The number of online learning users grows drastically and analyzing their learning behaviour has become a hot spot in recent years. However, there have been limited comparative studies on different learning processes, i.e. silent learning and active learning behaviour, and their influence on learning performance and learning experience. This study distinguishes between two types of learning behaviour, that is, silent learning which corresponds to content consumption, and active participation which corresponds to content contribution. By collecting and analyzing user log data from a Massive Online Open Course using OLS regression and sentiment analysis, we find that students’ reply, video watching, and test taking activities simultaneously enhance their learning performance and course learning experience. By contrast, comment behaviour does not indicate better student performance and learning experience. These findings provide novel insights in student silent and active participation, and point to future investigation of predictive studies based on real-time learning behaviour.