A lot of efforts has been done to implement a system of feedback blended learning systems in the past. New possibilities and new ideas makes it possible to improve these systems in order to maximize the effectiveness of teaching statistics to students at university level. The aim is both, a systematic implementation of teaching concepts in software using modern interactive tools and to invent new ways to teach statistical concepts in the area of statistical uncertainty and statistical modeling. This thesis presents the newly developed R package tguishiny (de Cillia et al., 2018), which gives teachers a way to create feedback-based courses for teaching R syntax and statistical concepts. Questions created with this package can contain interactive visualizations thanks to the powerful shiny framework. Answers submitted by students are summarized live in a “teacher view”. The ui design of the questions has been laregly influenced by the tGUIOnline project, which is a teaching tool developed in and used by the Statistics Austria institute.