Estrategia de evaluación para la identificación de deshonestidad academia en cursos en línea en ambientes de masividad

Abstract

Today, there is a growing evolution of communication technologies, giving way to the rise of online education [1]. For this reason, new educational modalities have been conceived, one of them is the MOOC, whose acronym stands for Massive Open Online Course; online courses that as its name indicates have as main characteristic that they are massive and open [2], inherit the advantages of traditional e-learning. Thus, they provide the opportunity to present all types of content, and allow access to education at any time and place [3], which leads many universities and traditional education institutions to be interested in adapting it to their educational modality [4]. The MOOC model is gradually being incorporated into universities and their training programmes through strategies such as MPOC (Mass and Private Online Courses), which are a variant that has been successfully applied in various educatio-nal settings. This success and the advantage of having a more controlled environment has led these courses to seek validation as credits within training programs [4]. It is here, where the incursion of these courses as an alternative to obtain academic recognition is present. Therefore, there is a need to control aspects such as possible academic dishonesty, to which these courses are vulnerable because they are develo-ped in online and mass environments [4]. Behaviors such as impersonation, creation of multiple accounts, use of materials or pages that are not permitted, copying or cheating during exams, were combated by implementing measures and proposing a solution that would mitigate such behaviors and take full advantage of the important benefits o ered by this form of education [5], [6], [7], [8]. In this degree work, the results obtained from the application of a mechanism for the implementation of an evaluation strategy on Openedx, which allows the identification of behaviors with suspicion of fraud through Learning Analytics techniques in a case study at MPOC level in the context of the University of Cauca, are evidenced. The mechanism is based on presenting a form with random answers, so that when reading respon-se statistics, patterns of coincidence are sought and fraud behaviors are identified. High expectations are placed on the impact of this proposal and it is hoped that the results obtained will be very useful in this area.