This work starts with the widespread view that learning processes benefit from computer-supported simulations. The software TinkerPlots (Konold & Miller 2017), which is used nationally and internationally, provides easy access to modeling and to performing stochastic simulations. For this reason, in order to show the possibilities and limitations of the software, the first part of this thesis examines how TinkerPlots supports stochastic simulations. Based on this tool analysis, a learning trajectory on statistical inference with randomization tests was developed for a course in mathematics education for primary school preservice teachers using a design-based research approach. This short sequence of lessons - a completely new topic for most participants - was incorporated at the end of the course. The design of the learning trajectory will be presented in the second part of the thesis together with an exploratory case study. After they experienced the new lessons, six of the preservice teachers participated in an exploratory case study in which they conducted, in pairs, a randomization test using TinkerPlots. The evaluation of this study is also explained in detail in the second part of this thesis. Thus, the present work contributes to the state of mathematics education research, firstly through the analysis of the TinkerPlots software, and secondly by examining an introduction to the logic of inference with randomization tests using TinkerPlots.