Anaysis of course structures and learners' engagement in MOOCs: the case of Thai MOOCs | |
| Author | Wanlipa Thongsuntia |
| Call Number | AIT Thesis no.CS-19-04 |
| Subject(s) | Educational evaluation--Data processing Distance education--Thailand Internet in education--Thailand Education, Higher--Computer-assisted instruction--Thailand |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Computer Science, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Thesis ; no. CS-19-04 |
| Abstract | Recently, Massive Open Online Courses (MOOCs) have achieved more than 100 million registered learners around the world. Based on several active research, the influential accomplishment of MOOCs is a number of learners’ engagement. MOOCs, nevertheless, has major issues including not only low instructional quality, but also high dropout rate. As from the above fact, the research question of this thesis is to find out how the difference in course structure design may affect learners engagement. With regard to the scope of this thesis, it focuses on Thai MOOC courses in STEM subject area. The dataset contains 28 STEM courses. The learners’ engagement dataset selects the top 20 of the most popular courses according to registered learners. This paper applies the learning analytics to analyze the patterns and clustering of two significant dimensions: i) course design & structure and ii) learners’ performance & engagement. Furthermore, this thesis explores the relationships of both dimensions by using data mining, machine learning, and visualization techniques. Regarding the course design & structure, there are four dimensions including course length & effort, Bloom’s taxonomy, number of learning components, and sequence components. Regarding learning analytics outcomes, the results show that courses with medium lengths and efforts have the highest percentage of passing learners and comprehensive learners in the dimension of course length & effort. Furthermore, Applying-focus is the best group which has the highest percent not only passing learners, also comprehensive learners in Bloom’s Taxonomy dimension. With clustering of components, the best group of Number of components is video-HTML focus. For sequence components clustering, the highest percentage of passing learners and comprehensive learners is in the group of discussion-HTML and video-HTML focus. The visualization of learning analytics are illustrated in MOOCA (Massive Open Online Course Analytics)1 . MOOCA is a web application implemented using Node.js, D3 and Plotly for visualization. In addition, MOOCA has an essential function as ‘Recommendation’ which advises users when they design the courses. The outcome of this function presents the cluster of the course. Consequently, the result shows that the course is in the group of high passing learners or not. |
| Year | 2019 |
| Corresponding Series Added Entry | Asian Institute of Technology. Thesis : no. CS-19-04 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Chutiporn Anutariya; |
| Examination Committee(s) | Dailey, Matthew N.;Suporn Pongnumkul; |
| Scholarship Donor(s) | Royal Thai Government Fellowship; |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2019 |