Constructs Influencing Students’ Behavioral Intention of Su- perstar Learning System

Superstar Learning System is an interactive information platform, where teachers and students can not only have an easy access to various learning resources, but also interact in the whole teaching and learning process. Although there have been a large number of studies devoted to the use of Superstar Learning System in education, very few of them explored its behavioral intention. In order to deal with this missing link in literature, we adopt a random sampling technique and a questionnaire survey to collect data. This study concludes that performance expectancy, effort expectancy, lecturer influence, peer influence, user innovativeness, interface simplicity, and multiple functions exert a positive influence on behavioral intention to use Superstar Learning System. We form a fit extended technology acceptance model (TAM) by including innovative constructs, i.e. lecturer and peer influences, user innovativeness, interface simplicity, and multiple functions. Interdisciplinary research is needed to explore an extended TAM to use Superstar Learning System in the future.


Origin of the Technology Acceptance Model
The Technology Acceptance Model (TAM) was proposed by Davis (1989), who adopted rational behavior theory to study users' acceptance of information systems. The original purpose of TAM was to explain the influencing constructs of computer technology acceptance. TAM proposes two major influencing constructs: perceived usefulness, which indicates the extent to which a user thinks that using a technology system improves the performance; and perceived ease of use, which means the degree to which a user thinks it is easy to use a technology system. Figure 1. The technology acceptance model (Davis et al., 1989) The Technology Acceptance Model (see Figure 1) considers that the use of technology is determined by behavioral intention, while behavioral intention is determined by the attitude toward the use and perceived usefulness. Perceived usefulness is determined by perceived ease of use and external variables. Perceived ease of use is determined by external variables. External variables include technology design characteristics, user characteristics (including perception forms and other personalities), task characteristics, nature of development or execution process, policy impact, and organizational structure, etc. They establish a connection between internal beliefs, attitudes, intentions, and differences among different individuals, environmental constraints, and controllable interference factors in the technology acceptance model.

TAM of Knowledge Management
System. An increasing number of organizations realize that organizational competitiveness depends on the effective management of intellectual resources, which makes knowledge management a very important organizational function (Davenport & Grover, 2001). Knowledge management includes a wide range of complex organizational, social and behavioral constructs. Nevertheless, information technology is still a major factor in the current study of knowledge management.
Based on the fact that knowledge management is supported by information-related technologies, the acceptance of knowledge management system is studied by using TAM.
This model mainly measures several main constructs of the technology acceptance model: perceived usefulness, ease of use, and user's intention to use knowledge management system, as well as their relationship with actual use.
Compared with Davis' original technology acceptance model, this research model did not consider the construct of attitudes to use, while Davis (1989) argued that attitudes to use partially mediated the effect of perceived usefulness on behavioral intention to use. External variables not involved, this research model does not include constructs that influence perceived usefulness and perceived ease of use.
TAM of Enterprise Resource Planning (ERP) application system. ERP is a system capable of handling multiple functions including finance, human resources, manufacturing, material management, sales, and distribution (Davenport, 2000). The implementation of ERP requires a lot of organizational resources and risks brought by a lot of investment. Compared with the implementation of a traditional simple information technology system, it is a completely different application area of information technology. In this research model, two constructs were included that influenced TAM, i.e. perceived usefulness and perceived ease of use (Lewis & Seibold, 1993). The model includes the main constructs of TAM and defines three external variables: the plan exchange of ERP system, the consensus of benefits generated by ERP system, and the training of ERP system . In the organization, the users who would most likely accept the ERP system were the senior managers. The plan exchange of the ERP system made the information about the ERP system flow from the senior managers to others (Caner, Jambulingham, & Gupta, et al., 2001). The consensus on the benefits of the ERP system (Mirani & Lederer, 1998) refers to the consensus among peers and managers on the value of the ERP system. The training of the ERP system (Potosky, 2002) includes internal and external training. TAM of the Internet Application. Many enterprises used the Internet, especially to share important information resources (Feher & Towell, 1997). The development of Internet-based systems and the establishment of intra-enterprise Internet can help break down the barriers of time and distance between suppliers and demanders in order to reduce costs and increase productivity. Enterprises use the Internet mainly to collect information, but how to obtain the expected information is a major concern (Soh, Mah, & Gan et al., 1997). Collecting task-related information has become a major aspect of using the Internet. More importantly, the performance of information processing was increasingly dependent on the matching degree between information and organizational tasks.
In this research model, TAM is used to evaluate individual performance. The assessment of the Internet use in daily work is mainly based on personal impressions of using task-related websites and Intranets. In order to empirically study the use of task-related Internet by staff, TAM and Choo's (1998) information behavior model are synthetically used. The information behavior model mainly explains how people reduce the uncertainty of tasks through the information demand-search-use cycle.
Based on the three-step cycle of information behavior model, the use of the Internet in work was explored from three aspects. In the aspect of information demand, users assessed whether the relevant information can solve the problem, and proposed related factors (Spink, Greisdof, & Bateman, 1998). In the information search stage, perceived usefulness, ease of use, and a personal factor-the state of imagination were proposed.
Revised TAMs. A more reliable TAM, i.e. the unified theory of acceptance and the use of technology (UTAUT) was formulated by Venkatesh et al. (2003), who aimed to integrate and analyze constructs in the TAM. In UTAUT there are three influencing constructs that determine behavioral intention, i.e. performance expectancy, effort expectancy, and social influence. One influencing construct, facilitating condition, is considered able to directly determine use behavior of mobile technology. This enhances the reliability of the acceptance model by extending the basic TAM. Characteristics of users, such as gender, age, experience, and voluntariness of use are deemed as significant moderating variables influencing behavioral intention and use behavior. It has been argued (Venkatesh et al., 2003) that UTAUT is significantly more reliable than the basic TAM in that the former is able to account for 70% of variances in the behavioral intention. Therefore, UTAUT provides a reliable model to test the usability of a novel mobile app (Ibrahim & Jaafar, 2011).
Later, many studies revised the basic TAM (e.g. Lee

Superstar Learning System
Superstar Learning System is an interactive information platform. Through this platform, teachers and students can not only have an easy access to various learning resources, interact in the whole teaching and learning process, but also retrieve relevant data and obtain timely feedback from either teachers or peers ( Figure 2). model. The teaching practice shows that this model combines the advantages of mobile teaching and traditional classroom teaching, which can not only improve the teaching quality of Integrated English courses, promote students' learning autonomy, but also meet the talent training objectives in the artificial intelligence era.

Constructs influencing behavioral intention
Although there have been a large number of studies devoted to the use of Superstar Learning System in education, very few of them explored its TAM. In order to deal with

Research questions and hypotheses
Therefore, we raised the research question: will performance expectancy, effort ex-

Theoretical framework
There are four core determinants in UTAUT: (1) Performance Expectancy (PE), (2) Effort Expectancy (EE), (3) Social Influence (SI), and (4) Facilitating Conditions (FC). PE refers to "the degree to which individuals feel that using systems is helpful to their work"; EE indicates "how much effort individuals need to make use of systems"; SI means "the degree to which individuals feel affected by the surrounding groups"; FC refers to "the degree of support that individuals feel for the use of systems in related technologies" (Venkatesh, 2003).
There are four moderators that have a significant impact on the above core determinants, i.e. gender, age, experience, and voluntariness of use. Venkatesh et al. (2003) also found that the combined use of more than two control variables would make the effect more significant on behavioral intention and use behavior. This study attempts to include new constructs, i.e. lecturer and peer influences, user innovativeness, interface simplicity, and multiple functions to test the TAM of Superstar Learning System, besides the explored constructs: performance expectancy, and effort expectancy.

Materials and Methods
This study adopts a random sampling technique to collect data in order to test proposed research hypotheses.

Participants
We randomly selected 281 tertiary students from Faculty of Business, Faculty of Western Languages, Faculty of Humanities & Social Sciences, Faculty of Middle Eastern Studies, and Faculty of Training in a state-owned university. They major in various disciplines, e.g. English Language, Linguistics, Business Administration, Journalism and Communication, International Relations, and International Politics. They, ranging from 19 to 26 years old, have normal literacy and psychological state based on their self-reports.

Research instruments
Research instruments used in this study were adapted from Abualaish & Love's peerreviewed article (2013). The questionnaire (Appendix A) included five adapted scales, aiming to determine five influencing constructs, i.e. performance expectancy, effort expectancy, lecturer influence, peer influence, and user innovativeness. Besides, we designed another two constructs that may influence the behavioral intention of Superstar Learning System, i.e. interface simplicity, and multiple functions. The questionnaire is internally consistent (α = 0.977) and externally valid (KMO=.943, P < .01) at the significance level .05.

A scale to measure performance expectancy
We used five-question items to determine performance expectancy such as "I find Superstar Learning System useful for my learning.", and "Using Superstar Learning System would enable me to achieve learning tasks more quickly." Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree. Five items were designed to measure effort expectancy, e.g. "I would find Superstar Learning System flexible and easy to use", and "Learning to operate Superstar Learning System does not require much effort". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure lecturer influence
We designed five items to determine lecturer influence, such as "I would use Superstar Learning System if it was recommended to me by my lecturers", and "I would like to use Superstar Learning System if my lecturers supported the use of it". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure peer influence
Peer influence was measured through five items, e.g. "Whether my classmates use Superstar Learning System influences my use of it", and "Whether my schoolmates use Superstar Learning System influences my use of it". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure user innovativeness
We identified user innovativeness through five items, e.g. "I like to experiment with new information technologies", and "When I hear about a new information technology I look forward to examining it". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure interface simplicity
The influencing construct "Interface simplicity" is determined by five items, e.g. "The interface of Superstar Learning System is easy to follow", and "The interface of Superstar Learning System is clearly organized". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure multiple functions
"Multiple functions" was identified via five items, e.g. "Superstar Learning System provides me with rich resources", and "Superstar Learning System provides me with many learning activities". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

A scale to measure behavioral intention
We also used five items to measure "behavioral intention", e.g. "I plan to use Superstar Learning System in my studies", and "I predict that I will use Superstar Learning System frequently". Each item is followed by a five-point Likert Scale, ranging from strongly agree, agree, neutral, disagree, to strongly disagree.

Superstar Learning System-assisted learning process
Superstar Learning System-assisted learning process is different from traditional learning without an aid of a mobile learning app. Superstar Learning System has the function of grouping students. It can set up learning groups according to the purpose of the study and the number of students. Generally, each group can contain 6 to 8 students. Teachers can distribute lecture materials to each group about one week before class. Each group member uses their spare time to consult, discuss and analyze the materials, and put Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 26 July 2021 doi:10.20944/preprints202107.0583.v1 forward plans within the prescribed time to prepare for classroom discussion. Classroom discussions are also conducted in divided groups on the mobile platform.
The contents of the solutions proposed by the groups may include learning experience and knowledge or explanation of related theoretical knowledge. Team members debate and answer questions raised by teachers and classmates. Teachers can give timely evaluation and feedback on the platform. In addition, the assessment of academic achievements includes the mutual evaluation of team members and the assessment of their classmates. This group-based learning model enables students to learn how to express their views clearly, how to deal with different views or opinions, and how to cooperate with each other and respect each other. It is conducive to strengthening students' team consciousness and enhancing their mutual cooperation and tolerance.
When examining the completion of students' learning tasks, teachers can adopt the way of "choosing persons" in the "activity" module. The random way of choosing persons to answer questions can urge students to prepare well before class. In the process of teaching, teachers can also arrange answers according to the key points of knowledge, and students can respond online. Of course, in the development of classroom teaching, teachers can also use the "sign-in" function in learning to allow students to complete the check-in within a specified time, to inspect the attendance of students, rather than using the traditional way to roll call names. It can not only save the classroom time, but also facilitate the records of each check-in situation on the platform, which is convenient for analysis of the performance statistics. It can also conduct a questionnaire survey on teaching methods, assessment methods, and personnel training through the "voting/questionnaire" module, as well as statistical analysis of obtained data.
Teachers can use modules such as "examination" and "homework" to assign corresponding questions to each course, or carry out chapter tests, give scores according to students' answers, and statistically analyze the results. Teachers can also reflect on teaching according to the results of the analysis, find out the problems existing in students, and further adjust the teaching plan, in order to implement and improve target teaching. In addition, the completion of each assignment and scores are recorded on the system platform for the final summary. It also saves the time of marking papers for teachers.
The discussion area in the course builds a platform for communication between teachers and students. Teachers can answer students' questions in time. Students can also exchange and discuss with each other, learn from each other, and broaden their views. In addition, according to the number of tasks completed by students, the length of video viewing, the number of discussions and visits, and the learning situation of students can be tracked through the "Statistics-Student Management and Supervision" interface of Superstar Learning System.
Applying the "notification" function of Superstar Learning System Communication Platform, we can also issue relevant learning task notifications, or course announcements such as adjusting classes, arranging experimental classes, and examinations to ensure that the information is conveyed in a timely manner.
The application of Superstar Learning System in the teaching process possibly provides great convenience for enriching the classroom teaching, which is conducive to the comprehensive improvement of students' professional, methodological and collaborative abilities, and promotes the overall achievement of quality education (Liu & Zhang, 2019).

Pearson correlation coefficients
Pearson correlation coefficients were computed to assess the relationship between performance expectancy, effort expectancy, lecturer influence, peer influence, user innovativeness, interface simplicity, multiple functions, and behavioral intention to use Superstar Learning System (Table 1). Table 1, there were strong, positive correlations between performance expectancy and behavioral intention (r = 0.887, p < .001), between effort expectancy and behavioral intention (r = 0.889, p < .001), between lecturer influence and behavioral intention (r = 0.876, p < .001), between peer influence and behavioral intention (r = 0.878, p < .001), between user innovativeness and behavioral intention (r = 0.773, p < .001), between interface simplicity and behavioral intention (r = 0.842, p < .001), between multiple functions and behavioral intention (r = 0.823, p < .001) at the significance level .01. Therefore, we rejected all the seven null research hypotheses.

The extended TAM
In order to answer the research question "will performance expectancy, effort expectancy, lecturer influence, peer influence, user innovativeness, interface simplicity, and multiple functions exert a positive influence on behavioral intention to use Superstar Learning System?", a WarpPLS SEM was operated to validate the extended TAM (See Figure 3).

Model fit and quality indices
As presented in Figure 3, the structural equation model is considered fit since the indices generally meet the required criteria (Kock, 2015
This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

Discussion
This discussion part explores the rationales for strong, positive correlations between performance expectancy, effort expectancy, lecturer influence, peer influence, user innovativeness, interface simplicity, multiple functions, and behavioral intention.

Performance expectancy and behavioral intention
The performance expectancy greatly influences behavioral intention to use mobile apps. In case users believe Superstar Learning System is useful for their learning, they will possibly try to use it, leading to enhanced behavioral intention. Teachers would praise students if they can complete learning tasks in a timely manner. Thus if students think Superstar Learning System can help them with their task, they would most likely use it. Collaboration is important in learning to improve students' academic performances. If Superstar Learning System is believed to enable students to improve their collaborative abilities and performances, they will possibly engage in the use of Superstar Learning System, which will enhance their behavioral intention.

Effort expectancy and behavioral intention
The perceived amount of effort spent using mobile apps will influence behavioral intention. In case students think Superstar Learning System, flexible and easy to use, does not require much of their effort, their intention of use will possibly be fortified. Furthermore, if Superstar Learning System can help them interact with their peers via clear and understandable instructions, their intention of use will also be strengthened.

Lecturer influence and behavioral intention
Superiors greatly influence students' decision of whether to use mobile apps. In case teachers suggest that students use Superstar Learning System, and use it themselves both in and out of class, students will possibly follow them. Teachers' further help and support for the use of Superstar Learning System will also definitely encourage students to use it, which will also consolidate their behavioral intention.

Peer influence and behavioral intention
Students tend to be in conformity with the majority of people surrounding them. In case that their classmates, schoolmates, friends, families, and other acquaintances use Superstar Learning System, they will also engage in the use of it, because they do not want to be isolated and idiosyncratic.

User innovativeness and behavioral intention
Personalities of students significantly influence the use of new technologies such as Superstar Learning System. If they like to try the new mobile app, are willing to examine it, and are ready to accept it, their behavioral intention will undoubtedly be augmented.

Interface simplicity and behavioral intention
The simple interface is also conducive to intention of use. If the interface of Superstar Learning System is concise, simple, easy to follow, clearly organized, and nearly outlined, students will possibly use it since it does not require much effort to learn.

Multiple functions and behavioral intention
Multiple functions are also an important construct influencing behavioral intention. Students will most likely not hesitate to use Superstar Learning System if it carries rich resources, contains many learning activities, includes various courses and learning approaches, and makes them follow the teacher easily and conveniently.

Conclusions
This section includes major findings, limitations of this study, as well as future research recommendations.

Major findings
This study concludes that performance expectancy, effort expectancy, lecturer influence, peer influence, user innovativeness, interface simplicity, and multiple functions exert a positive influence on behavioral intention to use Superstar Learning System. This is generally consistent with previous works (e.g. Venkatesh

Limitations
This study is limited to participants in a Chinese context, which might have entailed a contextual bias. Superstar learning system is an application widely used in China rather than across the world. The generalizability may be negatively influenced by this limitation.

Future research directions
Future research could extend the research context to a wider scope. Interdisciplinary research between computer sciences, education, linguistics, and other related disciplines is also needed to explore an extended TAM to use Superstar Learning System.