Computer science students need mathematical proof skills. At our University, these skills are being taught as part of various mathematics and computer science courses. To test the skills of our students, we have asked them to work out a number of exercises. We found that our students are not as well trained in basic proof skills as we would have hoped. The main reason is that proof skills are not emphasized enough. Our ndings are the result of a small experiment using a longitudinal measurement of skills. This method gives better insight in the skills of students than more traditional exam-based testing methods. Longitudinal measurement does not allow the students to speci cally prepare themselves for particular questions. The measurements thus relate to skills that are retained for a longer period of time. In our Department, erce debates have been held in the past discussing such issues as \what proof skills do our students have?". An important aspect of our work is that it tries...
Pieter H. Hartel, Bert van Es, Dick Tromp