Abstract:

A competence in the European Higher Education Area (EHEA) is defined as the ability of individuals to mobilize and organize their cognitive and affective resources to successfully address a complex situation. In this educational system it is considered that the continuous assessment of several competences encourages the autonomous learning of the students. The evaluation of competences is defined as the process of gathering evidences on the professional performance of a person, with the purpose of forming a judgment about his/her skills in relation to a professional profile. Any evaluation consists of selecting, collecting, analyzing, interpreting, and using certain information to make decisions. This is a complex process that requires defining profiles, organized around knowledge, skills and attitudes. It is also necessary to establish the evaluation criteria, identifying what the expected accomplishments are (indicators, demonstrations or evidence of each competence). In the university context, Professors dedicate a large amount of time to this evaluation, usually by conducting periodic exams throughout the course.
In Computer Science degrees, the evaluation of competences through practical exams is usually based on the software developed by the students. Professors typically evaluate the quality of software, once it is completed, by checking the source code. Therefore, the student does not know if he/she is reaching the quality criteria when he/she is developing the software. Then, it would be desirable to have tools that automate the process. These tools would help Professors in the evaluation of the source code and simultaneously, provide feedback to the students during the development of the software. As a consequence, the autonomous learning of students is encouraged.

The main idea of this paper consists of presenting a system oriented to the evaluation of competences in Computer Science degrees. The aim of this system is to provide a continuous evaluation of the competences achieved by students when they develop software. In particular, students can submit the developed software to the proposed system that will provide a feedback on the competences acquired and, simultaneously, the problems identified. The possibility of multiple evaluations over the same software can promote autonomous learning. In other words, the proposed system informs the student about the problems identified so far. Therefore, the student can understand and solve these problems. In this context, the Professor can provide to the system the evaluation criteria, according to the degree of acquisition of the different involved competences. This definition can be performed at two levels: (a) in terms of defining metrics for assessing the quality of developed software; and (b) in terms of defining tests (unit, integration, etc.) needed to verify the adequacy of the developed software to the specific requirements.
The architecture of the system consists of a server that will contain the software developed by the students together with a set of unit, integration, and performance tests. These tests will help to evaluate the software quality. Indeed, students will receive automatically a feedback of their work according to the metrics established by the Professor. In the current prototype, already implemented, they are integrated a Version Control System (Subversion), a Continuous Integration System (Hudson) and a Software Quality Platform (Sonar).

Keywords:

Continuous assessment, competences evaluation.