Análise do desempenho de estudantes no ensino híbrido com base nas interações

Abstract

Blended learning is a mode of education that integrates aspects of both face-to-face and online teaching. It has garnered significant interest in the field of education due to the increased flexibility it offers to both students and educators. Its most recent fundamental definition was provided by Garrison & Vaughan in 2008. Thus, it is crucial to delve deeper into the interactions and general characteristics of this mode of education to comprehend its impact on students' academic performance. Consequently, in the present study, a case study was conducted in a course offered in the blended learning model at the Faculty of Economics, Administration, and Accounting (FEAC) at the Federal University of Alagoas (UFAL). To achieve this, data regarding student interactions with the online learning environment and in-person classes and activities were analyzed. Data transformations were applied, and unsupervised machine learning algorithms were utilized to cluster students into different learning profiles, enabling the identification of characteristics for each type of student. Subsequently, visualizations of interactions that had the most influence on students' grades were generated, profiles with different levels of performance were identified, and personalized recommendations were created depending on how interactions occurred in each defined group. To validate the analysis results and the generated feedback, a self-assessment performance form was administered, collecting students' perceptions and suggestions for improvements to be applied in future courses following the blended learning model. Consequently, this study generates valuable insights into student interactions and performance in blended courses. It is concluded that the analyzed course had limited online content variety, recommending the addition of activities with scoring within the virtual learning environment (VLE) and moments in the course that assess student participation in classes, aiming to enhance future analyses in identifying performance profiles and providing feedback to students.