This article demonstrates how an adaptive instructional design for a physics module can be realized in a standard learning management system. We implemented a didactic design with physics-specific online exercises that were accompanied by either detailed or non-detailed instructions, depending on the results of the previous task (or a prior knowledge test for the very first exercise). This was realized by use of simple technological tools within the framework of a straightforward recommender system with four components. Consequently, students with less prior knowledge and/or lower learning achievements received more and different teaching assistance than those with high levels of prior knowledge or performance. This was done in the form of recommendations embedded within task feedback, suggesting which task to tackle next. We present first results which show that prior knowledge and online activity contribute to the learning progress in different ways depending on the type of task that was chosen. The detailed versions of the tasks were beneficial only to the students with lower or medium prior knowledge test scores while the students with higher levels of prior knowledge had less learning progress. In the future, our simple recommender system may serve as the basis for a more complex adaptive system, further closing the gap between research and practice in the field of technology-based adaptive learning. [For the complete proceedings, see ED600498.]