| Journal on Artificial Intelligence DOI:10.32604/jai.2022.030353 |  |
| Article |
Research on the Dissemination and Influencing Factors of Big Data and Artificial Intelligence Related Courses in Colleges and Universities-Taking MOOC as an Example
1School of Management, Wuhan Institute of Technology, Wuhan, 430079, China
2School of Management, Wuhan Institute of Technology, Wuhan, 430205, China
3School of Information Management, Wuhan University, Wuhan, 430072, China
4School of Management, Wuhan Institute of Technology, Wuhan, 430205, China
5Business Administration and Management, Hungarian University of Agriculture and Life Sciences, Godollo, 2030, Hungary
*Corresponding Author: Chen Xiaoyu. Email: 1091902674@qq.com
Received: 24 March 2022; Accepted: 16 May 2022
Abstract: The rapid development of information technologies such as artificial intelligence, Internet and big data has promoted the deep integration of technology and education, especially the rise of large-scale online courses, which provides a great opportunity for curriculum teaching reform in colleges and universities. At the same time, artificial intelligence, as a cutting-edge technology, has good development prospects and has become a popular professional course in colleges and universities, artificial intelligence technology has become the focus of subject education in many universities. The combination of online education and AI courses will also greatly enhance the enthusiasm of users and expand the dissemination scope of big data and AI related courses. Based on the grounded theory and technology acceptance model, this paper obtains the users’ perception evaluation of learning MOOC big data and artificial intelligence courses through field interviews, summarizes the core categories, constructs MOOC user acceptance model, collects sample data by questionnaire and makes an empirical study. It is found that perceived usefulness, expectation confirmation and primary communication effect have a significant positive impact on the secondary communication effect, and the content quality and socialization interaction indirectly affect the secondary communication effect of MOOC curriculum by affecting the primary communication effect.
Keywords: Artificial intelligence; User behavior; MOOC
Artificial intelligence (AI) was born in the 1950 s. At first, AI could only calculate, store and transmit information. Then, AI began to imitate people’s information perception function, resulting in great changes in human production and lifestyle. At present, emerging technologies represented by 5G, big data and artificial intelligence are in a period of highly active and intensive innovation, and an innovation network layout of interdisciplinary convergence, symbiosis and sharing of artificial intelligence has gradually formed. At the same time, as cutting-edge technologies, big data and artificial intelligence have good development prospects and become the focus of popular professional courses and discipline education in colleges and universities.
“Online education” and “online learning” have always been the focus of the education industry from germination to development [1]. The combination of online education and AI courses will also greatly enhance the enthusiasm of users and expand the dissemination scope of big data and AI related courses. As an online learning platform with a large audience, Massive Open Online Courses (MOOC) has been continuously introduced into college teaching activities because of its perfect functions, rich types of courses and large number of learners [2]. Colleges and universities use MOOC platform as a teaching auxiliary unit to provide students with autonomous learning opportunities, meet students’ knowledge needs, and promote the dissemination and sharing of knowledge. Although MOOC has attracted much attention as a new thing, its application and dissemination have become a more practical problem. What factors are playing an important role in the process of MOOC communication in universities has become the focus of subsequent research on the construction of MOOC in Colleges and universities.
This empirical study on the dissemination and influencing factors of big data and AI MOOC course takes college teachers and students as the survey object, mainly for the following reasons: (1) As of June 2021, the number of online education users in China has reached 325 million, accounting for 32.1% of the total Internet users, college teachers and students are an important part of online education users; (2) As the main audience of MOOC platform, college teachers and students use MOOC more frequently. Therefore, their use experience and feedback can play a certain role in optimizing MOOC learning experience and expanding communication ability.
This paper selects five universities: Wuhan University, Huazhong University of Science and Technology, Central China Normal University, Huazhong Agricultural University, and Wuhan Institute of Technology. According to the requirements of qualitative research methods of grounded theory, set the interview place, adopt the method of random sampling, and finally select 36 college students and 10 teachers as the interview objects.
This paper uses the grounded theory research method to obtain first-hand information through the combination of in-depth interview and focus group interview [3]. The interview was conducted from September 8 to September 20, 2021. Before the formal interview, a semi-structured interview outline was formed by referring to relevant literature [4], and guiding questions related to the service quality of mobile library were set. The interview was conducted from two aspects: the behavior of relevant communicators in the communication of MOOC big data and AI courses, and the attitude and behavior of the recipients. The interview contents are shown in Tab. 1.
During the interview, first of all, it is necessary to determine whether the interviewees have used MOOC and whether they are familiar with the operation and plate setting of MOOC, then conduct the interview around the semi-structured outline, and record the interview content in the form of text and recording. The interview methods include personal in-depth interview and focus group interview. We conducted 10 personal in-depth interviews for 20–30 min each; 7 groups of focus group interviews, each 30–40 min, with 2–6 interviewees, a total of 46. The interview was conducted in a relaxing and comfortable place, which achieved good results and obtained rich and reliable first-hand data.
2.3 Category Extraction and Analysis
Open coding is mainly to disrupt and decompose the obtained original data, and then recheck and summarize, in order to summarize and refine new concepts from the original sentences of the research object [5]. In this paper, 384 original sentences were obtained by dividing, recombining and refining the concept of data sentences. After labeling, conceptualizing and categorizing the 384 original sentences, 162 preliminary concepts were obtained, and 13 categories were obtained by further categorizing, respectively: convenience of content and form, content richness, content brand effect, diversity of ways, system stability, information accuracy, sense of use, timeliness of content, system interactivity, experience, system simplicity, social influence, system function completeness. The open coding process is shown in Tab. 2.
Axial coding integrates a higher level of abstract categories through analysis of the concepts formed by open coding and the relationship between concepts [6]. Through further refining the data, this study obtains five main categories: perceived usefulness, content quality, social interaction, self-efficacy and Expectation Confirmation. These five main categories show the relationship between independent categories. The main category and category logic formed by axial coding are shown in Tab. 3.
The task of selective coding is to systematically deal with the relationship between categories, determine the core category and secondary category, and form a grounded theory based on category relation. Core categories emerge naturally in open coding, and their main characteristics are: ①Core; ②Explanatory; ③Frequent reproducibility; ④It is easy to relate to other variables and has significance [7].
Through the progressive coding of selective coding, it is found that each main category focuses on the communication effect of big data and AI MOOC courses, so “the communication effect of big data and AI MOOC Courses” is defined as the core category. The relationship structure between the main categories and the representative statements of the respondents are shown in Tab. 4:
The theory finally formed by grounded theory is called substantive theory. It is not completely equivalent to the formal theory formed by quantitative research. Formal theory is considered to be a systematic theory that transcends specific situations and can be widely tried. The substantive theory is the revelation of specific phenomena and their internal relations [8].
Build the model according to the story line obtained by three-stage coding, as shown in Fig. 1.
Figure 1: Influencing factors model of MOOC course communication effect based on user perception
The basic function relationships contained in the model are: content quality, perceived usefulness, social interaction, Expectation Confirmation and self-efficacy. These five main categories have an impact on the communication effect of college big data and AI MOOC courses. Content quality includes four sub categories: convenience of content and form, content richness, content brand effect and timeliness of content; Socialized interaction includes two sub categories: system interactivity and social influence; Perceived usefulness includes two sub categories: system simplicity and system function completeness; Expectation Confirmation includes two sub categories: sense of use and experience; Self efficacy includes three sub categories: diversity of ways, information accuracy and system stability.
2.3.5 Theoretical Saturation Test
In order to ensure the reliability of the research, it is necessary to test the theoretical saturation of the coding results. Theoretical saturation means that the theory tends to be saturated when the data that can not further develop the characteristics of a certain category can not be obtained. After coding the previous data, this study selects samples for the second in-depth interview. This time, three students are selected for the interview. The interview time is about two hours, forming a new first-hand data. Through the sorting and analysis of the data, no new categories appear, indicating that the coding has reached theoretical saturation.
3 Research on Influencing Factors of MOOC Course Communication Effect of Big Data and AI in Colleges and Universities
Combined with the coding results of grounded interview, considering that “learners’ ability and level” and “platform and curriculum characteristics” will affect the effect of MOOC big data and AI courses in the communication process, three variables, self-efficacy, content quality and socialized interaction are added; the communication effect refers to the changes in the recipient’s psychology and behavior before and after the action of the transmitted information [9]. The two-level effect theory of communication holds that the ultimate purpose of communication, that is, the social premise of communication, can be achieved only when the transmitted information is sent to the receiver’s psychological system to transform the information into mentality energy, which is externalized or released into the receiver’s behavior and produce practical results. The transformation of information into psychological energy is the premise of effective communication and an indispensable key procedure in the process of communication. Therefore, in this study, the primary communication effect and secondary communication effect in the secondary effect theory are introduced as variables to build a model to systematically study the relevant factors affecting the communication effect of MOOC courses, and build a research model as shown in Fig. 2.
Figure 2: MOOC big data and AI course communication effect influencing factor model
The model variables of this study include three parts: platform and course characteristic variables, self-efficacy and information system model variables. The specific discussion and assumptions are as follows:
3.2.1 Platform and Course Characteristic Variables
(1) Content Quality
Content quality refers to the quality of big data and AI courses provided by MOOC platform, mainly involving video quality, chapter arrangement, key and difficult points and other relevant information [10]. If MOOC platform users perceive that the content quality of MOOC courses is high, and its updates are timely, it will enhance users’ cognition of the content quality of MOOC big data and AI courses, so as to improve their perceived usefulness and expectation confirmation level. Moreover, the MOOC big data and AI course content quality is high, users can get better satisfaction, and it will also be more conducive to improve the primary communication effect of MOOC courses (Lin, et al., 2012 [11]；Cheng, 2014 [12]). Thus, we hypothesize that
H1. The content quality of MOOC big data and AI courses positively influences the perceived usefulness of MOOC users.
H2. The content quality of MOOC big data and AI courses positively influences the expectation confirmation of MOOC users.
H3. The content quality of MOOC big data and AI courses positively influences the primary communication effect of MOOC users.
(2) Social Interaction
Social interaction refers to the interaction between users and teachers and other participants in the process of learning big data and artificial intelligence courses on the MOOC platform, including: interacting with teachers and other learners through commenting or giving likes to on course content; through communicating and interacting with teachers and students in the MOOC big data and artificial intelligence courses discussion area; forwarding and sharing MOOC big data and artificial intelligence courses. MOOC is different in form from traditional teaching methods. Users need to interact with teachers through the network, solve problems in the learning process, interact with other users in the virtual community, deepen their understanding of knowledge, and better master knowledge. If user interaction with teachers and other learners can be positively responded and supported, the user-perceived usefulness, expectation confirmation level, and dissemination effect of MOOC big data and artificial intelligence courses will be improved. Therefore, we hypothesize that
H4. The social interaction between MOOC big data and AI courses positively influences the perceived usefulness of MOOC users
H5. The social interaction between MOOC big data and AI courses positively influences the expectation confirmation of MOOC users
H6. The social interaction between MOOC big data and AI courses positively influences the primary communication effect of MOOC users
Learners’ self-efficacy refers to learners’ self-assessment of whether they can use their abilities to complete learning tasks [13]. In the MOOC learning process, if the user has a positive self-evaluation of his future learning achievements, he will be more active in the learning process, be more determined to complete the MOOC big data and AI course learning, actively complete various tasks in MOOC learning with a better attitude, and well understand and master the course knowledge points, and finally achieved good course results. On the one hand, when users’ self-efficacy continues to increase, they are more likely and willing to recommend the MOOC courses they have learned to others, which is conducive to the preliminary dissemination of MOOC big data and AI courses; On the other hand, the stronger the user’s sense of self-efficacy, the greater the possibility of digest knowledge after learning big data and AI courses, which will help to promote the further dissemination of course content and enhance its dissemination effect. Thus, we hypothesize that
H7. The self-efficacy of MOOC big data and AI course users positively influences the perceived usefulness
H8. The self-efficacy of MOOC big data and AI course users positively influences the expectation confirmation
H9. The self-efficacy of MOOC big data and AI course users positively influences the primary communication effect
H10. The self-efficacy of MOOC big data and AI course users positively influences the secondary communication effect
3.2.3 Information System Model Variables
(1) Perceived Usefulness
Perceived usefulness refers to the degree to which users perceive the use of information systems to improve their performance [14]. In MOOC learning environment, if users perceive that using MOOC big data and AI course services can improve their learning efficiency, deepen their understanding and mastery of knowledge, and enhance users’ sense of usefulness, the more useful they perceive, the more likely they are to recommend them to their classmates and friends for use and learning, so as to enhance the communication effect of MOOC big data and AI courses. Moreover, they will better digest the knowledge they think useful, which will also be conducive to the in-depth dissemination of MOOC big data and AI courses. Thus, we hypothesize that
H11. The perceived usefulness of MOOC big data and AI course users positively influences the primary communication effect of the courses
H12. The perceived usefulness of MOOC big data and AI course users positively influences the secondary communication effect of the courses
(2) Expectation Confirmation
Expectation confirmation means that the user will compare the actual use expectation with the expectation before use. If the actual use expectation is higher than the initial expectation, the user’s expectation confirmation level will be higher [15]. Bhattacherjee(2001a) [16], Bhattacherjee(2008) [17] found that Expectation Confirmation affects users’ cognition of the usefulness of the information system. If the course quality and learning expectation provided by the MOOC platform exceed the users’ expected level, the users’ Expectation Confirmation level is higher, which will improve the users’ perceived usefulness. When the quality and learning expectation of big data and AI courses provided by MOOC platform exceed the user’s expected level, users will most likely recommend MOOC courses to people around them for learning, so as to improve the communication effect of MOOC big data and AI courses. Thus, we hypothesize that
H13. The expectation confirmation of MOOC big data and AI course users positively influences their perceived usefulness
H14. The expectation confirmation of MOOC big data and AI course users positively influences the primary communication effect of the courses
(3) Communication Effect
The information transmitted by the online open class can only pass through the viewer’s sensory organs and act on their psychological system, which is in line with the viewer’s psychological choice, convert the received information into mentality energy, and produce the primary communication effect of the information dissemination of the online open class; and in the viewer’s own practice and topic interaction, the mentality energy is externalized into behavior to produce a visible social communication effect, which is the secondary communication effect [18]. Thus, we hypothesize that
H15. The primary communication effect of MOOC big data and AI courses positively influences the secondary communication effect
This paper collects data by means of network questionnaire. The questionnaire includes two parts of information: one is the survey on the basic situation of MOOC users, and the other is the specific variable measurement scale. Based on the mature scale at home and abroad, combined with the characteristics of MOOC big data and AI course education, the variable measurement scale is designed, and the scale is improved according to the opinions of experts, the final scale was formed.
The questionnaire was distributed through the online questionnaire platform, and users who had studied big data and AI related courses on the MOOC platform were invited to fill in the questionnaire with the help of wechat, QQ and other social media. Finally, 223 valid questionnaires were recovered. Among them, men accounted for 40.4% and women accounted for 59.6%. The educational background included undergraduate (82.5%) and master’s degree (17.5%); The weekly learning time is divided into less than 1 h (51.6%), 1–3H (39.9%), 3–5 h (4.5%) and more than 5 h (4%); The duration of use is divided into less than 3 months (67.3%), 3–6 months (17.9%), 6–12 months (5.4%) and more than 12 months (9.4%). It can be seen that the survey sample covers all kinds of people who study MOOC big data and AI courses at different times. It is representative and can be analyzed in the next step.
Smartpls2.0 was used to test the reliability and validity of the questionnaire, cronbach ‘ s α values are all greater than 0.6665, which is within the acceptable range(Chin, 1998 [19]), the composite reliability of each variable is greater than 0.7960, indicating that the measurement model has good reliability (Fornell, et al., 1981 [20]), can be used for the next regression analysis.
Smartpls2.0 is used to verify the model path, and the results are shown in Fig. 3. According to the model path coefficient and its significance level, only the hypothesis “H5: The social interaction between MOOC big data and AI courses positively influences the expectation confirmation of MOOC users, and the other 14 assumptions are tenable, as shown in Tab. 5.
Figure 3: Model path coefficient value and significance level
Perceived Usefulness (β = 0.2351, p < 0.001) and Primary Communication Effect (β = 0.4553, p < 0.001) significantly and positively affect the Secondary Communication Effect of MOOC courses on big data and AI, Expectation Confirmation indirectly affects the Secondary Communication Effect of MOOC courses on big data and AI by affecting Perceived Usefulness and Primary Communication Effect, which is consistent with the relevant research conclusions on the continuous use of information systems. From the perspective of path coefficient, the Primary Communication Effect of big data and AI courses in MOOC has a greater force on the Secondary Communication Effect, reaching 0.4553, and the force of Perceived Usefulness on the Secondary Communication Effect has reached 0.2351. For each unit of user’s Perceived Usefulness, its Secondary Communication Effect will increase by 0.2351; for every unit increase in Primary Communication Effect, its Secondary Communication Effect will increase by 0.4553.
Content Quality has a significant positive impact on Perceived Usefulness (β = 0.2446, P < 0.001), Expectation Confirmation (β = 0.1888, P < 0.001) (β = 0.1520, P < 0.001), Social Interaction has a significant positive impact on Perceived Usefulness (β = 0.2966, P < 0.001) and Primary Communication Effect (β = 0.1098, P < 0.001), the impact of Social Interaction on users’ Expectation Confirmation did not reach a significant level. From the perspective of path coefficient, the force of Content Quality on Perceived Usefulness is large, which is 0.2446.
Self-efficacy has a significant positive impact on Perceived Usefulness (β = 0.1350, P < 0.001), Expectation Confirmation (β = 0.2575, P < 0.001), Primary Communication Effect (β = 0.2754, P < 0.001) and Secondary Communication Effect (β = 0.1487, p < 0.001).
Through the interview on the acceptance of College MOOC big data and AI course users, this paper extracts variables, constructs theoretical models, and puts forward corresponding research hypotheses. Combined with relevant theories, this paper collects data to empirically verify the relevant influencing factors of MOOC big data and AI course communication effect, so as to provide practical data support for follow-up research.
The research data show that content quality, social interaction, self-efficacy, perceived usefulness and Expectation Confirmation all have a certain impact on the development of college big data and artificial intelligence courses on MOOC platform. In order to further improve users’ acceptance of MOOC big data and artificial intelligence courses and enhance the influence of MOOC college big data and artificial intelligence courses among teachers and students, this paper mainly puts forward targeted suggestions from three aspects: user needs, content quality and user experience, so as to provide some reference and suggestions for promoting the dissemination and user acceptance of college big data and artificial intelligence courses in MOOC environment.
4.2 Development Countermeasures and the Suggestions
4.2.1 Meet User Needs and Improve Perceived Usefulness
Under the learning environment of MOOC courses on big data and AI, the mobile and convenient characteristics of MOOC attract users to learn not only in traditional course, but also in online courses for consolidating the knowledge they have learned and expand their knowledge. Users form their Expectation Confirmation, Perceived Usefulness and Evaluation of Satisfaction during and after use. The more MOOC courses on big data and AI meet their own expectations, the higher the users’ Expectation Confirmation. Good Perceived Usefulness and better Primary Communication Effect will enhance the Secondary Communication Effect of MOOC courses. At the same time, users’ Perceived Usefulness will enhance the user experience, so as to improve the Primary Communication Effect. Therefore, the producers and providers of MOOC courses on big data and AI need to strive to meet the practical needs of university users’ online learning, enhance users’ Perceived Usefulness, attract users to use fragmented time for online learning, and improve the utilization of online learning resources.
4.2.2 Improve Content Quality and Encourage Interactive Communication
When users of MOOC courses on big data and AI conduct online learning, the Content Quality of the courses will positively affect users’ Perceived Usefulness of the course, the Expectation Confirmation after use and its Primary Communication Effect, thus affecting the Secondary Communication Effect of MOOC courses on big data and AI. Social Interaction refers to the communication and interaction between users, teachers and other learners in the learning process of MOOC courses on big data and AI. It can help users solve the problems and puzzles they encounter in learning. Through discussion with teachers and other learners, they can deepen their understanding and mastery of learned knowledge, and enhance their Perceived Usefulness and Satisfaction. MOOC is an online classroom where advantageous schools promote their excellent courses online for more people to learn. While paying attention to the Content Quality, MOOC should also attract the active participation of users, strengthen the interactive communication between users and teachers and other scholars, activate the online learning atmosphere, help users better participate in it and master the course knowledge.
4.2.3 Create A Successful Experience and Enhance User’s Confidence
The learning of online MOOC courses on big data and AI is different from the traditional classroom face-to-face teaching. Users need to use mobile terminal equipment and mobile network for learning. If users believe that they can deal with various problems encountered in the learning process with their own ability, it will positively affect their sense of Self-efficacy and Expected Results, so as to enhance their Expectation Confirmation level, then affect their Perceived Usefulness and Primary Communication Effect, and finally affect the Secondary Communication Effect of MOOC courses on big data and AI. Therefore, in order to expand the influence and dissemination scope of big data and AI courses, users should enhance their successful experience and self-confidence; Secondly, we should enhance the emotional communication and interaction between teachers and users, other learners and users. The guidance of teachers and the communication and discussion between users and other participants will spark their thoughts and improve their enthusiasm and self-confidence; In addition, users’ self-monitoring ability should be enhanced, Users’ self-monitoring is conducive to users’ self-summary and self-growth, enhance their sense of Self-efficacy, and finally improve the effect of users’ online learning [21].
4.3 Limitations and Future Directions
Although this paper discusses the influencing factors and paths of MOOC big data and artificial intelligence course dissemination, there are still some deficiencies that need to be improved in the follow-up research. Firstly, this paper explores the possible influencing factors and the relationship between influencing factors through the interview method, but the number of interviews is small, and the extraction of influencing factors may not be comprehensive enough. In the follow-up, it is necessary to expand the scope of interviews and improve the preciseness of the research. Secondly, the courses communication effect proposed in this paper is a dynamic and continuous process. In the follow-up research, the cross time point survey method should be used for longitudinal tracking test.
Funding Statement: The author(s) received no specific funding for this study.
Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.
References
1. O. N. Almotiry, M. Sha, M. P. Rahamathulla and O. Salih, “Hybrid cloud architecture for higher education system,” Computer Systems Science and Engineering, vol. 36, no. 1, pp. 1–12, 2021. [Google Scholar]
2. X. Guan, “From ‘Benefiting the public’ to ‘Custom-tailoring’: Study on the development from MOOC to SPOC,” Library Work in Colleges and Universities, vol. 35, no. 1, pp. 19–21, 2015. [Google Scholar]
3. C. Dunne, “The place of the literature review in grounded theory research,” International Journal of Social Research Methodology, vol. 14, no. 2, pp. 111–124, 2011. [Google Scholar]
4. D. Lee, S. L. Watson and W. R. Watson, “The relationships between self-efficacy, task value, and self-regulated learning strategies in massive open online courses,” International Review of Research in Open and Distributed Learning, vol. 21, no. 1, pp. 23–39, 2020. [Google Scholar]
5. J. R. Cutcliffe, “Methodological issues in grounded theory,” Journal of Advanced Nursing, vol. 31, no. 6, pp. 1476–1484, 2000. [Google Scholar]
6. H. F. Lin and R. Shang, “Definition of organizational routine and its construct dimensions: A grounded theory research,” Journal of Management Science, vol. 30, no. 6, pp. 113–129, 2017. [Google Scholar]
7. X., D, Jia and X., H., Tan, “The actual value of the classical grounded theory and its spirit to China management research,” Chinese Journal of Management, vol. 7, no. 5, pp. 656–665, 2010. [Google Scholar]
8. D. Walker and F. Myrick, “Grounded theory: An exploration of process and procedure,” Qualitative Health Research, vol. 16, no. 4, pp. 547–559, 2006. [Google Scholar]
9. O. Hasbiansyah, “Characteristic of moslem intellectual, a perspective of communication psychology,” Mediator: Jurnal Komunikasi, vol. 3, no. 1, pp. 151–156, 2002. [Google Scholar]
10. Q. Li and N. Liu, “Design and making method of teaching videos in MOOC——an empirical study of course based on coursera and ed X platforms,” Modern Educational Technology, vol. 26, no. 7, pp. 64–70, 2016. [Google Scholar]
11. W. S. Lin and C. H. Wang, “Antecedences to continued intentions of adopting e-learning system in blended learning instruction: A contingency framework based on models of information system success and task-technology Fit,” Computers & Education, vol. 58, no. 1, pp. 88–99, 2012. [Google Scholar]
12. Y. M. Cheng, “Extending the expectation-confirmation model with quality and flow to explore nurses’ continued blended e-learning intention,” Information Technology & People, vol. 27, no. 3, pp. 230–258, 2014. [Google Scholar]
13. E. Alqurashi, “Self-efficacy in online learning environments: A literature review,” Contemporary Issues in Education Research, vol. 9, no. 1, pp. 45–52, 2016. [Google Scholar]
14. Y. Chen and X. L. Du, “Individual behaviors on post-adoption usage of information systems: An empirical study,” Journal of Intelligence, vol. 32, no. 10, pp. 193–200, 2013. [Google Scholar]
15. Z. Zhang, Y. N. Wang, X. H. Chen and Y. Gao, “An empirical research on the factors affecting learners’ continuous learning intention in MOOC: A modified expectation-confirmation model,” E-Education Research, vol. 37, no. 5, pp. 30–36, 2016. [Google Scholar]
16. A. Bhattacherjee, “Understanding information systems continuance: An expectation-confirmation model,” Mis Quarterly, vol. 25, no. 3, pp. 351–370, 2001. [Google Scholar]
17. A. Bhattacherjee, J. Perols and C. Sanford, “Information technology continuance: A theoretic extension and empirical test,” Data Processor for Better Business Education, vol. 49, no. 1, pp. 17–26, 2008. [Google Scholar]
18. F. Jordán, W. C. Liu and A. J. Davis, “Topological keystone species: Measures of positional importance in food webs,” Oikos, vol. 112, no. 3, pp. 535–546, 2006. [Google Scholar]
19. W. W. Chin, “The partial least squares approach to structural equation modeling,” Modern Methods for Business Research, vol. 295, no. 2, pp. 295–336, 1998. [Google Scholar]
20. C. Fornell and D. F. Larcker, “Structural equation models with unobservable variables and measurement errors,” Journal of Marketing Research, vol. 18, no. 2, pp. 39–50, 1981. [Google Scholar]
21. R. L. Oliver, “A cognitive model of the antecedents and consequences of satisfaction decisions,” Journal of Marketing Research, vol. 17, no. 4, pp. 460–469, 1980. [Google Scholar]
 | This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. |