Explaining Students’ Continuance Intention to Use Mobile Web 2.0 Learning and Their Perceived Learning: An Integrated Approach

Problem Statement and Research Contributions

Utilization of MW2.0 in the educational context is considered as recent phenomenon in which theoretical underpinnings from two associated areas of learning and social networking sites (SNSs) are still in their infancy. In the literature, there is a scarcity of studies investigating the factors influencing the deployment of MW2.0 as pedagogical tools in higher education (R.-B. Wang & Du, 2014). The ubiquitous attribute of MW2.0 does not in themselves result in the sustainable adoption of these tools for learning activities. Consequently, there is a need to explore the determinants and factors that influence the adoption behavior in the higher education context. By identifying the determinant factors, demands of students and lecturers can be aligned with strategic planning of universities, better and significant integration of technology with learning and teaching can be made, and further policy decisions can be enhanced. Perspectives related to the use of MW2.0 in work and learning contexts are varied. Limited research investigated how students use MW2.0 tools and services for their learning activities (C.-H. Wong et al., 2015) and further motivations of technology adoption by students (Novakovich, Miah, & Shaw, 2017).

As higher education institutes (HEIs) attempt to deploy and implement new technological approaches into their teaching and learning activities, it is essential that research establishes how MW2.0 tools are used outside of classrooms by students and what are their preferences to use these tools to support their learning. Limited studies investigated empirically the adoption and usage of MW2.0 for learning purposes in developing counties, particularly Malaysia, and further its impact on students’ learning and academic performance. The main focus of previous studies conducted in Malaysia was mostly on the adoption of mobile devices but not on MW2.0 as a new pedagogical tool (Tan et al., 2012). Accordingly, this study aims to fill the gaps in the literature by developing and testing the model that investigates the determinants of MW2.0L adoption by students and further its impact on their perceived learning (PL). Hence, this study tries to answer the following research questions:

What factors impact the intention of students to continue use of MW2.0L?

This study contributes to the growing body of knowledge in MW2.0L. It is built upon existing literature on students’ use of technology by conducting an empirical research about the extent to which students utilize MW2.0 tools as pedagogical tools. Understanding the utilization of technologies by students as learning tools in their personal networking environments is important, since it aids researchers and practitioners in understanding how to incorporate new technologies (here MW2.0) into teaching and learning activities to meet students’ needs in a purposeful manner.

Furthermore, this research contributes to understanding of the approach that HEIs can take to prepare students to participate in online learning networking environments. This study not only provides guidelines to students to adopt MW2.0 for their learning purposes but also provides guidance to educators about how to use technology in their teaching activities. This research contributes to understanding of how teachers can put efforts to develop advantages of mobile technologies and social networking to enhance academic performance of students. Finally, through understanding the adoption behavior of MW2.0 as pedagogical tools, this study contributes to understanding of how MW2.0 providers can integrate learning features to further development of applications.

The article is organized as follows. The following section presents the theoretical foundations of the study. Then the research model and development of related hypotheses are provided. Methodology of the study is then discussed. Then analysis and results are then presented. The results of the study and the research findings are then reported and discussed. The final section provides implications of the study both for literature and practice together with brief conclusions and directions for future studies.

Mobile Web 2.0 Learning

According to T. Cochrane and Bateman (2013), MW2.0 tools and services are the ones which are specifically designed and formatted to be used with mobile devices. Social constructivist learning environments (SCLEs) and MW2.0 tools share several synergies. SCLEs developed in the context of social constructivism theory are described as “a way of knowing in which students or learners construct their new understanding and knowledge during the process of social interaction with others” (Sthapornnanon, Sakulbumrungsil, Theeraroungchaisri, & Watcharadamrongkun, 2009, p. 1). In such environments, students’ learning is occurred when they share their background information and are actively participating in a give and take cooperating activities. In social constructivism theory, the emphasis is on students in that students would learn better through SIs with peers. Collaborative group works, user-generated contents, formative feedbacks, and many other processes provided by Web 2.0 tools and services similar to those utilized in SCLEs, motivated many educators and instructors to harness such tools in their leaning and pedagogy activities (T. D. Cochrane, 2014).

Following the aforementioned definitions and discussions and following the studies conducted by T. D. Cochrane (2014) and T. Cochrane and Bateman (2013), MW2.0L refers to “appropriating the benefits of web 2.0 and pedagogy 2.0 anywhere anytime using MW2.0 and wireless mobile devices (or WMDs), in particular WiFi (wireless Ethernet) and 3G (third generation mobile ‘broadband’) enabled smartphones, and 3G enabled netbooks” (T. D. Cochrane, 2011, p. 13).

Technology Acceptance Models

Fishbein and Ajzen (1975) developed TRA which “brought a compelling and coherent structure on the field of attitudes” (Sheppard, Hartwick, & Warshaw, 1988, p. 340). While targeting to develop “a cumulative body of knowledge in the attitude area” (Fishbein & Ajzen, 1975, p. 520), TRA differentiated attitude (AT), beliefs, intention, and behavior. The cornerstone of this theory is that “human beings are usually quite rational and make systematic use of information available to them” (Ajzen & Fishbein, 1980, p. 5). TRA asserts that the immediate predictor of behavior is intention, where intention is predicted by AT and subjective norm (SN). Numerous studies proven the capability of the model to predict individuals’ behavior or intention in various context such as green product consumption (Paul, Modi, & Patel, 2016), cyberbullying prediction (Doane, Pearson, & Kelley, 2014), Internet banking (Yousafzai, Foxall, & Pallister, 2010), SNSs usage for sustainable purchase behavior (Reiter & Connell, 2017), and organ donation (S. H. Wong & Chow, 2016). However, TRA model was not without critiques. Criticisms have highlighted that TRA model was not successful in explaining “irrational decisions, habitual actions and other unintentional behaviors” (James, 2015, p. 32). Furthermore, while it considers the extent to which the desired outcome is delivered by the innovation, it ignores decision-makers’ degree of outcome that they desire (James, 2015). Moreover, scholars reported deficiency of TRA regarding its tendency toward subjective data because of its reliance on subjects’ self-reporting (Ajzen, 1988; Ajzen & Fishbein, 1980). Despite all the criticisms, TRA remained “a powerful tool” (Jackson, Quaddus, Islam, & Stanton, 2006, p. 6). TPB was developed by Ajzen (1985) to overcome identified deficiencies with TRA (Ajzen, 1985, 1991). Accordingly, the construct of perceived behavioral control (PBC) as the determinant of intention and behavior was added to the model. TPB has successfully been applied in explaining various technology adoption behaviors such as e-learning adoption (Chu & Chen, 2016), green purchase behavior (Liobikienė, Mandravickaitė, & Bernatonienė, 2016), SNSs adoption (Y. P. Chang & Zhu, 2011), and Internet banking behavior (Yousafzai et al., 2010). Alike TRA, TPB considers rational choices toward a specific behavior such as a tradeoff between cost and benefit (Ajzen, 1991). And also like TRA, while TPB considers the extent to which a desired outcome is delivered, it ignores the desired outcome considered by the decision-maker. The technology acceptance model (TAM) was developed by Davis (1986) to explain technology adoption behavior and has become “the leading model in explaining and predicting system use” (Chuttur, 2009). TAM which was widely applied in information systems research, provided concise clarification of technology usage behavior by focusing on two predictors of perceived usefulness (PUS) and perceived ease of use (PEU; Davis, 1989). TAM has been widely utilized to explain acceptance behavior of various technologies in different situations such as renewable energy technology acceptance (Kardooni, Yusoff, & Kari, 2016), e-learning adoption (Tarhini, Hone, Liu, & Tarhini, 2016), mobile apps usage for learning purposes (Wai et al., 2016), wearable technology adoption behavior (Y. W. Ha, Kim, Libaque-Saenz, Chang, & Park, 2015), and social media adoption by business-to-business organizations (Siamagka, Christodoulides, Michaelidou, & Valvi, 2015).

Technology-to-Performance Chain Model

To gain further understanding of the factors that influence students’ PL outcomes in the context of MW2.0L, it would be appropriate to refer to models and theories that have shown promises in predicting information system success. Technology-to-performance chain (TPC) model proposed by D. Goodhue and Thompson (1995) is one such model.

TPC postulates that the success of information systems depends on recognizing both the task in which the technology is used for and the fit between the task and the technology (D. Goodhue & Thompson, 1995). Task-technology fit (TTF) is defined as “the degree to which a technology assists an individual in performing his or her portfolio of tasks” (D. Goodhue & Thompson, 1995, p. 216). In the case of students’ use of MW2.0L, TTF refers to the capabilities of MW2.0 tools to act as a platform to enhance personal learning, sharing of knowledge, facilitating students’ engagement, college experiences, and educational practices (C.-H. Wong et al., 2015).

As shown in Figure 1, TTF is determined by task characteristics, technology characteristics, and individual characteristics. TTF is influencing performance directly and utilization indirectly through precursors of utilization such as AT toward technology, beliefs, social norms, expected consequences, and facilitating conditions. The use of technology is impacting performance directly. According to TPC, an information technology (IT) would have positive impact on individuals’ performance once it is fit to the task that it is supposed to support and it should be used by users. OPEN IN VIEWER

The influence of TTF has been investigated using parts of the TPC model in various domains. D. Goodhue and Thompson (1995) initially investigated the impact of TTF on individuals’ performance by adopting a subset of TPC using the participants from transport and insurance companies and reported the strong support of predicting performance by TTF. They also found some support of technology characteristics and task characteristics in explaining TTF.

Researchers investigated parts of TPC in learning and education domain as well, including learning management systems adoption (T. J. McGill & Klobas, 2009), e-book adoption (D’Ambra, Wilson, & Akter, 2013), mobile learning (S. Y. Park, Nam, & Cha, 2012), and SNSs adoption (Rad, Dahlan, Iahad, & Zakaria, 2014). Staples and Seddon (2004) tested the most comprehensive model of TPC by investigating the use of library cataloging system by staff and the use of spreadsheets and word processing by students. They found the strong supports for the influences of TTF on performance and ATs and beliefs on usage.

The TPC model and the role of TTF have not yet been investigated in the context of MW2.0L. Given the need for the extensive research on the factors determining the successful utilization of MW2.0L and the relevance of TPC model, it could be appropriate framework for the purpose of this study.

Uses and Gratifications Theory

The uses and gratifications (U&G) theory refers to “the study of the gratifications or benefits that attract and hold audiences to various types of media and the types of content that satisfy their social and psychological needs” (Dunne, Lawlor, & Rowley, 2010, p. 47). Since MW2.0 is media, U&G becomes an appropriate theory to the current study (Y. W. Ha et al., 2015). This theory is considered as an ideal theoretical lens to investigate the motivations and gratifications that users seek to obtain when they use MW2.0 tools (Chaouali, 2016). Users’ will is necessary in using MW2.0 tools (Xu, Ryan, Prybutok, & Wen, 2012) which is different from traditional media in that users were exposed to them unwillingly. Regarding latter services like MW2.0 tools and services, users only access to the service when they have the appropriate application on their smartphones (Shin, 2011). Thusly, MW2.0 satisfies three expectations of U&G theory which are (a) people who are using MW2.0 tools and services are active users of the media, (b) people select MW2.0 tools and services according to their goals and purposes, and (c) users are aware of their motivations for selecting MW2.0 tools and services (L. Ha & Fang, 2012). U&G theory has been applied in the context of SNSs usage and adoption extensively (Hartmann, Apaolaza, He, Barrutia, & Echebarria, 2017).

Technological Convenience (Mobility)

The gratification-opportunity of MW2.0L is related to their technological convenience. The most convenience opportunity obtained through MW2.0L is their mobility which “permits a spatial and temporal flexibility” (Lo & Leung, 2009). Compare to other medium, mobility is an attribute of MW2.0 tools and services which reflects their ubiquity (i.e., anytime-anywhere). Accordingly, users can access to their Web 2.0 applications through their mobile devices without time and place restrictions (Nikou & Bouwman, 2014). Convenience of users is enhanced by using mobile devices (Y. W. Ha et al., 2015) which make them to be an integral part of people lives (Nikou & Bouwman, 2014). In the context of MW2.0L, mobility has been found as an important factor in determining individuals’ AT and behavior in using MW2.0 tools and services (Chaouali, 2016; Y. W. Ha et al., 2015; Nikou & Bouwman, 2014). Therefore, in education context, we can also posit that:

Hypothesis 1 (H1): Mobility positively influences students’ behavioral intention to continue use of MW2.0L.

IE (Cognitive Needs)

In the current context and following the study by Nambisan and Baron (2007), cognitive gratifications refer to “product-related learning”—that is, gaining better knowledge and understanding regarding the product, its underlying technologies and its usage. In the literature, cognitive needs also refer to “IE” as well, which reflects “people’s desire to increase awareness and knowledge of one’s self, others, and the world” (Shao, 2009, p. 10). In previous studies, similar concepts such as utilitarian, learning, and information were used reflecting the concept of cognitive needs (Calder, Malthouse, & Schaedel, 2009; Nambisan & Baron, 2007).

MW2.0 tools and services hold valuable collective knowledge regarding its usage for learning purposes which is generated and disseminated through continued learners and instructors interactions (R.-B. Wang & Du, 2014; C.-H. Wong et al., 2015).

Hausman and Siekpe (2009) reported that there is a positive and significant relationship between the content of a website and user’s AT toward that website, in that, ATs of users would be improved by making the content of a website more informative. In the context of virtual customer environments’ adoption, Nambisan and Baron (2007) showed that customers’ perception of interaction-based benefits (i.e., cognitive gratifications) would positively shape their usage behavior of these tools. In the context of MW2.0, users are able to obtain recent information which is retrieved through a network of acquaintances, to acquire information on recent trends related to them, and further to obtain answers to various questions (Humphreys, 2007). In the context of SNSs, Van-Tien Dao, Nhat Hanh Le, Ming-Sung Cheng, and Chao Chen (2014) revealed that information seeking is an important predictor of consumers’ perceived value of social media advertising. Hence, following the aforementioned discussion, we can posit that:

Hypothesis 2 (H2): IE positively influences students’ behavioral intention to continue use of MW2.0L.

Social Interactions

According to Phang, Kankanhalli, and Sabherwal (2009), SI refers to “individuals feel at ease and comfortable to engage in interpersonal communication exchanges through the technology-enabled space” (p. 729). Van-Tien Dao et al. (2014) argued that,

compared with other potential motives for joining social media communities, such as information seeking, SNS users are likely to put more weight on SI and connection to build and maintain their social relationships with others, as well as to seek social support and a sense of ‘belongingness’. (p. 277)

RC (Hedonic Gratification)

Through the lens of U&G theory, another gratification which motivates individuals toward the use of a specific media is the gratifications for RC. This category of gratifications is also recognized as “hedonic gratifications” in the literature which are the ones related to strengthening one’s joyful experience and refers to “aesthetic related to deviation, resting, enjoyment, and spending time” (Y. W. Ha et al., 2015, p. 429). In the literature, authors used similar concepts such as “escapism,” “entertainment,” and “passing time” (Calder et al., 2009; Nambisan & Baron, 2007). Hedonic gratifications are considered as “self-protective strategy to prevent seemingly disastrous consequences” (Kashdan, Barrios, Forsyth, & Steger, 2006, p. 1301) and also is referred as “avoidance coping strategy” to relief aversive feelings associated with problems of everyday life (Norton, 2012). In the studies by Chen, Clifford, and Wells (2002) and Hausman and Siekpe (2009), the authors reported that the ATs of users are more favorable toward the websites that their contents are more entertaining. Related to mobile SNSs, boredom can be eliminated by “engaging in dialog with acquaintances during [. . .] spare time or examining the information generated from networks” (Y. W. Ha et al., 2015, p. 429). In the context of SNS usage through mobile devices, Cheng et al. (2015) identified needs for entertainment and seeking fashion or status as different types of recreational needs which influence civic engagement. However, in their study, they have found that needs for entertainment and fashion or status were among the weak predictors of mobile SNSs usage. In this study, gratifications for RC are developed as a higher order construct and further is measured by two variables of needs for entertainment and fashion or status. Therefore, based on the aforementioned discussion, we can postulate that:

Hypothesis 4 (H4): Gratifications for RC (measured by the needs for entertainment and fashion/status) positively influences students’ behavioral intention to continue use of MW2.0L.

Perceived Ease of Use

PEU is a construct included in the original TAM. PEU is defined as “the degree that using a specific technology will be free from effort” (Davis, 1989, p. 320). Prior studies reported that there is a positive relationship between PEU and intention (E. Park, Baek, Ohm, & Chang, 2014). E. Park et al. (2014) found that PEU is an important determinant of users’ AT toward the use of mobile social network games. Consequently, based on the previous studies that showed PEU is an important determinant of behavioral intention, we can also state that:

Hypothesis 5 (H5): PEU positively influences students’ behavioral intention to continue use of MW2.0L.

Perceived Usefulness

According to TAM, PUS strongly affects users’ behavior toward the acceptance of a specific technology. Davis, Bagozzi, and Warshaw (1989) defined PUS as “the perceived degree to which an individual believes that using a specific service or system improves his or her task performance” (p. 320). This definition relates the concept of PUS to one’s “task performance.” Prior studies in the field of technology acceptance reported significant relationship between PUS and intention (Y. W. Ha et al., 2015; E. Park et al., 2014; Ratna & Mehra, 2015). Wu and Chen (2016) reported that PUS is an important factor in explaining students’ intention to use massive open online courses (MOOC). Related to the adoption of Moodle in a blended learning setting, Yeou (2016) showed that PUS plays an important role in determining students’ intention behavior. Accordingly, following aforementioned logic and related to MW2.0L, we can hypothesize that:

Hypothesis 6 (H6): PUS positively influences students’ behavioral intention to continue use of MW2.0L.

Perceived Playfulness

Since accessing Web 2.0 tools through mobile devices are considered as complex hedonic systems in which content and entertainment services are delivered ubiquitously (Shin & Shin, 2011), playfulness of learning activities through MW2.0 tools and services can be a significant determinant of user’s acceptance of MW2.0L. Due to proliferation of hedonic mobile systems, playfulness has been considered as an important research subject in the IS literature (Sanakulov & Karjaluoto, 2015). In the study by Van der Heijden (2004) which has been conducted on the acceptance of hedonic information systems, the author reported that playfulness dominates usefulness in influencing users’ behavior to adopt the technology. The crucial role of playfulness in explaining users’ adoption of new technologies has been reported vastly in the literature (Davis, Bagozzi, & Warshaw, 1992; Yang et al., 2016). In a recent study by Yang et al. (2016), authors reported the significant effect of playfulness on users’ mobile SNSs engagement. In the context of SNSs, Turel and Serenko (2012) showed the positive significant influence of users’ playfulness on their usage of SNSs. In other words, when users perceive mobile SNSs enjoyable, they tend to “to put more effort into the usage, focus longer, involve deeper, and utilize the services repeatedly in the future” (Yang et al., 2016, p. 489). Accordingly, the more playful and enjoyable the MW2.0 tools are in using learning activities, users are more likely to engage using MW2.0L. Hence, in this study, when the use of MW2.0 for pedagogical purposes is accompanied with playfulness, users are more likely to use them. Therefore, we posit that:

Hypothesis 7 (H7): Perceived playfulness positively influences students’ behavioral intention to continue use of MW2.0L.

Personal Innovativeness in Information Technology

According to Rogers (1995), innovative individuals would able to handle uncertainties and would have higher intentions to adopt and accept innovations. Agarwal and Prasad (1998) point out that innovative individuals are more inclined to use technology since they develop more positive beliefs toward the technology. Wood and Swait (2002) recognize innovativeness as a personality trait since it is only observed in selected individuals. J. Lu, Yao, and Yu (2005) reported that individuals with greater level of innovativeness in IT are going to have better perceptions toward innovations. In the context of MW2.0L, it can be stated that individuals’ perceptions toward the decisions to adopt such tools would be restrained by innovative mindset, in other words, innovative people are anticipated to generate more positive ATs toward the new IT (López-Nicolás, Molina-Castillo, & Bouwman, 2008). Crespo and del Bosque (2008) reported that individuals with higher degrees of PIIT are more inclined to adopt B2C e-commerce. In the context of learning, Liao, Huang, Chen, and Huang (2015) and Tan, Ooi, Leong, and Lin (2014) showed that students with more innovative have greater intentions to adopt new technologies in their learning and pedagogy activities. Hence, in the context of MW2.0L we can posit that:

Hypothesis 8 (H8): PIIT positively influences students’ behavioral intention to continue use of MW2.0L.

Attitude

AT is referred to as the degree to which an individual assesses a behavior as favorable or unfavorable (Fishbein & Ajzen, 1975). The relationship among AT and intention developed in TPB model postulates that AT plays an evaluative predisposition role to behavior (Ajzen, 1985). Previous studies on technology adoption vastly reported that AT is one the most important determinants of intention to adopt technology (Zhou, 2016). For instance, in the context of e-learning adoption, Ratna and Mehra (2015) showed that favorable AT of students towards e-learning would eventually lead them to the actual use of e-learning. In the context of MOOC, Zhou (2016) reported that AT was the great determinant of students’ intention to use. In the context of mobile SNSs adoption, AT found as the significant determinant of users’ behavior toward the actual use of mobile SNSs (Y. W. Ha et al., 2015). Therefore, we can hypothesize that the more favorable is students’ AT toward MW2.0L, the more is their behavioral intention to continue use:

Hypothesis 9 (H9): AT positively influences students’ behavioral intention to continue use of MW2.0L.

Perceived Behavioral Control

Lee (2009) reported that individuals may not found themselves qualified in using the technology even when they have positive AT toward it. The author further argued that the intention to use the technology may increase when one perceive that he or she has control over its use. The relationship among PBC and intention received extensive empirical support in previous studies in the literature (Cheon, Lee, Crooks, & Song, 2012; Chu & Chen, 2016; Zhou, 2016). Previous studies reported that PBC would influence intention to adopt technology directly or through AT (Baker, Al-Gahtani, & Hubona, 2007). In other words, PBC suggests that individual’s evaluation capabilities and resources play an important role in predicting intention. In the context of e-learning, Ndubisi (2004) showed that PBC was significant in determining students’ intention to adopt. However, an insignificant relationship of PBC and intention was reported by Sawang, Sun, and Salim (2014). Accordingly, since incongruent impact of behavioral control on intention is reported in previous studies, further investigation of this factor’s impact on intention to adopt MW2.0L is needed. Hence, the following hypothesis is developed:

Hypothesis 10 (H10): PBC positively influences students’ behavioral intention to continue use of MW2.0L.

Subjective Norms

SN is introduced as one of the important determinants of behavioral intention by Ajzen (1985). However, mixed results are reported in technology adoption literature regarding the impact of SN on intention. While some studies revealed significant positive influence of subjective norm on intention (S. Lin, Zimmer, & Lee, 2013), other studies reported insignificant relationship among these two factors (Teo, 2011). For example, stronger subjective norm was found to be important in motivating students’ behavioral intention toward the use of e-learning (Baker et al., 2007; Sawang et al., 2014), while in the study by Teo (2011), the author reported that the impact of subjective norm on teachers’ intention to adopt new technologies in their classes was insignificant.

Other studies investigated the relationship among SN and intention in contingency models. For instance, Srite (2006) suggested that factors such as culture may moderate their relationship. The author reported that in more collective culture, like China, where people care more about interpersonal relationships, the impact of SN was stronger on the intention. However, in individualistic cultures, the impact was weak (Srite, 2006). Moreover, Cheung and Vogel (2013) found that not all components of SN are going to influence students’ behavioral intention to adopt e-learning. They have found that influences from peers have greater impact on students’ intention rather than the norms from media and teachers. Since incongruent conclusions are reported from previous studies regarding the impact of SN on intention, further clarification is needed about this relationship. Accordingly, the following hypothesis can be postulated:

Hypothesis 11 (H11): SN positively influences students’ behavioral intention to continue use of MW2.0L.

Task-Technology Fit

The utilization in this study is considered as students’ behavioral intention to continue use of MW2.0L. Previous studies showed that TTF substantially influences the intention of individuals to adopt new innovations (H.-P. Lu & Yang, 2014). In this study, TTF is defined as the extent to which the technological functions provided by MW2.0 tools and services assist students in achieving their learning objectives. TPC model asserts that the utilization of the technology depends on the fit between the technology and the task it supports. In the context of virtual learning environment for students and instructors, researchers showed that there are positive links between TTF and students’ grades (T. J. McGill & Hobbs, 2008), and instructors’ teaching skills (McGill, Klobas, & Renzi, 2011). In the study by Raven, Leeds, and Park (2010), the authors investigated the use of digital video tools supporting oral presentations within classes through the lens of TPC model and they found that these tools significantly improved presentation skills of students.

Most of university students have smartphones with MW2.0 tools installed on their devices which support communication, collaboration, e-learning, online lecturing, and other learning-related services. As MW2.0 tools and services increasingly support activities and tasks related to pedagogical purposes, users will be more likely to have strong motivations to continue use of such tools and services (Yi et al., 2016). Hence, we can assume that students’ use of MW2.0L tends toward mandatory.

Accordingly, following the earlier discussion and based on previous studies (T. McGill et al., 2011; T. J.McGill & Hobbs, 2008; T. J. McGill & Klobas, 2009; Yi et al., 2016), the following hypothesis is developed:

Hypothesis 12 (H12): TTF positively influences students’ behavioral intention to continue use of MW2.0L.

Continuance Intention to Use MW2.0L

Delone and McLean (2003) argued that although the adoption models like TAM had sidelined the outcomes for system use, successful model of information system indicates the direct link between net benefit, which is a variable related to performance outcome, and system usage. Petter and McLean (2009) claimed that this relationship has been vastly tested and validated in literature. Related to the use of technology for learning purposes, this relationship shows students’ academic performance will be improved when they take part in online activities (Mohammadyari & Singh, 2015; Padilla-MeléNdez, Del Aguila-Obra, & Garrido-Moreno, 2013; Soffer & Yaron, 2017). Northrup (2001) posited even though there is a physical separation between learners and educators in the e-learning environment, this learning environment could significantly increase students’ engagement, and this heightened engagement could improve learning outcomes as well as critical thinking skills and problem solving. In the context of technology adoption for learning purposes, several studies examined the continuance intention as the dependent variable (R.-B. Wang & Du, 2014; Wu & Chen, 2016), while Cheng (2011) proposed that the adoption of a web-based learning system could positively affect perceived impacts on performance. Online learning environment allows the learners to control the pace of learning, study autonomously, and discover things independently (Islam, 2016). Consequently, related to this study’s context, the use of MW2.0L would provide students self-directed learning opportunities and in turn, increased learning effectiveness. However, according to Islam (2013), insignificant relationship of technology usage for learning purposes and students’ performance has been reported by some studies. Such opposing findings create a need to further scrutinize this relationship related to MW2.0L usage and its relationship with students’ PL. Hence, the following hypothesis can be developed:

Hypothesis 13 (H13): MW2.0L continuance intention positively influences students’ PL.

Measurement

Almost all the contemporary social science researches either use the existing measurement items in the literature that perform well or modify the existing ones (Ramirez, David, & Brusco, 2013). The same approach was followed in this study, and measurement scales were adopted from previously validated sources. Accordingly, items to measure gratifications of technological convenience, IE, SI, and RC were adapted from the studies by Leung and Wei (2000), Cheng et al. (2015), and Y. W. Ha et al. (2015). AT toward MW2.0L, PP, PUS, and PEU were retrieved from E. Park et al. (2014) and further refined to fit the context of current study. To measure PIIT, the current study utilized original measurement items retrieved from the study by J. Lu et al. (2005). TTF was measured utilizing the items retrieved from the study by H.-P. Lu and Yang (2014). PBC and SN constructs were measured using the items retrieved from Cheon et al. (2012) who conducted a study on m-learning adoption from the lens of TPB. CITU MW2.0L was measured utilizing the items retrieved from the study by Wu and Chen (2016) on the use of MOOC which further were refined to fit the context of this study. Finally, students’ PL of MW2.0L was measured utilizing the items retrieved from the study by Soffer and Yaron (2017).

The content validity of the instrument was assessed by asking experts in the field of technology adoption for learning purposes selected from academia and practice who possess relevant experience and qualifications to justify the questionnaire, ¹ particularly in relation to the elements of each concept.

Sample and Data Collection

The nonprobability sampling procedure has been selected for this study for following reasons. First, as long as the target population to conduct this research was students within Malaysian public universities with MW2.0L experience, the number of elements in this population is unknown and cannot individually be identified. Second, access to all public university students in Malaysia was not promising since such process would be costly and time consuming and the location of universities is dispersed in Malaysia.

To decide the minimum sample size, we utilized the G*Power software with the settings of 12 predictors, 0.15 as the effect size, 0.5% as the error Type I (α), and the power of 0.80. The minimum required sample size was equal to 127. Therefore, it was decided that data collection for the purpose of this study would be greater than the identified minimum required number.

Accordingly, a structured questionnaire derived from the literature was designed and distributed among the students of top five public universities in Malaysia. Utilizing the intercept survey method, 500 questionnaires were distributed among the students. Since the purposive sampling method was found suitable to this study, experienced respondents in using MW2.0L were filtered only by asking them before the questionnaire was administered. After around 3 months, a total of 485 completed questionnaires were received. By reviewing the received questionnaires, 28 of them were dropped because of missing responses in the main variables or the same responses for all the items. Finally, there were 456 usable completed questionnaires in hand and ready to further analysis.

Data Analysis

To analyze the data and examine the research hypotheses, we have adopted the partial least square (PLS)-based structural equation modeling (SEM. According to Hair, Ringle, and Sarstedt (2011), PLS-SEM is causal modeling approach which focuses on “maximizing the explained variance of the dependent latent constructs” (p. 139). This study used PLS-SEM for some reasons: First, this study aimed to identifying the determinants of intention to adopt mobile SNSs for learning purposes. This is consistent with the rules of thumb proposed by Hair et al. (2011). PLS is an appropriate technique when the goal is predicting the key target constructs or identifying the key antecedent constructs. Likewise, the proposed research model in this study included independent variables (i.e., PULS, PEOU, PTTF, and PENJ) as well as moderating variable (i.e., mobile SNSs experience). Furthermore, PLS is suitable when one dependent variable (i.e., PULS) becomes an independent variable in subsequent relationships (Gholami, Sulaiman, Ramayah, & Molla, 2013). Further, PLS has an advantage over regression in which it analyzes the entire model including the moderating effects as a unit, rather than dividing it into pieces (D. L. Goodhue, Lewis, & Thompson, 2012), and appropriate for analyzing complex models (Hair et al., 2011).

The SmartPLS M2 Version 2 has been used to analyze the data in this study following a two-step analysis approach. Two models are used in PLS analysis: (a) measurement model that relates indicator items to their relevant latent constructs and (b) structural model which relates different latent constructs to each other (Hair, Ringle, & Sarstedt, 2013). Construct is an unobserved variable which is measured by some observed variables known as indicators or measurement items. To perform the two-step analysis, we followed the stages recommended by Hair et al. (2013).

Sample Profile

Table 1 exhibits the profile of respondents, in which 34.65% of respondents were male, and 65.35% were female in which this ratio is usually observed in Malaysian public universities. Most respondents were master students (43.86%), aged between 21 and 25 years (48.46%) and from Universiti Teknologi Malaysia university (32.46%). All the respondents indicated that they have used Facebook for their formal and informal learning purposes. WhatsApp (84.87%) and YouTube (68.64%) were among the other mostly used MW2.0 applications utilized for learning purposes. As exhibited in Table 1, 77% respondents indicated that they have more than 3 years of experience in using MW2.0 tools for their learning and pedagogy purposes. More than 90% of respondents indicated that their main purpose of using MW2.0 applications for their learning purposes was communication with peers, lecturers, or research supervisors. Sharing information and homework- or assignment-related tasks were the two other motivations of using MW2.0 tools among students for their learning purposes.

OPEN IN VIEWER

Table 1.

Sample Profile.

	Frequency	Percentage	Cumulative
Gender
Male	158	34.65	34.65
Female	298	65.35	100
Age
Under 20 years old	135	29.61	29.61
21–25	221	48.46	78.07
26–30	54	11.84	89.91
31–40	28	6.14	96.05
Above 40 years old	18	3.95	100
Academic qualification
Bachelor	100	21.93	100
Master	200	43.86	78.07
PhD and above	156	34.21	34.21
University
Universiti Malaya	57	12.5	12.5
Universiti Putra Malaya	91	19.96	32.46
Universiti Teknologi Malaysia	148	32.46	64.92
Universiti Kebangsaan Malaysia	80	17.54	82.46
Universiti Sains Malaysia	80	17.54	100
Mobile Web 2.0 application
Facebook	456	100
WhatsApp	387	84.87
YouTube	313	68.64
Twitter	33	7.24
Instagram	43	9.43
LinkedIn	97	21.27
Other	78	17.11
Experience
Less than 1 year	6	1.32	1.32
1–2	23	5.04	6.36
2–3	76	16.67	23.03
More than 3 years	351	76.97	100
Purpose of using mobile Web 2.0 learning
Share information with classmates  or interest groups	325	71
Educational discussions	166	36
Homework and assignments	242	53
Study groups	141	31
Communication with students, lecturers,  or supervisors	414	91
Research process	31	7
Other	27	6

Furthermore, in the questionnaire, respondents were asked to indicate to what extent MW2.0L helped them in contributing to a learning environment or achieving their learning goals. These questions were followed with a text box in that they were asked to comment on their responses. Figures 3 and 4 illustrate the responses to these two questions. OPEN IN VIEWER

OPEN IN VIEWER

Figure 4.

Results of the structural model.

As shown in Figures 3 and 4, it was obvious that the majority of participants were considerably positive toward the application of MW2.0L both for achieving their learning goals and contributing to learning environments. Moreover, further investigation in the comments provided by the respondents under each question revealed that they were positive about MW2.0L in very specific applications of such tools such as communication or graded homework or assignments.

Common Method Bias

Cross-sectional nature of the current study suggests the probability of existing common method bias (CMB). Accordingly, Harman’s one-factor test (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) and marker variable technique (Lindell & Whitney, 2001) were applied to assess CMB of the current study. The results of the analysis showed that the largest variance which was explained by a single factor was 27.6% indicting that the majority of the variance was not accounted by one factor (Podsakoff et al., 2003). To perform the marker variable technique, we have assigned all the items as the indicators of a new common method factor and further reestimated the model. The results on analysis revealed that there was no positive correlation among the common method factor (i.e. marker variable) and the variables of the research model. Hence, CMB issues were not detected in the present study.

Measurement Model Assessment

As suggested by the Hair et al. (2013), the psychometric properties of reflective measures were evaluated for their internal consistency, indicator reliability, convergent validity, and discriminant validity. Internal consistency was estimated using composite reliability (CR) and Cronbach’s alpha. Indicator reliability was assessed in term of item loading, while convergent validity was evaluated using the value of average variance extracted (AVE). Fornell-Larker criterion together with the evaluation of cross-factor loading ² was used to assess the discriminant validity. Tables 2 and 3 exhibit the validity and reliability assessments of constructs.

OPEN IN VIEWER

Table 2.

Factor Loadings and Reliability.

Construct	Item	OL	Cronbach’s α	CR	AVE
Attitude	AT01	0.905	0.871	0.921	0.795
	AT02	0.883
	AT03	0.886
Continuance intention to use	CITU1	0.891	0.865	0.917	0.787
	CITU2	0.889
	CITU3	0.881
Information exchange	IE01	0.903	0.890	0.932	0.819
	IE02	0.907
	IE03	0.905
Perceived behavioral control	PBC1	0.910	0.891	0.932	0.821
	PBC2	0.903
	PBC3	0.905
Perceived ease of use	PEU1	0.908	0.897	0.936	0.829
	PEU2	0.908
	PEU3	0.916
Personal innovativeness in information technology	PIIT1	0.890	0.876	0.924	0.802
	PIIT2	0.905
	PIIT3	0.891
Perceived learning	PL01	0.891	0.879	0.925	0.805
	PL02	0.900
	PL03	0.900
Perceived playfulness	PP01	0.892	0.864	0.917	0.787
	PP02	0.884
	PP03	0.885
Perceived usefulness	PUS1	0.907	0.892	0.933	0.822
	PUS2	0.907
	PUS3	0.906
Recreations	RC01	0.850	0.882	0.919	0.738
	RC02	0.860
	RC03	0.862
	RC04	0.864
Social interaction	SI01	0.895	0.917	0.942	0.801
	SI02	0.890
	SI03	0.905
	SI04	0.891
Subjective norms	SN01	0.904	0.892	0.933	0.822
	SN02	0.909
	SN03	0.906
Technological convenience	TC01	0.907	0.889	0.931	0.818
	TC02	0.900
	TC03	0.906
Task-technology fit	TTF1	0.900	0.883	0.927	0.810
	TTF2	0.893
	TTF3	0.907

OPEN IN VIEWER

Table 3.

Discriminant Validity According to Fornell-Larcker Criterion.

	AT	CITU	IE	PBC	PEU	PIIT	PL	PP	PUS	RC	SI	SN	TC	TTF
AT	0.891
CITU	0.866	0.887
IE	0.836	0.871	0.905
PBC	0.801	0.854	0.798	0.906
PEU	0.818	0.864	0.856	0.802	0.911
PIIT	0.786	0.755	0.747	0.723	0.724	0.896
PL	0.644	0.677	0.642	0.595	0.628	0.591	0.897
PP	0.808	0.780	0.776	0.749	0.753	0.753	0.577	0.887
PUS	0.803	0.873	0.833	0.830	0.835	0.710	0.607	0.737	0.907
RC	0.720	0.696	0.699	0.672	0.665	0.679	0.587	0.659	0.648	0.859
SI	0.845	0.876	0.841	0.841	0.847	0.740	0.629	0.769	0.862	0.689	0.895
SN	0.812	0.862	0.811	0.825	0.822	0.717	0.592	0.743	0.864	0.663	0.848	0.907
TC	0.844	0.875	0.845	0.852	0.822	0.745	0.611	0.783	0.847	0.687	0.869	0.849	0.905
TTF	0.790	0.853	0.809	0.831	0.826	0.699	0.583	0.731	0.851	0.617	0.802	0.849	0.821	0.900

Note. AT = attitude; CITU = continuance intention to use; IE = information exchange; PBC = perceived behavioral control; PEU = perceived ease of use; PIIT = personal innovativeness in information technology; PL = perceived learning; PP = perceived playfulness; PUS = perceived usefulness; RC = recreation; SI = social interaction; SN = subjective norm; TC = technological convenience; TTF = task-technology fit.

As exhibited in Table 2, the outer loadings of all indicators met the threshold of 0.7. The values of Cronbach’s alpha and CR of all reflective constructs are well above the acceptable threshold of 0.7. Furthermore, all the AVE values exceeded the threshold of 0.5. Hence, the measurement model reflected the valid and reliable indicator reliability, internal consistency, and convergent validity regarding the values of their outer loadings, Cronbach’s alpha (and CR), and AVE, respectively.

Table 3 exhibits the Fornell-Larcker criterion evaluation, in which the square roots AVE of each construct should be greater than its correlation with other constructs. The measurement model showed well-established discriminant validity as both criteria of cross-loading assessment and Fornell-Larcker were met.

The results of analysis showed that there is no issue regarding the measurement model of the study related to its validity and reliability, and hence, the collected data can further be utilized to assess the structural model. The structural assessment of the research model is provided in the following section.

Structural Model Assessment

Following the examination of the measurement model, the structural model was tested by estimating the paths between the independent constructs and the intention. To assess the structural model, there are several criteria to be reported as suggested by Hair et al. (2013). To begin with, the structural model should be assessed against collinearity issues by examining the values of variance inflation factor (VIF). Then, the significance and relevance of the structural model should be assessed by means of path coefficient values and standard errors (t values) which are obtained through the application of bootstrapping technique. Afterwards, the level of R-square (R²) needs to be reported, which reflects the amount of variance of the dependent construct which is explained by its predictor constructs in the structural model (Hair et al., 2011). The value of R² ranges from 0 to 1, and the higher values indicate the better prediction capability of the model via PLS path modeling.

To begin with, the structural model should be assessed against collinearity issues by examining the values of VIF and the level of tolerance. Statistical software package, IBM SPSS, was utilized to perform collinearity assessment. As presented in Table 4, the values VIF passed the thresholds (less than 5). Hence, it can be concluded that the all the constructs included in the proposed model aiming to predict the CITU and PL were not correlated and consequently no construct is needed to be eliminated from the model.

OPEN IN VIEWER

Table 4.

Collinearity Assessment for the Structural Model.

First set DV = Continuance intention to use		Second set DV = Perceived learning
Construct	VIF	Construct	VIF
Attitude	2.139	Continuance intention to use	1.000
Information exchange	1.899
Perceived behavioral control	2.328
Perceived ease of use	1.112
Personal innovativeness in information technology	1.879
Perceived learning	1.999
Perceived playfulness	2.658
Perceived usefulness	2.318
Recreations	3.014
Social interaction	1.879
Subjective norms	2.005
Technological convenience	1.699
Task-technology fit	3.211

The second step after checking the structural model regarding the collinearity issues, focused on the assessment of path coefficients that are the result of the PLS algorithm. Following the recommendation of Hair et al. (2013), the bootstrapping approach was handled using mean replacement as the missing values algorithm and a 5,000 as the number of resamples. Figure 3 and Table 5 show the significance testing results of the structural model path coefficients.

OPEN IN VIEWER

Table 5.

Summary of the Structural Model.

Hypothesis	Relationship	R 2	β	t	p	Result
H1	TC → CITU		0.093	2.284	.011	Supported
H2	IE → CITU		0.142	3.679	.000	Supported
H3	SI → CITU		0.098	2.696	.004	Supported
H4	RC → CITU		0.011	0.366	.357	Not supported
H5	PP → CITU		0.000	0.003	.499	Not supported
H6	PEU → CITU		0.106	2.985	.002	Supported
H7	PUS → CITU		0.124	3.161	.001	Supported
H8	PIIT → CITU		0.011	0.325	.373	Not supported
H9	AT → CITU		0.161	3.680	.000	Supported
H10	PBC → CITU		0.099	2.637	.004	Supported
H11	SN → CITU		0.080	2.231	.013	Supported
H12	TTF → CITU	0.883	0.099	2.954	.002	Supported
H13	CITU → PL	0.458	0.677	18.705	.000	Supported

The R² values of CITU and PL which are the two dependent constructs of this study were 0.883 and 0.458 which are relatively substantial and weak, respectively. As shown in Figure 3, all the path coefficients were significant (with the p values less than .05, .01, and .001) except the path coefficients of the relationships of RC gratifications → CITU, PP → CITU, and PIIT → CITU. Having obtained these results, the research hypotheses were examined.

Hypothesis 1 states that technological convenience as one of the gratifications of MW2.0L sought by students influences their intention to continue use of these tools for their learning purposes. From Figure 3 and the results exhibited in Table 5, it can be observed that there is a positive and significant impact of technological convenience on CITU (β = 0.093, t = 2.284, p < .05). The results revealed that technological convenience was statistically related to CITU, and hence, Hypothesis 1 was accepted. Hypothesis 2 states that IE as another gratification of MW2.0L influences the continuance intention of students to use MW2.0L. Related parameters are significant (β = 0.142, t = 3.679, p < .001), and thus Hypothesis 2 is supported. Hypothesis 3 suggests that SI is another MW2.0 gratification sought by students which influences their intention to continue these tools for their learning purposes. The results of the analysis revealed that Hypothesis 3 with the path coefficient of 0.098 is significantly supported (t = 2.696, p < .01). Hence, Hypothesis 3 is accepted. Hypothesis 4 posits that RC measured by entertainment and fashion or status is another factor which motivates students to continue use of MW2.0L. However, the results of the analysis indicated that while it had positive influence on the CITU, its impact was statistically insignificant (β = 0.011, t = 0.366, p > .05), and hence, Hypothesis 4 was rejected. Hypothesis 5 posits that PP of MW2.0L motivates students to continue use of these tools. However, results of analysis (β = 0.000, t = 0.003, p > .05) revealed that the influence of PP on CITU was insignificant. Hence, Hypothesis 5 was rejected. The influence of PEU of MW2.0L on its CITU was investigated in Hypothesis 6. The results of data analysis (β = 0.106, t = 2.985, p < .01) revealed that PEU was significantly impacting CITU. Hypothesis 7 postulates that students’ beliefs toward the usefulness of MW2.0L would influence their intention to continue use of these tools for learning purposes. Statistical analysis of collected data (β = 0.124, t = 3.161, p < .001) showed that CITU was significantly predicted by PUS. Hypothesis 8 investigated the impact of personal innovativeness in IT on continued use of MW2.0L. However, the results (β = 0.011, t = 0.325, p > .05) revealed that this construct had no significant effect on students’ intention to continue use of MW2.0L. Hypothesis 9 posited that students’ AT toward MW2.0L is another important factor in explaining their intention to use. Results of the data analysis (β = 0.161, t = 3.680, p < .001) showed that AT was positively and significantly impacting students’ CITU MW2.0L. Accordingly, Hypothesis 9 was accepted. Hypothesis 10 postulated that PBC is positively impacting students’ CITU MW2.0L. The results of data analysis showed that the path coefficient (β = 0.099) of the relationship perceived behavior control → CITU was statistically significant with the t value of 2.638 (p < .01); hence, Hypothesis 10 was accepted. The impact of SN on students’ intention to continue use of MW2.0L was developed in Hypothesis 11. The results (β = 0.080, t value = 2.231, p < .05) indicated that SN impacts CITU positively and significantly. Hence, Hypothesis 11 is accepted. Hypothesis 12 postulates that students’ perceived fit between their learning activities (task) and MW2.0 tools (technology) positively impact their decision to continue use of MW2.0L. Data analysis revealed that the relationship TTF → CITU with the path-coefficient of 0.099 was statistically significant (t value = 2.954, p < .01), and hence Hypothesis 12 was accepted. PL of students regarding the use of MW2.0L was investigated in Hypothesis 13 which posits that students’ CITU MW2.0L is positively explaining their PL. Results showed that, the relationship of CITU → PL was statistically significant (β = 0.677, t value = 18.705, p < .001), and hence, the hypothesis was accepted.

As t statistic produced in SmartPLS is easily inflated with large sample sizes, the path coefficients with values of less than 0.10 can be statistically significant. As occurred in this study with the sample size of 456, some relationships were found significant while having path coefficient less than 0.10. For example, constructs of technological convenience and SN with path coefficients of 0.093 and 0.080 were found statistically significant while they are not practically significant. In this case, f-square (f²) is measured to demonstrate the actual strength of the effect of each construct. In the following section, f² of the accepted hypotheses in the current research model is calculated and interpreted.

Assessment of f² Effect Sizes

Effect size is used to measure the relative impact of an individual independent construct on its dependent one (Hair et al., 2013). It measures the extent to which the R² value changes when a certain construct is omitted from the model, and hence, the substantial impact of the omitted construct on explaining the model is reflected. Hence, the substantive impact of the omitted construct on the dependent variable can be interpreted by the value of f² effect size. To this end, the f² effect size test was used (Cohen, 1988). The effect size is calculated as:

f 2 = R included 2 - R excluded 2 1 - R included 2

where R included 2 and R excluded 2 are the R² values of the research model after analysis when the target independent variable is included in or excluded from the model. Table 6 exhibits the results f² effect size related to the accepted hypotheses of the research model in this study. The test is performed after removing the rejected hypotheses from the model (i.e., RC gratifications, PP, PIIT).

OPEN IN VIEWER

Table 6.

Values of f² Effect Size of Latent Variables.

Construct	R included 2	R excluded 2	f2 effect size	Category
Technological convenience	0.883	0.881	0.017	No effect
Information exchange	0.883	0.879	0.034	Small
Social interaction	0.883	0.881	0.017	No effect
Perceived ease of use	0.883	0.881	0.017	No effect
Perceived usefulness	0.883	0.880	0.026	Small
Attitude	0.883	0.877	0.051	Small
Perceived behavioral control	0.883	0.881	0.017	No effect
Subjective norms	0.883	0.882	0.009	No effect
Task-technology fit	0.883	0.881	0.017	No effect

Based on the categorizing approach suggested by Cohen (1988) and further recommended by Hair et al. (2013), the results of effect size analyses presented in Table 6 reflect that among the factors influencing the CITU MW2.0L, IE, PUS, and AT had substantives effects on the dependent variable. Technological convenience, SI, PEU, PBC, SN, and TTF were found to have no effects on CITU regarding the values of their f² effect size.

Based on the results presented in Table 6, AT had the highest effect size on explaining the intention to continue use of MW2.0L, which can be interpreted as a small effect. These results suggest that the majority of the variance of the dependent construct (i.e., CITU) was explained by the combination of all the independent constructs available in the model rather than their independent contribution.

Technological Convenience

According to the results of this study, technological convenience of MW2.0L influenced significantly students’ behavioral intention to continue use of these tools for their learning purposes. Hence, accessing to information anytime and anywhere is the motivation behind the use of MW2.0L. Finding of this study is in line with other studies in the literature in the context of mobile social media adoption (Cheng et al., 2015; Y. W. Ha et al., 2015) and mobile learning studies (Iqbal & Bhatti, 2017; Sabah, 2016). For example, in the study conducted by Cheng et al. (2015), authors reported accessibility as a unique motivation behind the use of social network sites on mobile devices.

While technological convenience was found significant in explaining students’ CITU behavior, the results of this study exhibited in Table 5 suggest that it is not the strongest predictor of intention behavior compared with other factors. Same results reported by Wei and Lo (2006) on the use of cell phone use among the users, where mobility was found to have the least variance in explaining usage behavior among other motives. Cheng et al. (2015) also reported that while accessibility of mobile devices is important in motivating users to use social media tools through mobile devices, but other gratifications such as cognitive needs and affection needs were found more significant than this motive. Hence, it can be explained that mobility and accessibility of mobile devices complements other motives of using MW2.0L by students.

Information Exchange

IE (i.e., cognitive needs) in the context of MW2.0L refers to the students’ use of MW2.0 tools to “seek for information in order to be critical and creative thinker” (Hashim, Tan, & Rashid, 2015, p. 5). IE motive of MW2.0L implies that, to fulfill cognitive needs, these tools should have the capability to motivate learners to seek and exchange information-knowledge which is related to their learning purposes. Cognitive needs of learners are fulfilled by the collaboration and interaction with other learners and instructors through social media tools. Accordingly, information-related motives which are related to enhance the knowledge and awareness of students in their courses are considered as another important determinant of CITU behavior.

Utilizing MW2.0L is different with conventional classroom forums available in learning management systems in that seeking information and exchanging information and knowledge is quite simple and easy in these tools. In the study conducted by Leung and Zhang (2016), authors reported that heavy users of tablets mostly valued information-seeking gratification. It was natural since users who heavily use tablets to seek and exchange information valued tablets’ big screen and easy to use. Sample explanation can be applied to MW2.0L as well. Students who valued IE gratification of MW2.0L found these tools as mediums that can easily collaborate and interact with other students and simply seek required information and exchange valuable knowledge. Seeking and exchanging information through MW2.0L vary according to students’ cognitions need, their perceptions toward MW2.0L and further their SIs (Asghar, 2015).

Social Interaction

“Education is social practice” (Kim, Kwon, & Cho, 2011, p. 1513). Social constructivism suggests that effective and efficient learning can be provided through the SI among learners and instructors. The results of analysis revealed that SI is one of the significant motivations in which students are intend to continue use of MW2.0L. SI as the higher order construct measured by affection and recognition needs implies that the more students participate in MW2.0L activities, the more they tend to be socially present and be acknowledged among others (Kim et al., 2011).

Following the study conducted by Hamid, Waycott, Kurnia, and Chang (2015), SI through MW2.0L as a gratification sought by students can be mapped to three forms of interactions that learning through MW2.0 tools offer. First type of interaction occurs among students themselves (i.e., student-student interaction). Despite the student-student interaction opportunity provided by MW2.0L, some cautions are needed to be considered as well. For example, due to the complexity of interactions through MW2.0 tools, identity and privacy of users interacting through such mediums are high concerns (Kimmons, 2014). Hence, designers and developers of MW2.0 tools should apply such consideration in their development. Another consideration concerns the appropriate behavior of students’ interaction through MW2.0 tools which is known as “netiquette.” Waycott, Sheard, Thompson, and & Clerehan (2013) suggested that lecturers are needed to control and moderate students’ online interactions to ensure that such behaviors are not going to be occurred.

Second type of interaction provided by MW2.0L focuses on student–lecturer interactions. As suggested by Hamid et al. (2015), learning through social media tools enhanced the interaction among students and their teachers in the context of Malaysia. While it is considered beneficial to interact with lecturers through MW2.0 tools, it can rise workload challenges to lecturers when such interactions are happening outside of normal office working hours. While some studies found such type of interaction beneficial (e.g., Hamid, Waycott, Kurnia, & Chang, 2014; Hamid et al., 2015), other researchers found such informal interaction as a barrier to the adoption social media tools in higher education and pedagogical purposes (e.g., Schroeder, Minocha, & Schneider, 2010).

The third type of interaction that can be occurred by learning through MW2.0 tools is student–content interaction. Hamid et al. (2015) suggest that learning through social media tools would enhance students’ critical thinking ability and further may help their self-monitoring of learning progress.

Recreation

RC measured as a higher order construct in this study (measured by two variables of entertainment and fashion or status) was found to have an insignificant influence on students’ CITU MW2.0L. It was hypothesized that students would engage in MW2.0L for their leisure and amusement needs. However, it was found that students would continue to use such tools beyond recreational purposes. This finding correlates with the study conducted by N. Park, Kee, & Valenzuela (2009) on the gratifications of Facebook groups engagement by students.

One explanation for the rejection of this hypothesis would be, while the majority of the current study’s respondents are considered as heavy users of MW2.0L with more than 3 years of experience, they are not considering fashion or status as a motivation to continue use of MW2.0L. According to Leung and Zhang (2016), users who find this gratification as a motive to use a system are the ones who tend to be observed as trendy and fashionable. Heavy users who already have experience of using the system may have passed this stage which is to make a fashion statement. This argument is in accordance with the study conducted by Quan-Haase and Young (2010). They have reported that users who use instant messaging solely to be considered fashionable they are among the ones who tend to use these systems less.

Several studies reported social media tools entertaining for their users (Leung, 2013; Leung & Zhang, 2016). For example Leung (2013) reported that Net Geners find Facebook and blogs an appropriate place for entertainment, and Boomers reported forums interesting environment to start entertaining discussions. Accordingly, for future studies, the influence of entertainment on the CITU would be appropriate.

Perceived Playfulness

The results of this study revealed that PP has no direct effect on continuance intention of MW2.0L usage which contradicts with previous studies on e-learning (Padilla-MeléNdez et al., 2013), m-learning (e.g., Y. S. Wang, Wu, & Wang, 2009), and social media usage (e.g., Hung, Tsai, & Chou, 2016). The result reported in the current study which is in line with the study conducted by Iqbal and Qureshi (2012) suggests that students who intend to continue use of MW2.0L are not motivated by enjoyable and playfulness contents of these tools. One explanation would be, while MW2.0L is attractive, enjoyable, and playful, but as the majority of respondents are students with higher degrees of education (master and PhD), they care less about this characteristic of MW2.0L in their rational decision toward the continue of usage.

However, given the voluntary usage of MW2.0L and that target people are not only students with higher degrees of education but also from very diversified backgrounds of study with different levels of education, it would be crucial to make contents of MW2.0L playful and enjoyable to further attract more users to MW2.0L systems. Accordingly, MW2.0L designers are advised to refer to the framework proposed by J. Chung and Tan (2004) in developing playfulness of mobile learning systems.

Perceived Ease of Use

PEU was another factors impacting students’ CITU MW2.0L. While the impact was significant, its effect on behavioral intention was less significant than PUS as another construct of TAM. The result contradicts with the findings reported by previous studies conducted in the context of e-learning systems (e.g., Castañeda, Muñoz-Leiva, & Luque, 2007; K. M. Lin, 2011). However, there is one difference between such studies and the research conducted in this study. In studies which found greater impact of PEU on behavioral intention than PUS, respondents reported less experiences or even no experiences in using e-learning systems. Respondents in this study reported that they have used MW2.0 tools for their learning purposes for at least 3 years. Hence, they would automatically look for more complex attributes of such systems to be useful to their learning processes (usefulness).

Therefore, to enhance users’ perceptions toward the ease of use MW2.0L, developers are advised to put more efforts on the design of instructional contents which are considered as focal point in motivating students toward the acceptance of e-learning systems (K. M. Lin, 2011). Hence, it would be suggested that designing MW2.0 platforms toward to be easy to use and having user guides associated with such tools with acceptable degrees of quality and clarity are considered as important approaches to motive and retain less experienced users to continue use of MW2.0L.

Perceived Usefulness

PUS, which reflects the perceived assessment whether a particular technology enhances the job performance, was found as another important predictor of students’ behavioral intention to continued use. Consistent with other studies (Wu & Chen, 2016; Yeou, 2016) which found PUS as an important factor in deciding the adoption of learning-related technologies, this study also reported its importance in motivating students in using MW2.0L for their pedagogical activities. It can be interpreted that students who believe in use-performance relationship with MW2.0L, they may believe that their learning performance would improve, and hence, they are more inclined to use these applications.

Since the majority of the current study’s respondents were among the experienced users in using MW2.0 tools in learning activities, they already have deep perceptions regarding the usefulness of MW2.0L. However, for light users with less experiences, previous studies reported that users with less experiences of using an information system would perceive that system heuristically (Castañeda et al., 2007). It means such users are motivated to use a specific system according to attributes which are easy to evaluate (easy to use). Hence, related to the context of this study, when users have enough experiences in using MW2.0 tools for learning purposes, they would perform more systematic processes to evaluate the more complex specifications of MW2.0 tools (i.e., usefulness). Accordingly, the key mission to enhance students’ behavioral intention to continue use of MW2.0L would be apply strategies to improve their perceptions toward the usefulness of such tools for learning and pedagogy purposes.

Personal Innovativeness in Information Technology

Hypothesis 8 postulated that students’ innovativeness in IT would positively influence their intention toward continuance usage of MW2.0L. However, the relationship PIIT → CITU was found insignificant. Liu, Chen, Sun, Wible, and Kuo (2010) reported that students with innovative minds are more inclined to accept mobile learning. Agarwal and Prasad (1998) argued that innovative individuals are tend to look for relative advantages of the target technology and further its ease of use aspects, and hence, they show higher degrees of usage intention. Regarding the absence of the relationship PIIT → CITU which same result is also observed in previous studies as well (J. Lu et al., 2005), it can be implied that, since the majority of this study’s respondents are master level students, their decision to whether continue to use MW2.0L is grounded in their rationality rather than curiosity and braveness (J. Lu et al., 2005; Tan et al., 2014).

Since the result of this study regarding the influence of innovativeness on behavioral intention contradicts with what is reported in the literature (Crespo & del Bosque, 2008), one possible reason would be related to the personality trait of respondents in this study. Yesil and Sozbilir (2013) reported that innovative behavior is more obvious in individuals with the personality trait of openness to experience. Hence, individuals with characteristics such as curiosity, intelligence, and flexibility as the traits of openness to experience would perceive e-learning systems and technologies more useful to their learning processes. Hence, future studies are advised to take individuals’ personality traits into consideration as moderator variables to further seek knowledge regarding the impact of innovativeness on behavioral intention among students with different personalities.

Attitude

AT, which is referred as the one’s overall evaluation of performing a certain behavior, has already been reported as the most important factor in motivating individuals toward technology acceptance (Cheng et al., 2015; Soffer & Yaron, 2017; Wai et al., 2016). According to Ardies, De Maeyer, Gijbels, and van Keulen (2015), the factor that is used to measure the extent of users’ ambitious to use educational technologies is AT, and whether this technology has positive or negative impact on their learning outcome.

It is argued that, if there is a tendency in students to act positively toward a subject, for example, MW2.0L, then they would have more interest in that subject (Boser & Daugherty, 1998). Obviously, students are not going to continue use of MW2.0L unless they view how MW2.0L is supposed to enhance their learning and the outcomes resulting from them, positively. Furthermore, the analyses showed that exchanging information among the peers and lecturers and findings MW2.0 tools useful for learning and pedagogy activities are other motives to continue use of MW2.0L.

Thus, the challenge for HEIs to motivate their students to use MW2.0L is (a) to reinforce the positive AT of students who are already committed to adopt MW2.0L, (b) to change the AT of those students who view MW2.0L negatively, and (c) to provide interactive environments in MW2.0 among students and lecturers where they can find these tools useful to their learning purposes and an environment to exchange their information. Understanding AT toward MW2.0L would help in understanding its strengths and weaknesses, would assist to determine its readiness, and further would facilitate its required infrastructure.

Perceived Behavioral Control

PBC, as defined the extent to which students perceive the ease or difficulty of using MW2.0L, was found to have a significant effect on CITU which correlates with the findings of other studies in the literature in the context of online learning technologies (e.g., Cheon et al., 2012; Chu & Chen, 2016; Zhou, 2016). The significant impact of PBC on intention to use implies the importance of motivational implications (Chu & Chen, 2016). Hence, in the context of MW2.0L, if an individual perceives he or she has not enough capabilities or resources to use the system, even he or she has positive AT toward the system and has the support of important others, that person would be still unmotivated to use MW2.0L.

While the impact of PBC was supported in this study, mixed results are reported in the literature on the relationship between PBC and behavioral intention. For example, Nasco, Toledo, and Mykytyn (2008) conducted a study on the adoption of e-commerce among Chilean small- and medium-sized enterprises and they reported an insignificant correlation among PBC and adoption behavior. However, M. K. Chang (1998) reported that PBC is the strongest predictor of behavioral intention, Venkatesh, Morris, Davis, and Davis (2003) showed that PBC is significant only in some specific relationships. One possible explanation for the low impact of PBC on students’ continuance intention behavior can be explained by the cultural analysis study conducted by Hofstede (2003). The author argued that cultural difference may play important role in explaining PBC. Since Malaysians live in a culture which is less prone to risk taking, they may resist changes to their learning approaches through innovative technologies. Hence, further studies would be required to take culture construct into consideration for further investigation of the impact of PBC on intention toward continued usage.

Subjective Norms

The results of this study revealed that SNs were significantly influencing students’ continuance intention behavioral toward the use of MW2.0L. Following the definition of SN proposed by Agudo-Peregrina, Hernández-García, and Pascual-Miguel (2014) in the context of e-learning, the current study redefined SN as “the extent to which a student perceives a pressure from members in his or her environment to use MW2.0 tools for learning purposes” (p. 303).

In the context of this study, it can be argued that when students perceive that people (peers and lecturers) who are important to them think they should use MW2.0 tools and services for learning purposes, they consequently would incorporate important others’ beliefs to their own beliefs, and hence, they would perceive that MW2.0 tools are useful to their learning purposes (Abdullah & Ward, 2016).

One possible explanation regarding the low impact of SN on students’ intention to continue use of MW2.0L, as suggested by Hossain and de Silva (2009) would be the multidimensional nature of SN. In the context of e-learning systems such as MW2.0L, SN that influences students’ behavioral intention to continue use may stem from peers, lecturers, institution itself, or other external influences. Decomposition of SN correlates with the differentiation of strong and weak social ties found in online learning environments (Zhou, 2016). The study conducted by on the acceptance of Google Apps confirms this argument. In their study, they have reported that students’ peers had significant effect on their intention use, while instructors were found to have an insignificant effect. Hence, as SN is measured reflectively in this study, future studies are recommended to refine SN as decomposed variables and further study the influences of different sources on behavioral intention.

Task-Technology Fit

Another significant factor impacting students’ behavioral intention to continue use of MW2.0 tools for learning purposes was TTF. Finding of this study regarding the impact of TTF on behavioral intention correlates with findings of other studies reported in the literature in the context of educational technologies (e.g., D’Ambra et al., 2013; Wu & Chen, 2016). For example, Wu and Chen (2016) reported that students who find a fit between their learning task and the capabilities provided by MOOC would find these tools more useful to their learning purposes. The results of this study demonstrated that students who perceive MW2.0 to coincide with the needs of their learning activities getting done are more intent to use the functionalities available in MW2.0. This finding is consistent with results of other studies investigating the impact of perceived TTF on the utilization of technology for learning purposes. For example, in the study by Baleghi-Zadeh, Ayub, Mahmud, and Daud (2014), the authors reported that when the functionalities of learning management systems are fit with the task, students’ behavioral intention to use the system is increased.

Since the target respondents of this study are students solely which data were collected quantitatively, future studies can investigate perceptions of instructors and lecturers to find out what specifications are important to them to find MW2.0L fit to their task. Furthermore, qualitative methods such as interviews or focus groups would assist researchers to gain deeper understanding of users’ willingness to use such tools for learning purposes.

Continuance Intention to Use

In this study, we have hypothesized that MW2.0L CIT is positively influencing students’ PL. Such hypothesis is echoed in the TPC model of D. Goodhue and Thompson (1995) as they postulate that utilization of technology influences performance. Performance which in this study is reflected in PL was greatly explained by students’ MW2.0L CIT. This finding is in accordance with the results of previous studies on investigating the impact of IS continuance intention and performance. For example, Urbach, Smolnik, and Riempp (2010) reported that employees’ continuance intention of using portal would enhance their job performance and effectiveness. Regarding the context of this study, we can state that students’ learning activities via MW2.0 tools which assists them in transferring knowledge, sharing information, collaboration over subjects, and generally ubiquity properties of these tools maybe are the key issues that enhance learning of learners.

Implications for Research

Although many studies are conducted to investigate the adoption of e-learning, m-learning, and Web 2.0 learning in many countries, the author of this research argued that few studies explored the adoption of MW2.0L and specifically within developing countries including Malaysia. While there are some studies investigated the adoption and diffusion of mobile SNSs within Malaysian HEIs (e.g., C.-H. Wong et al., 2015), there were no comprehensive model in the literature to investigate the important determinant of MW2.0L adoption and acceptance by Malaysian students. The lack of empirical studies on the adoption of MW2.0L would resulted in the little knowledge and understanding of the adoption of such educational technologies. Hence, this research and its results contributed to the MW2.0L adoption knowledge and literature, by reviewing the literature related to e-learning, m-learning, and Web 2.0 learning conducted both in developed and developing countries, eliciting important factors in the adoption of such technologies, and further adapt such factors to the context of MW2.0L adoption within developing countries in general and to Malaysia in specific.

This study investigated continuance intentions of students in MW2.0L and further their PL outcomes. According to other researchers (T. D. Cochrane, 2014; Soffer & Yaron, 2017; C.-H. Wong et al., 2015), few studies investigated the continuance intention of MW2.0 usage and further its impact on learners’ PL. Accordingly, the following findings were contributed to the MW2.0L literature:

Significant determinants of MW2.L were identified;

This study investigated empirically the factors that would influence the students’ intention toward the use of MW2.0L in Malaysia as an example of developing country. Although, this study was conducted in the Malaysian education settings and culture, the results of this study would be useful to other developing countries with similar educational contexts and settings as well.

Implications for Practice

Learning process has already been changed within Malaysian public universities (Mohammad, Mamat, & Isa, 2012), however, policy-makers of HEIs should investigate and consider students’ usage and acceptance of MW2.0L prior to development and implementation process. Findings of this study exhibited high intention of Malaysian higher education students to continue use of MW2.0L which it can be served as an evidence to support such initiatives. However, the literature reported that the utilization of educational technologies is usually hindered by the lack of sufficient resources (Khan, Al-Shihi, Al-Khanjari, & Sarrab, 2015). Moving toward the increased use of MW2.0L requires reliable technological resources and infrastructure within universities such as wireless and high speed Internet. Such facilities and infrastructures are required to be facilitated by Ministry of Education Malaysia and further with the aid of Ministry of Communications and Multimedia Malaysia.

Universities are required to provide a well-resource mobile learning facility center. Within such center, students and lecturers who have little knowledge and experience in using MW2.0 tools and services for learning and pedagogy activities would be trained. This would be an opportunity for universities to motivate their students and lecturers to appreciate innovative technologies in formal and informal learning processes.

Faculties should incorporate MW2.0 tools and services in their courses. Such procedure requires faculty members to expand their communications with students out of classes to have a more collaborative learning environments. The approach that faculty members can take to increase the engagement of students in MW2.0L would be the implementation of learning strategies to help students to put their knowledge in practice. To do so, faculties may consult with their lecturers who are expert in educational technologies to alter their course materials and learning procedures to further alter their conventional teaching strategies.

Findings of this study showed that students of Malaysian public universities have high intention to use MW2.0L. Hence, developing mobile-friendly contents is crucial in the success of the acceptance of such systems. Performance expectancy in the architecture of MW2.0L should be emphasized by such designers while ensuring contents reliability and high-quality services. According to the results, while there is no matter regarding the ease of use of the system, MW2.0L would not be used if students find such tools not useful to their learning purposes. Furthermore, the application of this study’s results is not limited to students of HEIs. For example, in the context of health care in that mobile devices are used vastly in health-supported applications, designers and developers can consider the important factors found in this study in developing health-related MW2.0L tools. Accordingly, designer can focus on the factors of PUS and IE by increasing patients’ access to health care and health-related information, improving patients’ ability to diagnose and track diseases, and further expanding health educators’ access to ongoing medical trainings and educations, which in turn would increase their acceptance and continued usage of MW2.0L.