useR! 2024

We present a way to create web-based learning applications and track the user interactions to facilitate educational research and improve the learners' experience. The main ingredient is a new R package, learnrextra, which extends the well-known learnr package by interaction tracking, layout improvements and a summary panel. It allows authors to create learning applications in the familiar RMarkdown format or as Shiny applications. User interactions can then be tracked in a highly granular, configurable and anonymous way, e.g. mouse movements, clicks, exercise submissions, etc. Tracking data for all learning applications is collected via a REST interface.  An administration interface allows to manage learning applications, set up experiments (A/B testing), monitor tracking and download collected data. We will present our software and demonstrate how to create applications and experiments for learning analytics in R. We will share results from our experiments with students in class and discuss challenges we faced. All components are open-source: https://github.com/orgs/IFAFMultiLA/repositories. An extensive documentation is available at https://ifafmultila.github.io.