Abstract:

Among the many difficulties appeared with the arrival of COVID19 we find those associated with educational training. This new scenario requires adapting the traditional in-person approach to another one capable of facing the drawbacks associated with distance learning. In the university context, we have observed that virtual classes following the pre-COVID19 methodology have little feedback from the students. There are also significant difficulties in the evaluation process through non-face-to-face exams, mainly related with the responses authorship. To solve these problems we have applied a project-based approach that replaces the traditional exam. Each student, or group of students, has a different project. The base constraints are: every project must cover all the learning objectives, with the same amount of required work from the students and the same difficulty. The main challenge is that all the projects must be different, such that it would not be possible to solve one project just by copying the solution from other project. And if one tries to use the solution from other project in his/her project, she/he will have to use in fact the intended learning outcomes to be able to use that information in her/his own project. Also, it is important to keep the teacher work load at a reasonable level. The practical implementation consist of dividing the concepts in the course in groups, and for each group of concepts to build a number of different project cases, just for the concepts in the group. The course project is composed by one single project case for each group of concepts. And the set of all the course projects results from the combination of the project cases. For instance if we have 3 groups of concepts and 5 project cases for each group, then we have 5*5*5 = 125 possible course projects. We use a simple algorithm (in MATLAB) to build the combinations. After applying this methodology, we have observed an increase in student participation both during classes and in course forums, and also a proactive attitude towards self-learning. Instead of punishing the interchanges of information among student, this approach promotes the peer-learning. This exercise replaces the exam that remains as a testimonial evaluating instrument with a lower weight. As a result of our experience beyond the aforementioned, we have observed an improvement in the final marks, and subjectively, the student seems to worry more on the learning than on the marks.

Keywords:

Virtual class, configurable random projects, self-learning, peer learning, algorithm.