The study adds to the body of knowledge about different types of feedback. Feedback is considered a fundamental component for supporting and regulating learning processes. Especially in computer-based and self-regulated learning environments, the nature of feedback is of critical importance. Hence, this study investigates different types of automatically generated feedback. Seventy-four students participated in the experimental study, in which they had to write texts and create concept maps reflecting their understanding of climate change while receiving three different forms of model-based feedback. Results indicate that concept maps were structurally and semantically more similar to an expert solution than written texts. However, the form of the automatically created model-based feedback did not influence the type of representation. (Contains 4 tables and 1 figure.)