In Massive Open Online Courses (MOOCs), learners can post both text comments and overall ratings regarding the courses. There is growing interest in assessing the consistency of online reviews and the determinants of learner satisfaction. This study analyses the disconfirmation effect between textual review topics and the determinants of learner satisfaction in MOOCs. The MOOCs are categorised under three disciplines - Social Science, Technical Science, and Humanities & Natural Science. A crawler was employed to collect the corpus, extracting 93,679 reviews of 5,214 online courses from a Chinese university MOOC platform (icourse163.org). Textual analytics was used in the topic extraction. The empirical results suggest a strong disconfirmation effect between textual reviews and the determinants of learner satisfaction, i.e., not all textual review topics affect the overall learner satisfaction. Compared with positive reviews, negative (and neutral) reviews have a stronger disconfirmation effect. Further, the antecedents of learner attention are course-discipline specific. The disconfirmation effect is course-discipline dependent, with the most prominent for Technical Science courses, and the least for Humanities & Natural Science courses. This study provides a framework to guide platform managers and course instructors in better course delivery and enhancing overall learner satisfaction.
- Data science
- Disconfirmation effect
- Online learning
ASJC Scopus subject areas
- Library and Information Sciences