Framework para construção de objetos de aprendizagem com learning analytics

Abstract

The use of Learning Analytics techniques has stood out as an alternative to better understand students' behavior and performance in virtual environments. In this context, the possibility of introducing such techniques in the production of Learning Objects opens up, improving these resources and providing feedback on the students' evaluation and on the effectiveness of the Learning Object in relation to its pedagogical objectives. However, there are many challenges associated with this introduction, including the lack of standardization of the captured metrics, difficulties in standardizing data, and interoperability problems, in addition to the limited reuse of Learning Objects containing such techniques. In this sense, this thesis presents a framework to support the introduction of Learning Analytics techniques in Learning Objects. For its elaboration, systematic reviews on the subject were carried out, a survey with researchers in the area was realized and exploratories prototyping on the application of Learning Analytics techniques in Learning Objects were done. Based on the triangulation of these results, the framework presents an architecture aligned with principles of reuse for production of Learning Objects with Learning Analytics techniques, a set of analytical data that can be obtained from such techniques and a set of relevant practices to be integrated into the Object production process, the latter being an extension of the QPPOA framework (Qualification of the Learning Object Production Process). The use of this framework can assist Learning Object production centers in the introduction of Learning Analytics practices, in order to provide a pedagogical, technical and managerial improvement of this production.