TrainBayes for School



Project partners


  1. TrainBayes (DFG)
  2. siMINT (BMBF)
  3. FEHLBa (DFG)
  4. TrainBayesS (DFG)
  5. TrainBayes for School (Müller-Reitz-Foundation)



  1. Material from the project TrainBayes (DFG)

Transfer of the material of the DFG-projekt TrainBayes into School

The project TrainBayes for School aims at making the training materials from the completed DFG project TrainBayes directly accessible for mathematics lessons in school. In school, the ability to use Bayes' formula to evaluate hypotheses in uncertain situations is considered an essential component of probabilistic thinking, which should already be introduced in lower and upper secondary school and can be seen, for example, in everyday risk assessment (e.g. when assessing a corona pandemic). For this reason, the understanding of so-called Bayesian situations and of Bayes' formula have been discussed in mathematics education research for decades. Numerous findings from previous research show, however, that students, laypersons and experts alike often fail to use Bayes' formula adequately. In the completed DFG project TrainBayes (EI 773/4-1 and KR 2032/6-1), materials were therefore developed for four different training courses on Bayesian thinking in the professions of medicine and law (in which Bayesian thinking is particularly relevant) and their effect on promoting Bayesian reasoning was studied. The project results show that all training courses lead to significant short and medium-term learning success. In order to be able to measure the training successes, various authentic contexts were developed in both professions (medicine and law) in which Bayesian thinking was tested. The materials developed in TrainBayes are particularly valuable for schools for two reasons:
  1. The helpful training materials are based on the format of so-called "natural frequencies", which is still rarely used in schools, as well as promising visualizations that are unusual in school. The DFG project has shown that training with those strategies, that are uncommon in schools, lead to significantly better learning gains than training with the strategies established in schools. For this reason alone, the transfer of these training materials to schools is favorable.
  2. 2. The authentic tasks developed in TrainBayes are a valuable extension for school modeling tasks, in which artificially embedded contexts without authentic statistical information are often used, especially in stochastics lessons.
In the project TrainBayes for School, the digital materials developed in TrainBayes are adapted in such a way that they can be can be used directly at school.  


Project leadership:
Theresa Büchter
Prof. Dr. Andreas Eichler