Revisão sistemática sobre mecanismos de avaliação automática em objetos de aprendizagem sobre programação

Abstract

In programming courses, students should be attentive and focused on the concepts, code syntax, and logic of algorithms in order to become good programmers. A great deal of information is presented to students in classroom, thus making very important the application of practical exercises to improve the learning experience in computing. After all, according to Piaget, learning is strongly linked to experience and practice. However, instructors of programming disciplines faces difficulties to apply exercises at the end of each topic due to the time required for correction and feedback to the student. One solution to this problem would be to use automatic assessment tools. Automatic evaluation systems can bring benefits to learning because students can solve exercises, undergo the automatic evaluation program and get feedback on their work at the end of each task. Thus, special attention must be taken on the integration of these mechanisms and content of disciplines, providing proper learning objects about programming. The objective of this project was to conduct a systematic mapping of automatic evaluation mechanisms and the application of them in learning objects about programming. For this systematic review were collected 17 studies. By reading them, we identified 10 tools that are already being used for introductory courses in computing. It was observed the purpose of automated evaluation tools and its features. Such tools contributes to the teaching and development of students in introductory programming courses, also aiming at teaching basic concepts of object-oriented programming and software testing. Most of the tools evaluates and provides automatic grades to students, and in some cases even verify the originality of code is assessed, looking for plagiarism among students’ responses. We could also identify studies that spurs student to perform collaborative corrections and even to self assess their work. Overall, we identified positive results with the integration of these mechanisms in learning objects. Such mechanisms improve the didactic of classes, contributing to the effectiveness of learning of programming.