction from DeLone and McLean (2003), and educational quality along with actual use from Hassanzadeh et al (2012) To ensure the validity of the instrument, the first Confirmatory Factor Analysis (FCFA) was taken Studying the interior structure of a collection of indices and validity measures, this approach sought to evaluate factor loadings and relationships between a collection of indices and their corresponding factors As seen in Table 2, in the FCFA of sample group (20% of total), except for four indices, almost all indices received the standardized factor loadings larger than the recommended value (0.4); thus, having excluded the invalid indices, the model was tested with other selected indices so that the instrument to be valid 4.1 Data collection The research aimed to understand the e-learning satisfaction and intention towards actual use of e-learning in Tehran early 2014 This period was marked by recent developments in Iran 365 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Table The research instrument Construct Question Source Factor loading Educational quality E-learning assures the presents of students E-learning provides collaborative learning E-learning provides required facilities such as chat and forum E-learning provides possibility of communicating with other students E-learning provides possibility of learning evaluation E-learning is appropriate with my learning style Chang and Chen (2009) Hassanzadeh et al (2012) Lee (2010) Lee (2010) 0.77 0.71 0.69 0.66 Hassanzadeh et al (2012) Vernadakis, Antoniou, Giannousi, Zetou, and Kioumourtzoglou (2011) 0.32 0.63 Wang and Wang (2009) 0.78 Au, Ngai, and Cheng (2008) Andrade and Bunker (2009) 0.81 0.73 Au et al (2008) 0.66 is aesthetically satisfying optimizes response time is user friendly provides interactive features between users and Ho and Dzeng (2010) DeLone and McLean (2003) Ozkan and Koseler (2009) Ozkan and Koseler (2009) 0.54 0.76 0.63 0.62 possesses structured design has flexible features has attractive features is reliable is secure Ho and Dzeng (2010) Au et al (2008) Wang, Wang, and Shee (2007) Ozkan and Koseler (2009) Ozkan and Koseler (2009) 0.74 0.66 0.83 0.71 0.67 information that is relevant to my needs comprehensive information information that is exactly what I want me with organized content and Au et al (2008) Ho and Dzeng (2010) Wang and Wang (2009) Ozkan and Koseler (2009) 0.82 0.64 0.72 0.60 up to date content and information required content and information Wang and Liao (2008) Wang et al (2007) 0.65 0.57 Wang and Liao (2008) DeLone and McLean (2003) DeLone and McLean (2003) DeLone and McLean (2003) DeLone and McLean (2003) 0.64 0.69 0.34 0.49 0.74 Service quality Technical system quality Information quality E-learning provides a proper online assistance and explanation E-learning department staff responds in a cooperative manner E-learning provides me with the opportunity of reflecting views E-learning provides me with courses management E-learning E-learning E-learning E-learning system E-learning E-learning E-learning E-learning E-learning E-learning provides E-learning provides E-learning provides E-learning provides information E-learning provides E-learning provides Perceived ease of use E-learning E-learning E-learning E-learning E-learning is is is is is easy to use easy to learn easy to access easy to understand convenient Perceived usefulness E-learning E-learning E-learning E-learning E-learning E-learning E-learning helps to save time helps to save cost helps me to be self-reliable helps to improve my knowledge helps to improve my performance is effective is efficient DeLone and McLean (2003) Ho and Dzeng (2010) Chiu and Wang (2008) Hassanzadeh et al (2012) Hassanzadeh et al (2012) DeLone and McLean (2003) DeLone and McLean (2003) 0.64 0.48 0.64 0.69 0.73 0.76 0.64 Satisfaction E-learning is enjoyable I am pleased enough with e-learning system E-learning satisfies my educational needs I am satisfied with performance of system E-learning is pleasant to me E-learning give me self-confidence DeLone and McLean (2003) Lee (2010) Lee, Yoon, and Lee (2009) Wu et al (2010) Lee (2010) DeLone and McLean (2003) 0.75 0.64 0.63 0.37 0.57 0.59 Intention to use I tend to use e-learning system I believe that use of e-learning is available I am likely to use e-learning system in the near future Lin (2007) Lin (2007) Lin (2011) 0.54 0.31 0.66 Actual use I use e-learning on daily basis I use e-learning frequently I visit e-learning often DeLone and McLean (2003) DeLone and McLean (2003) DeLone and McLean (2003) 0.57 0.56 0.49 which push researchers and educators to take a pedagogical view towards developing educational applications to promote teaching and learning; hence, this study offers an appropriate window for studying variations in educator’s intention The sample is taken from the students of four public universities of Tehran including Elm-o-Sanat, Amir Kabir, Shahid Beheshti, and Tehran universities The final questionnaire was arrived at after examining theoretical literature and studies undertaken by previous researchers based on which indices were selected (Table 2) The research used stratified sampling – since it was concerned with different attributes of research population The research model uses a cross-sectional survey In fact, the research model is investigated based on views expressed by the respondents at one point of time This approach, as one of the common approaches, was taken due to theoretical and survey limitations In the Cochran formula for finite population, with Z a2 value of about 1.96, e value less than 0.1 of about 0.099 and q value of about 0.5, each university was calculated at a minimum of 81 Students had to confirm they are users of e-learning system before the questionnaire was released to them A total of 420 students were selected Next, participants were intercepted in randomly chosen faculties where questionnaires were physically administered to them There were a total of 105 questionnaires for each university in three main faculties, out of which 390 were gathered This research is practical 366 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Table Sample selection University Students (N) in each university n for each university Frequency of sample/population Percentage (%) Elm-o-Sanat Shahid Beheshti Amir Kabir Tehran Total 500 550 580 500 42,200 81 81 81 81 324 97/105 98/105 97/105 98/105 390/420 24.6 25.6 24.6 25.1 100 Table The demographic characteristics of the sample Frequency Percentage Gender Female Male Total 192 198 390 49.2 51.8 100.0 Age 20 20–30 Total 48 342 390 12.3 87.7 100.0 Education BA MA Total 110 280 390 28.2 71.8 100.0 in nature and the goal was to conduct it from an extensive perspective; it was thus, exploratory and descriptive in approach Alpha Cronbach for the questionnaire emerged to be 0.839 n¼ N  z a2  p  q e2  N  1ị ỵ z a2  p  q Formula Cochran formula for finite population Data analyses 5.1 Response rate and representatives Table summarizes the response rate Three hundred and ninety out of four hundred twenty questionnaires were collected with valid data The discard rate was low The total population of Iranian students by sex and age group was obtained from Iran Center of Census and Statistics This was compared to the gender and age distribution of the sample in order to test its’ representativeness In terms of gender, the distribution of the sample was 51.8% male and 49.2% female According to the Technology and Science Minister’s latest report, by end of 2013, the male to female population of student ratio in Iran was 47% and 53%; thus the sample appeared to be representative in terms of gender distribution Having analyzed the demographic characteristics of e-learning students, it was concluded that most of them (87.7%) were in the age group of 20–30 years followed by those in the age group of 20-years (12.3%) The population of Iranian e-learning students shared a similar age distribution of 78% and 22% respectively This indicates that the sample is representative of the Iranian e-learning population In addition, MA students (71.8%) dominated other groups Table presents the demographic characteristics of the sample 5.2 Exploratory and confirmatory analysis To perform an exploratory analysis, convergent and discriminant validities and scale reliability are considered (Fraering and Minor, 2005) Convergent validity measures whether items can effectively reflect their corresponding factors, while discriminant validity measures whether two factors are statistically different from each other (Anderson and Gerbing, 1988) Following the two-step approach proposed by Anderson and Gerbing (1988), we first examined the measurement model to test its reliability and validity Then we examined the structural model to test the model fitness and the relationships between variables Table lists Average Variance Extracted (AVE), Composite Reliability (CR), R square (R2), Communality, and Cronbach alpha values, and standardized factor loadings As seen in Table 5, almost all factor loadings are larger than 0.4, while t-values (shown in Fig 3) indicate that all of them are significant at 0.05 All AVEs exceed 0.5, all CRs (the degree to which items are free from random error and therefore render consistent results) exceed 0.7, and all communalities exceed 0.7 showing minimally accepted construct reliability (Gefen, Straub, & Boudreau, 2000) Thus, the scale has a good convergent validity In addition, all alpha values are larger than 0.7, showing good reliability (Nunnally, 1978) On the other hand, intention – with an R2 of about 0.63 is proven to be well predicted by its predictors and the remaining 0.36 is the prediction error Besides, satisfaction, with an R2 of about 0.22 is partially forecasted by its predictor, and the remainder 0.77 is the prediction error Therefore, users’ intention is proved to be a strong predictor of their actual use of e-learning At last, actual use – with an R2 of about 0.73 is proven to be well predicted by its predictors which are users’ intentions and satisfaction Moreover, the indices used for ‘‘satisfaction’’, ‘‘intention’’, and ‘‘actual use’’ gained larger factor loadings than the recommended values which underlines their careful selection To examine the discriminant validity, the squared roots of the AVEs are compared with the factor correlation coefficients As listed in Table 6, for each factor, the square root of AVE is larger than its correlation coefficient with other factors, showing good discriminant validity (Gefen et al., 2000) In the second step, we employed structural equations modeling by LISREL 8.80 to estimate the structural model 5.3 Path coefficient As listed in Table 7, among the factors influencing satisfaction, information quality (c = 0.29, p < 0.01) and technical system quality (c = 0.29, p < 0.01) showed the greatest effects, educational quality (c = 0.27, p < 0.01) and service quality (c = 0.24, p < 0.01) had significant paths as well Among the factors influencing intention to use, technical system quality (c = 0.23, p < 0.01), and service quality (c = 0.17, p < 0.01), information quality (c = 0.13, p < 0.01) had respectively significant positive paths However, educational quality (c = 0.03) showed no significant effect in this regard Perceived usefulness (b = 0.52, p < 0.001) had significant positive path towards intention, while perceived ease of use (c = 0.07) showed no significant effect on the intention to use Furthermore, perceived ease of use (c = 0.16, p < 0.001) had a significant effect on perceived usefulness Satisfaction (b = 0.52, p < 0.001) also appeared to have a significant positive path towards intention to use Finally, satisfaction (b = 0.18, p < 0.001) and intention to use 367 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Table Main statistics Factor Item Standardized loading AVE CR R square Communality Alpha Actual use Act1 Act2 Act3 0.73 0.70 0.69 0.8354 0.8775 0.7373 0.8354 0.8655 Ease of use Eas1 Eas2 Eas3 Eas4 0.86 0.81 0.88 0.85 0.7813 0.8674 0.7813 0.8551 Educational quality Edu1 Edu2 Edu3 Edu4 Edu5 0.80 0.79 0.73 0.75 0.70 0.7623 0.8691 0.7623 0.8605 Information quality Inf1 Inf2 Inf3 Inf4 Inf5 Inf6 0.85 0.72 0.83 0.73 0.80 0.84 0.7669 0.8750 0.7669 0.8692 Intention to use Int1 Int2 0.78 0.77 0.8605 0.8799 0.6312 0.8605 0.8589 Satisfaction Sat1 Sat2 Sat3 Sat4 Sat5 0.86 0.82 0.80 0.76 0.79 0.7578 0.8779 0.2228 0.7578 0.8586 Service quality Ser1 Ser2 Ser3 Ser4 0.88 0.85 0.82 0.88 0.7705 0.8641 0.7705 0.8501 System quality Sys1 Sys2 Sys3 Sys4 Sys5 Sys6 Sys7 Sys8 Sys9 0.82 0.83 0.80 0.72 0.85 0.74 0.82 0.84 0.77 0.7735 0.8842 0.7735 0.8819 Perceived usefulness Per1 Per2 Per3 Per4 Per5 Per6 Per7 0.75 0.82 0.81 0.88 0.83 0.89 0.87 0.7616 0.8739 0.7616 0.8679 0.0230 Table The square root of AVE (italic at diagonal) and correlation coefficients ACT EAS EDU INF INT SAT SER SYS USE ACT EAS EDU INF INT SAT SER SYS USE 0.9140 0.2809 0.0875 0.5938 0.8859 0.4216 0.5845 0.4418 0.1027 0.8839 0.6039 0.2927 0.1829 0.0253 0.1296 0.0965 0.1518 0.8730 0.2311 0.0575 0.1569 0.1054 0.3416 0.1302 0.8757 0.5221 0.2664 0.6000 0.4084 0.1302 0.9276 0.2237 0.5277 0.6015 0.1291 0.8705 0.2208 0.0090 0.7979 0.8777 0.5352 0.1328 0.8794 0.1283 0.8726 (b = 0.85, p < 0.001) both positively affected actual use of e-learning Therefore, all paths except H2 and H9 are supported Path coefficients and their significances are listed in Figs and 5.4 Measurement of the model fitness To ensure that the measurement model possesses a sufficiently good model fit, the overall model fit is assessed in terms of seven common measures: Normed v2- the ratio of v2 to the degree of freedom, Goodness of Fit Index (GFI), Comparative Fit Index (CFI), Normed Fit Index (NFI), Non-Normed Fit Index (NNFI), Incremental Fit Index (IFI), and Root Mean Square Error of Approximation (RMSEA) A model fit is usually considered strong when Normed v2 is smaller than 3, GFI is larger than 0.8, CFI, NFI, NNFI, and IFI are larger than 0.9, and RMSEA is around 0.06 Table lists the recommended and actual values of fit indices The actual values of all fit indices were better than the recommended values, showing a superior fit 368 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Table Path coefficients and significances Question H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 * ** *** Path Educational ? satisfaction Educational ? intention Service ? satisfaction Service ? intention System ? satisfaction System ? intention Information ? satisfaction Information ? intention Ease of use ? intention Ease of use ? usefulness Usefulness ? intention Satisfaction ? intention Satisfaction ? actual use Intention ? actual use Path coefficient ** 0.27 0.03 0.24** 0.17*** 0.29** 0.23*** 0.29** 0.13** 0.07 0.16* 0.1*** 0.52*** 0.52*** 0.85*** Supported or not Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes If the indirect path gains a greater effect than that of direct one, then indirect path would prove to be mediator – that is a direction which leads us to actual use faster As showed in Table 10, to examine the mediating effect of perceived usefulness in the relationship between ease of use and intention, both the direct and indirect effects of perceived ease of use were tested Given the insignificance of the path ease of use-intention (0.07), perceived ease of use can only affect intention through usefulness (0.083); therefore, the mediating role of usefulness (H10) is proven to be significant at the 0.05 significance level Discussion Significance codes: 0.05 Significance codes: 0.01 Significance codes: 0.001 5.5 Path analysis As seen in Fig 4, the path to ease of use: usefulness–intention– actual use was tested via path analysis, in which the path was proved to be significant at q = 0.000 On the other hand, the path ease of use–intention did not prove to be significant at q = 0.000 which underlies the insignificant path coefficient derived from structural model The ratio of path loading to standard error indicates that path loadings are greater than twice their standard errors showing convergent reliability; their variances also substantiate the decision to use The results of path analysis involving regression coefficients and their significances are listed in Table In view of the fact that user satisfaction and intention to use both affect users’ actual use positively, it can be concluded that educational quality, service quality, technical system quality, and information quality – those with significant effects – positively affects users’ actual use – all indirectly and through satisfaction and intention In fact, the e-learning system posses users’ relative confidence about educational quality, service quality, technical system quality, and information quality; among which technical system quality appears to have a greater positive effect than others This confirms what Hassanzadeh et al (2012) concluded in their study in which technical system quality was found to be the strongest factor affecting users’ satisfaction of e-learning system in Iran Alsabawy et al (2013), Motaghian et al (2013), Saba (2013), Tajuddin et al (2013), Kim et al (2012), and Islam (2012), in their studies into e-learning systems, found that system quality positively affects user satisfaction as well, which corresponds with the studies undertaken by Wang and Chiu (2011), Rai et al (2009), and Wu et al (2008) in e-learning context Islam (2012), Cheng (2012), Li et al (2012), Ramayah et al Fig Standard coefficients 369 H Mohammadi / Computers in Human Behavior 45 (2015) 359–374 Fig Significance values Table The values of fit indices Fit indices x2/df (GFI) (IFI) (CFI) (NNFI) (NFI) (RMSEA) Actual Recommended 2.90 0.80 0.90 >0.90 0.90 >0.90 0.92 >0.90 0.92 >0.90 0.05