Correlation study between the access mark and the performance in project-based and standard courses

The access mark to engineering studies is often used as an a priori success estimator. In our institution, we have observed that the correlation of the access mark with the grades obtained in project-based courses (R=0.52) is slightly lower than the one obtained with the average of the other non-project-based courses (R=0.58), and is especially low in capstone projects (R=0.31). Project-based and Challenge-based courses are one of the most acknowledged ways of promoting the learning of transversal skills, specifically innovation and entrepreneurship skills. In our institution, ICT engineering bachelor students perform a project-courses path, with three subjects of growing complexity in the 2nd, 3d and 4th year. While the first two are partially guided and with challenges proposed by the faculty members, the 3d one is a 12 ECTS capstone project with challenges proposed by industry or external institutions. In this study, we have analyzed the performance of the students along 10 academic years (2011-2012 to 2020-2021). Not only the correlation with the access mark in these courses is lower but the prediction interval is also different. While it is almost impossible that a student with a low access mark gets an outstanding average mark in the bachelor and vice-versa, there are students with a low access mark which have an outstanding performance in the capstone project and students with a very high access mark and with high results in analytical courses but with a poor performance in capstone projects. Therefore, a different kind of skills are promoted in these courses.