Teaching Model-Driven Engineering from a model-testing perspective

In Model-Driven Engineering (MDE), code is automatically generated from a model. However, even assuming that the code generated is error-free, there is no guarantee that the model used correctly reflects the requirements of the software system. This requires validation in the form of model testing. Although the practice of teaching modelling to students has been well researched, students still often struggle with modelling. Preliminary results of the thesis show that students' models often contain deadlock situations and that students' cognitive strategies when designing test cases for models often lack in testing for undesired scenarios. This thesis therefore looks at the challenge of teaching conceptual modelling via a new perspective, namely that of model testing. The thesis investigates whether teaching conceptual modelling via a testing perspective can be beneficial for students' modelling skills. Hereunto, we have designed two tools that provide feedback on model testing. These tools are being evaluated in terms of teaching students desired modelling skills. One of the tools had already been found to yield an increase for students' model coverage. The other tool will firstly be evaluated in a similar way. After which both tools will be evaluated in terms of yielding other desired modelling skills such as model evaluation capabilities.