ABSTRACT
Blended learning has risen in primary and secondary schools due the COVID-19 pandemic, and is expected to remain after the pandemic. Despite this, the large majority of research into blended learning has studied university student populations. Thus, this paper seeks to research differences in usage between and in primary and secondary schools. Data were collected from a prototype learning portal, resulting in a dataset of 803 students from 12 schools with 45 teachers. Teachers and students could perform diverse actions on the platform, coded as 19 student features and 6 teacher features. These features were used to perform cluster analysis with k-means clustering on the entire dataset, split between teachers and students. Differences in primary and secondary schools were also analyzed. Clustering on the entire student dataset resulted in two clusters that differed on amounts of learning activities and learning tracks. Primary and secondary school students differed on the amount of learning activities, amount of learning tracks, and ratings of fun. Clustering on teachers resulted in two clusters differing in exploration and personalization. These results might be due to differences implementations of blended learning, where young students begin with simple rotation and complexity in implementation increases with age. These results might also be due to differences in teacher competencies. These results belie the importance of research into blended learning in pre-university education, filling a gap in literature and providing guidelines for coaching and use of multiple learning applications. © 2021, Springer Nature Switzerland AG.