A teaching experiment was carried out in a university-level thermodynamics course using adaptive and interactive e-learning material, created in the new Moodle question type Stateful extending the original e-learning platform STACK. The system collects data about the students that is used to algorithmically classify them according to their behaviour in solving problems. It is observed that the classification of this data predicts students' success in the other parts of the course for a majority of students. Also, the classification is statistically consistent with Thermodynamic Concept Survey and Maryland Physics Expectation Survey.