This research shows, on a wide dataset of students, that evaluation methods to promote active learning improve students' academic success and performance. On a database of 4622 engineering and design students, we demonstrate that this result holds when correcting the selection bias--more engaged students opt for active learning. The intriguing result is that the percentage of students that opt for active learning decreases along time. The discussion points to the role of faculty as crucial to decrease the resistance of students to active learning. The general methodological approach in this paper may be useful to assess the performance of any change in a subject's evaluation.