Understanding Learner Dissatisfaction in Online Data Science Courses: An Analysis of Learning Experience

The growing popularity of online learning, especially in skill-based areas like data science, has been accompanied by mixed reviews from online learners. This study analyzes 842 negative learner reviews collected during February 2025 from Coursera-based massive open online courses using a mixed-methods approach to examine the major themes, structure, and intensity of dissatisfaction. A multistage thematic analysis was conducted to identify various dissatisfaction themes, followed by salience–valence analysis to evaluate frequency and severity. To enhance methodological rigor, interrater reliability was established, indicating excellent agreement between coders. The findings suggest that content and pedagogical shortcomings are strongly associated with learner dissatisfaction, whereas technical and service issues intensify criticism. In addition, learner dissatisfaction was further classified into factual and opinion-based categories. The findings suggest that factual and opinion-based dissatisfaction are nearly equally distributed; however, a substantial proportion of reviews (76.6%) exhibit both elements, indicating that dissatisfaction is predominantly hybrid in nature. Overall, the study offers a framework for understanding negative learner experiences in data-intensive courses.